Обсуждение:Мечеть

2020

Мече́ть (ар. ٌمَسْجِد — масджид‎) — молитвенное (богослужебное) архитектурное сооружение.

История

Первоначально в мечетях собирались для совершения общей молитвы, слушания проповедей и обсуждения общественных дел. В них хранили казну, объявляли народу указы, проводили судебные разбирательства. Постепенно социально-политическая функциямечетей ослабевала. Однако и по сей день мечети продолжают быть символом единения мусульман.

Первой мечетью была Кааба [3:96]. По словам пророка Мухаммада ﷺ после Каабы была построена мечеть аль-Акса в аль-Кудсе (Иерусалиме). Эти две мечети, а также мечеть Пророка в Медине являются местами паломничества мусульман. В первые годы зарождения ислама мусульмане совершали молитвы в специально отведенных для этого местах и зданиях в отдаленных районах Мекки. Сам Пророк ﷺ совершал молитвы и в окрестностях Каабы. Затем центром мусульманской общины Мекки был дом аль-Аркама. После того, как Умар ибн аль-Хаттаб принял ислам мусульмане получили возможность совершения намазов у Каабы.

Однако первой мечетью в Мекке считается дом сподвижника пророка Аммара ибн Ясира. Вторую мекканскую мечеть основал Абу Бакр. После совершения хиджры пророк Мухаммад ﷺ распорядился построить мечеть в пригороде Медины в аль-Кубе, относительно которой был ниспослан следующий аят Корана: …мечеть, основанная на богобоязненности с первого дня, — достойнее, чтобы ты в ней стоял [9:108]. Мечеть аль-Куба имела вид огороженного четырёхугольного двора с навесом, крытым пальмовыми листьями. После неё была построена мечеть Пророка в Медине.

В каждом мусульманском городе существуют «соборные» (джума) мечети, служащие для больших коллективных молитв в пятницу и в праздничные дни, а также квартальные мечети. В некоторых странах существуют женские мечети, в которых допускают только женщин.

Архитектура

В разных районах мусульманского мира сложились различные архитектурные типы мечетей. Согласно постепенно сложившейся традиции, представляет собой отдельно стоящее здание с куполом-гамбизом, иногда мечеть имеет внутренний двор (Мечеть Аль-Харам). Мечети имеют один или несколько минаретов, с которых провозглашается призыв к молитве (азан). Молитвенный зал с колоннадой и балконом лишён изображений, но на стенах могут быть начертаны строки из Корана на арабском. Стена, обращённая к Мекке, отмечена пустой нишей, михрабом, в которой молится имам, и в направлении к которой молящиеся должны повернуться лицом. Пол мечети обычно устилается коврами. Справа от михраба расположена кафедра-минбар, с которой проповедник имам читает свои проповеди верующим во время пятничной и праздничной молитвы. Многие мечети имеют открытый двор с водоёмом для ритуального омовения. При мечетях, как правило, работают школы-медресе, в которых обучают основам исламской религии, чтению Корана и т. д..

Не смотря на то, что посещение мечети желательно совершать в состоянии ритуальной чистоты, она не имеет сакрального значения и никак не освящается. Женщины молятся отдельно от мужчин в специально отгороженной части общего зала или на галереях. Для них также оборудуются отдельные места для совершения ритуального омовения. Мечети содержатся на частные пожертвования или государственные субсидии. Служители мечети — имамы, хатибы, муэдзины — выступают лишь как организаторы ритуала, не обладающие святостью и благодатью.

Этикет посещения мечети

В исламе существуют следующие правила посещения мечети:

• Зайдя в мечеть, посетитель должен поприветствовать людей салямом;
• Если в мечети никого нет, то следует сказать: «ас-Саляму ‘аляйни ва ‘аля ‘ибади-л-Ляхи с-салихин».
• Прежде чем сесть, мусульманину следует совершить двухракаатную молитву — тахийят аль-масжид.
• Необходимо оберегать мечеть от грязи.
• Желательно как можно больше поминать Всевышнего Аллаха.

Запрещено:

• Нельзя громко разговаривать, шуметь, толкать людей или другим образом мешать молящимся.
• Обсуждать мирские дела
• Проходить перед совершающим молитву;

По словам аль-Куртуби тому, кто соблюдает эти и другие правила мечети, она станет «местом безопасности и защиты от шайтана» (Тафсир аль-Куртуби).

Цитаты

Известный табиин Саид ибн аль-Мусаяб, рахимахуЛлах, сказал: «Тот, кто сидит в мечети, на самом деле сидит в компании Всевышнего Аллаха» (Тафсир Куртуби, р.6, стр. 560.).

Encyclopaedia of Islam, Second Edition / Masd̲j̲id

---

Article Table of Contents

I. In the central Islamic lands

II. In Muslim India

III. In Java.

V. In China.

VI. In East Africa.

(a.), mosque, the noun of place from sad̲j̲ada “to prostrate oneself, hence “place where one prostrates oneself [in worship]”. The modern Western European words (Eng. mosque, Fr. mosquée, Ger. Moschee, Ital. moschea) come ultimately from the Arabic via Spanish mezquita.

I. In the central Islamic lands

A. The origins of the mosque up to the Prophet’s death.

The word msgdʾ is found in Aramaic as early as the Jewish Elephantine Papyri (5th century B.C.), and appears likewise in Nabataean inscriptions with the meaning “place of worship”, but possibly, originally “stele, sacred pillar”. The Syriac form msgdʾ and Amharic masged are late loans from Arabic, though Geʿez mes̲h̲gād “temple, church” may be a genuine formation from the verbal root s - g- d (itself certainly borrowed from Aramaic). The form ms 1 gd “oratory, place of prayer” occurs also in Epigraphic South Arabian (A. F. L. Beeston et alii, Sabaic dictionary, Louvain-Beirut 1982, 125). The Arabic masd̲j̲id may thus have been taken over directly from Aramaic or formed from the borrowed verb (see A. Jeffery, The foreign vocabulary of the Qur’ān, Baroda 1938, 263-4).

1. The Meccan period. The word is used in the Ḳurʾān especially of the Meccan sanctuary (al-Masd̲j̲id al-ḥarām, sūra II, 139, 144, 145, 187, 192, 214; V, 3; VIII, 34; IX, 7, 19, 28; XVII, 1; XXII, 25; XLVIII, 25, 27); according to later sources, this was already the usage in the Meccan period (ca. al-Yaʿḳūbī, Taʾrīk̲h̲. i. 285, 12). According to tradition, the term al-Masd̲j̲id al-aḳṣā (sūra XVII, 1) means the Jerusalem sanctuary (according to B. Schrieke, in Isl., vi [1915-16], 1; cf. Horovitz, in ibid., ix [1919], 159 ff., the reference is rather to a place of prayer in heaven); and in the legend of the Seven Sleepers, masd̲j̲id means a tomb-sanctuary, probably Christian, certainly pre-Islamic (sūra XVIII, 20). The word is also applied to pre-Islamic sanctuaries, which belong to God and where God is invoked, although Muḥammad was not always able to recognise the particular cult associated with them. It is undoubtedly with this general meaning that the word is used in this verse of the Ḳurʾān: “If God had not taken men under his protection, then monasteries, churches and places of prayer (ṣalawāt) and masād̲j̲id would have been destroyed” (sūra XXII, 41). The word is also used in a ḥadīt̲h̲ of an Abyssinian church (al-Buk̲h̲ārī, Ṣalāt, bāb 48, 54; Muslim, Masād̲j̲id, tr. 3) and in another of Jewish and Christian tomb-sanctuaries (al-Buk̲h̲ārī, Ṣalāt, bāb 55; Muslim, Masād̲j̲id, tr. 3). Even Ibn K̲h̲aldūn can still use the word in the general meaning of a temple or place of worship of any religion (Muḳaddima, faṣl 4, 6 at the end). There is therefore no question of a word of specifically Muslim creation. This is in entire agreement with Muḥammad’s original attitude to earlier religions. Just as Abraham was a Muslim, so David had a masd̲j̲id (al-Ṭabarī, Taʾrīk̲h̲, i, 2408, 7 ff.).

To the Prophet, the Meccan sanctuary always remained the principal mosque, known as Bayt Allāh even before the time of the Prophet. It was a grave charge brought against the Ḳurays̲h̲ in the Meccan period that they drove the believers out of al-Masd̲j̲id al-ḥarām (sūra II, 214; V, 3; VIII, 34; XXII, 25; XLVIII, 25), which was considered all the more unjust as they worshipped the true lord of the sanctuary. To the true God belonged al-masād̲j̲id (sūra LXXII, 18, Meccan); it was therefore an absurdity for the godless to prevent the worship of God in “God’s own mosques” (sūra II, 108). The result was that it was revealed in the year 9/630-1: “It is not right for polytheists to frequent the mosques of God” (sūra IX, 17 f.) and the opponents of the new religion were therefore excluded from the sanctuary. The Sīra agrees with the Ḳurʾān, that the sanctity of al-Masd̲j̲id al-ḥarām to which Muḥammad had been used from childhood was always regarded by him as indisputable. Like other Meccans, he and his followers regularly made the ṭawāf around the Kaʿba and kissed the Black Stone (e.g. Ibn His̲h̲ām, 183, 12 ff.; 239, 8; 251, 15); it is frequently stated that he used to sit in the masd̲j̲id like his fellow-citizens, alone or with a follower or disputing with an opponent (Ibn His̲h̲ām, 233, 16; 251, 15; 252, 14; 259; 260; 294; 18 f.). It is related that he used to perform the ṣalāt between the Yaman corner and the Black Stone, apparently from the narrator’s context, very frequently (Ibn His̲h̲ām, 190, 9 ff.). After his conversion, ʿUmar is said to have arranged that believers performed the ṣalāt unmolested beside the Kaʿba (Ibn His̲h̲ām, 224, 13 f., 17 f.). How strongly Muḥammad felt himself attached to the Arab sanctuary is evident from the fact that he took part in the traditional rites there before the hid̲j̲ra (sūra CVIII, 2); in the year 1/622-3, one of his followers, Saʿd b. Muʿād̲h̲, took part in the pilgrimage ceremonies, and in the year 2/623-4 he himself sacrificed on 10 D̲h̲u ’l-Ḥid̲j̲d̲j̲a on the muṣallā of the Banū Salima. He therefore, here as elsewhere, retained ancient customs where his new teaching did not directly exclude them. But when an independent religion developed out of his preaching, a new type of worship had to be evolved.

In Mecca, the original Muslim community had no special place of worship. The Prophet used to perform the ṣalāt in secret in the narrow alleys of Mecca with his first male follower ʿAlī and with the other earliest Companions also (Ibn His̲h̲ām, 159, 166, 13 ff.). The references are usually to the solitary ṣalāt of the Prophet, sometimes beside the Kaʿba (Ibn His̲h̲ām, 190, 9 ff.), sometimes in his own house (Ibn His̲h̲ām, 203, 6 f.). That the believers often prayed together may be taken for granted; they would do so in a house (cf. Ibn His̲h̲ām, 202). Occasionally also ʿUmar is said to have conducted the ritual prayer with others beside the Kaʿba (Ibn His̲h̲ām, 224) because ʿUmar was able to defy the Ḳurays̲h̲. When the Prophet recited in the mosque the revelation, later abrogated, recognising Allāt, al-ʿUzzā and Manāt, according to the story, not only the believers but also the polytheists present took part in the sud̲j̲ūd (al-Ṭabarī, i, 1192 f.). Abū Bakr is said to have had a private place of prayer (masd̲j̲id) in Mecca in his courtyard beside the gate; the Ḳurays̲h̲, we are told, objected to this because women and children could see it and might be led astray by the emotion aroused (Ibn His̲h̲ām, 246; al-Buk̲h̲ārī, Ṣalāt, bāb 86; Kafāla, bāb 14 etc.; Maẓālim, bāb 22).

In the dogma taught by Muḥammad, a sanctuary was not a fundamental necessity. Every place was the same to God, and humility in the presence of God, of which the ritual prayer was the expression, could be shown anywhere; hence the saying of the Prophet that he had been given the whole world as a masd̲j̲id, while earlier prophets could only pray in churches and synagogues (al-Wāḳidī, tr. Wellhausen, 403; Corpu iuris di Zaid b. ʿAlī, ed. Griffini, 50 and p. clxxix; al-Buk̲h̲ārī, Ṣalāt, bāb 56; Tayammum, bāb 1; Muslim, Masād̲j̲id, tr. 1), and also the saying: “Wherever the hour of prayer overtakes thee, thou shalt perform the ṣalāt and that is a masd̲j̲id” (Muslim, Masād̲j̲id, tr. 1). That he nevertheless remained firmly attached to the traditional sanctuary of the Kaʿba, produced a confusion of thought which is very marked in sūra II, 136 ff. When in Medina he was able to do as he pleased, it must have been natural for him to create a place where he could be undisturbed with his followers and where they could perform the ritual ṣalāt together.

2. The foundation of the Mosque in Medina. According to one tradition, the Prophet came riding into Medina on his camel with Abū Bakr as ridf surrounded by the Banū Nad̲j̲d̲j̲ār. The camel stopped on Abū Ayyūb’s fināʾ. Here (according to Anas) the Prophet performed the ṣalāt, and immediately afterwards ordered the mostque to be built and purchased the piece of land from two orphans, Sahl and Suhayl, who were under the guardianship of Muʿād̲h̲ b. ʿAfrāʾ, for 10 dīnār s, after declining to accept it as a gift; he lived with Abū Ayyūb until the mosque and his houses were completed. During this period he performed the ṣalāt in courtyards or other open spaces (al-Buk̲h̲ārī, Ṣalāt, bāb 48; Muslim, Masād̲j̲id, tr. 1; Aḥmad b. Ḥanbal, Musnad, iii, 212 above; Ibn His̲h̲ām, 336; al-Ṭabarī, i, 1258 f.; al-Masʿūdī, Murūd̲j̲, iv, 140-1 = §1469). According to this tradition, the building of the mosque was intended by the Prophet from the first and the choice of the site was left to the whim of his mount. According to another tradition, the Prophet took up his abode with Abū Ayyūb, but during the first period of his stay in Medina he conducted the ṣalāt in the house of Abū Umāma Asʿad, who had a private masd̲j̲id, in which he used to conduct ṣalāts with his neighbours. The Prophet later expressed the desire to purchase the adjoining piece of ground, and he bought it from the two orphans, who according to this tradition, were wards of Asʿad (al-Balād̲h̲urī, Futūḥ al-buldān, 6; cf. Wüstenfeld, Gesch. d. Stadt Medina, 60). The site was covered with graves, ruins (k̲h̲irab ; also ḥart̲h̲, al-Ṭabarī, i, 1259, 17; 1260, 1; cf. Aḥmad b. Ḥanbal, Musnad, iii, 212, 7, perhaps due to an old misreading) and palm-trees and was used as a place for keeping camels (and smaller domestic animals, al-Buk̲h̲ārī, Wuḍūʾ, bāb 66). The site was cleared, the palms cut down and the walls built. The building material was bricks baked in the sun (labin) (Ibn His̲h̲ām, 337; al-Buk̲h̲ārī, ṣalāt, bābs 62, 65; according to one tradition they were baked at the well of Fāṭima, Wüstenfeld, Stadt Medina, 31); in plan it was a courtyard surrounded by a brick wall on a stone foundation with three entrances; the gateposts were of stone. On the ḳibla side (i.e. the north wall), at first left open, the stems of the palm trees which had been cut down were soon set up as columns and a roof was put over them of palm-leaves and clay. On the east side two huts of similar materials were built for the Prophet’s wives Sawdāʾ and ʿĀʾis̲h̲a; their entrances opened on to the court and were covered with carpets; they were later increased so that there were nine little houses for the Prophet’s wives. When the ḳibla was moved to the south, the arbour at the north wall remained; under this arbour called ṣuffa or ẓulla the homeless Companions (Ahl al-Ṣuffa [q.v.]) found shelter (al-Buk̲h̲ārī, ṣalāt, bābs 48, 62; Wüstenfeld, Medina, 60 f., 66; al-Diyārbakrī, Taʾrīk̲h̲ al-K̲h̲amīs, Cairo 1302, i, 387 ff.; on the ṣuffa, 387 in the middle; 391 after the middle; cf. L. Caetani, Annali dell’ Islām, i, 377 f.). In seven months, the work was completed (Wüstenfeld, Medina, 59), according to others in the month of Ṣafar of the year 2 (Ibn His̲h̲ām, 339, 18 f.). The mosque was very simple. It was really only a courtyard with a wall round it; the ṣuffa already mentioned supplied a shelter on the north side, while on the south side, later the ḳibla side, an arbour was probably built also, for the Prophet used to preach leaning against a palmtrunk and this must have been on the ḳibla side. How large the arbours were cannot be ascertained. The mosque was the courtyard of the Prophet’s houses and at the same time the meeting-place for the believers and the place for common prayer.

According to the sources, it was the Prophet’s intention from the very first to build a mosque at once in Medina; according to a later tradition, Gabriel commanded him in the name of God to build a house for God (al-Diyārbakrī, i, 387 below); but this story is coloured by later conditions. It has been made quite clear, notably by L. Caetani (Annali dell’ Islām, i, 432, 437 ff.) and later by H. Lammens (Moʿâwia, 8, 5, 62; idem, Ziād, 30 ff., 93 ff.) that the earliest masd̲j̲id had nothing of the character of a sacred edifice. Much can be quoted for this view from Ḥadīt̲h̲ and Sīra (cf. Annali dell’ Islām, i, 440). The unconverted T̲h̲aḳafīs were received by the Prophet in the mosque to conduct negotiations and he even put up three tents for them in the courtyard (Ibn His̲h̲ām, 916; al-Wāḳidī-Wellhausen, 382); envoys from Tamīm also went freely about in the mosque and called for the Prophet, who dealt with them after he had finished prayers (Ibn His̲h̲ām, 933 f.; al-Wāḳidī-Wellhausen, 386). Ibn Unays brought to the masd̲j̲id the head of the Hud̲h̲alī Sufyān, threw it down before the Prophet and gave his report (Ibn His̲h̲ām, 981; al-Wāḳidī-Wellhausen, 225). After the battle of Uhud, the Medina chiefs spent the night in the mosque (al-Wāḳidī-Wellhausen, 149). The Awsīs tended their wounded here (ibid., 215 f.; al-Ṭabarī, i, 1491 f.); a prisoner of war was tied to one of the pillars of the mosque (al-Buk̲h̲ārī, Ṣalāt, bāb 76, 82; cf. 75). Many poor people used to live in the ṣuffa (al-Buk̲h̲ārī, Ṣalāt, bāb 58); tents and huts were put up in the mosque, one for example by converted and liberated prisoners, another by the Banū G̲h̲ifār, in whose tent Saʿd b. Muʿād̲h̲ died of his wounds (ibid., bāb 77; Ibn al-At̲h̲īr, Usd al-g̲h̲āba, ii, 297). People sat as they pleased in the mosque or took their ease lying on their backs (al-Buk̲h̲ārī, ʿIlm, bāb 6; Ṣalāt, bāb 85; Ibn Saʿd, i, 124, 14); even so late as the reign of ʿUmar, it is recorded that he found strangers sleeping in a corner of the mosque (al-Mubarrad, Kāmil, 118, 15 ff.); the Prophet received gifts and distributed them among the Companions (Buk̲h̲ārī, Ṣalāt, bāb 42); disputes took place over business (ibid., bābs 71, 83) and in general, people conducted themselves as they pleased. Indeed, on one occasion some Sudanese or Abyssinians with the approval of the Prophet gave a display with shield and lance on the occasion of a festival (ibid., Ṣalāt, bāb 69; ʿĪdayn, bāb 2, 25; D̲j̲ihād, bāb 81); and on another a stranger seeking the Prophet, rode into the mosque on his camel (ibid., ʿIlm, bāb 6). So little “consecrated” was this, the oldest mosque, that one of the Munāfiḳūn or “Hypocrites”, ejected for scoffing at the believers, could call to Abū Ayyūb “Are you throwing me out of the Mirbad Banī T̲h̲aʿlaba ?” (Ibn His̲h̲ām, 362, 10 f.).

All this gives one the impression of the headquarters of an army, rather than of a sacred edifice. On the other hand, the mosque was used from the very first for the general divine worship and thus became something more than the Prophet’s private courtyard. Whatever the Prophet’s intentions had been from the first, the masd̲j̲id, with the increasing importance of Islam, was bound to become very soon the political and religious centre of the new community. The two points of view cannot be distinguished in Islam, especially in the earlier period. The mosque was the place where believers assembled for prayer around the Prophet, where he delivered his addresses, which contained not only appeals for obedience to God but regulations affecting the social life of the community (cf. al-Buk̲h̲ārī, Ṣalāt, bābs 70, 71); from here he controlled the religious and political community of Islam. Even at the real old sanctuaries of Arabia, there were no restrictions on what one could do; what distinguished the mosque from the Christian church or the Meccan temple was that in it there was no specially dedicated ritual object. At the Kaʿba also, people used to gather to discuss every day affairs and also for important assemblies, if we may believe the Sīra (Ibn His̲h̲ām, 183 f., 185, 1, 229, 8, 248, 257, 19). Here also the Prophet used to sit; strangers came to visit him; he talked and they disputed with him; people even came to blows and fought there (Ibn His̲h̲ām, 183-4, 185-6, 187-8, 202, 19, 257, 259; Chron. d. Stadt Mekka, ed. Wüstenfeld, i, 223, 11). Beside the Kaʿba was the Dār al-Nadwa, where important matters were discussed and justice administered (ibid., see index). From the Medina mosque was developed the general type of the Muslim mosque. It depended on circumstances whether the aspect of the mosque as a social centre or as a place of prayer was more or less emphasised.

3. Other mosques in the time of the Prophet. The mosque of the Prophet in Medina was not the only one founded by Muslims in his lifetime, and according to tradition not even the first, which is said to have been the mosque of Ḳubāʾ. In this village, which belonged to the territory of Medina (see Wüstenfeld, Geschichte der Stadt Medina, 126), the Prophet on his hid̲j̲ra stopped with the family of ʿAmr b. ʿAwf; the length of his stay is variously given as 3, 5, 8, 14 or 22 days. According to one tradition, he found a mosque there on his arrival, which had been built by the first emigrants and the Anṣār, and he performed the ṣalāt there with them (see Wüstenfeld, op. cit., 56; al-Balād̲h̲urī, Futūḥ al-buldān, 1; al-Diyārbakrī, i, 380-1). According to another tradition, the Prophet himself founded the mosque on a site, which belonged to his host Kult̲h̲ūm and was used as a mirbad for drying dates or, according to others, to a woman named Labba, who tethered her ass there (Wüstenfeld, Medina, 131; Ibn His̲h̲ām, 335; al-Ṭabarī, i, 1260, 6; Ibn Saʿd, i/1, 6; Masʿūdī, Murūd̲j̲, iv, 139; al-Diyārbakrī, 1, 381; al-Sīra al-Ḥalabiyya, Cairo 1320, ii, 58-9). Out of this tradition arose a legend based on the story of the foundation of the principal mosque in Medina. The Prophet makes (first Abū Bakr and ʿUmar without success, then) ʿAlī mount a camel, and at the place to which it goes builds the mosque with stone brought from the Harra; he himself laid the first stone, and Abū Bakr, ʿUmar and ʿUt̲h̲mān the next ones (al-Diyārbakrī, i, 381). The Prophet is said to have henceforth visited the mosque of Ḳubāʾ every Saturday, either riding or walking, and the pillar is still shown beside which he conducted the service (al-Buk̲h̲ārī, Faḍl al-ṣalāt fī Masd̲j̲id Makka wa ’l-Madīna, bāb 2, 4; Muslim, Ḥad̲j̲d̲j̲, tr. 94; al-Diyārbakrī, i, 382; al-Balād̲h̲ūrī, 5). We are occasionally told that he performed his ṣalāt on the Sabbath in the mosque at Ḳubāʾ when he went to the Banū ’l-Naḍīr in Rabīʿ I of the year 4/625 (al-Wāḳidī-Wellhausen, 161).

It is obvious that the customs and ideas of the later community have shaped the legend of this mosque. The only question is whether the old tradition that the mosque was founded either by the Prophet himself or even before his arrival by his followers is also a later invention. We thus come to the question whether the Prophet founded or recognised any other mosques at all than that of Medīna. Caetani, in keeping with his view of the origin of the mosque, was inclined to deny it, pointing to the fact that there was later an obvious tendency to connect mosques everywhere with the Prophet and that sūra IX, 108, strongly condemns the erection of an “opposition mosque” (Masd̲j̲id al-Ḍirār). The Ḳurʾān passage is as follows: “Those who have built themselves a masd̲j̲id for opposition (ḍirār) and unbelief and division among the believers and for a refuge for him who in the past fought against God and his Prophet; and they swear: We intended only good! God is witness that they are liars! Thou shalt not stand up in it, for verily a masd̲j̲id which is founded on piety from the first day of its existence has more right that thou shouldest stand in it; in it are men who desire to purify themselves” (sūra IX, 108-9). According to tradition, this was revealed in the year 9/630-1; when the Prophet was on the march to Tabūk, the Banū Sālim said to him that they had built a mosque to make it easier for their feeble and elderly people, and they begged the Prophet to perform his ṣalāt in it and thus give it his approval. The Prophet postponed it till his return, but then his revelation was announced, because the mosque had been founded by Munāfiḳūn at the instigation of Abū ʿĀmir al-Rāhib, who fought against the Prophet. According to one tradition (so Ibn ʿUmar, Zayd) the “mosque founded on piety” was that of Medina, from which the people wished to emancipate themselves; according to another (Ibn ʿAbbās), the reference was to that of Ḳubāʾ; Abū ʿĀmir and his followers were not comfortable among the Banū ʿAmr b. ʿAwf and therefore built a new mosque. According to some traditions, it was in D̲h̲ū Awān. The Prophet however had it burned down (al-Ṭabarī, i, 1704-5; Ibn His̲h̲ām, 357-8, 906-7; Ibn Saʿd, i/1, 6; al-Wāḳidī-Wellhausen, 410-11; al-Ṭabarī, Tafsīr, xi, 17 ff.; Wüstenfeld, Medina, 131; al-Sīra al-Ḥalabiyya, ii, 60; al-Balād̲h̲ūrī, 1-2; Muslim, Ḥad̲j̲d̲j̲, bāb 93). If the connection with the Tabūk campaign is correct, the Masd̲j̲id al-Ḍirār is to be sought north of Medina; the “mosque founded on piety” would then be the mosque of Medina rather than that of Ḳubāʾ which lies to the south of it. There is in itself nothing impossible about the rejection in principle of any mosque other than that of Medina. We should then have to discard the whole tradition, for, according to it, the Prophet was at first not unfavourably disposed to the new mosque, and his wrath, according to the tradition, arose from the fact that it had been founded by a refractory party. But as a matter of fact, there are indications that a number of mosques already existed in the time of the Prophet; for example, the verse in the Ḳurʾān, “in houses, which God hath permitted to be built that His name might be praised in them, in them men praise Him morning and evening, whom neither business nor trade restrain from praising God and performing the ṣalāt and the giving of alms”, etc. (sūra XXIV, 36-7). If this revelation, like the rest of the sūra, is of the Medinan period, it is difficult to refer it to Jews and Christians, and this utterance is quite clear: “Observe a complete fast until the night and touch thou them (i.e. women) not while ye are in the mosques” (sūra II, 183). This shows that there were already in the time of the Prophet several Muslim mosques which had a markedly religious character and were recognised by the Prophet.

That there were really public places of prayer of the separate tribes at a very early date is evident from the tradition that the Prophet in the year 2 offered his sacrifice on 10 D̲h̲u ’l-Ḥid̲j̲d̲j̲a/3 June 624 on the muṣallā of the Banū Salima. In addition, there are constant references to private masād̲j̲id where a few believers, like Abū Bakr in Mecca, made a place for prayer in their houses and where others sometimes assembled (al-Buk̲h̲āri, Ṣalāt, bābs 46, 87; Tahad̲j̲d̲j̲ud, bāb 30; cf. also Ad̲h̲ān, bāb 50).

B. The origin of mosques after the time of the Prophet.

1. Chief mosques. What importance the Medina mosque had attained as the centre of administration and worship of the Muslims is best seen from the fact that the first thought of the Muslim generals after their conquests was to found a mosque as a centre around which to gather.

Conditions differed somewhat according as it was a new foundation or an already existing town. Important examples of the first kind are al-Baṣra, al-Kūfa and al-Fusṭāt. Baṣra was founded by ʿUtba b. G̲h̲azwān as winter-quarters for the army in the year 14/635 (or 16/637 or 638). The mosque was placed in the centre with the Dār al-Imāra, the dwelling of the commander-in-chief with a prison and Dīwān in front of it. Prayer was at first offered on the open space, which was fenced round; later, the whole was built of reeds and when the men went off to war the reeds were pulled up and laid away. Abū Mūsā al-As̲h̲ʿarī [q.v.], who later became ʿUmar’s wālī, built the edifice of clay and bricks baked in the sun (labin) and used grass for the roof (al-Balād̲h̲urī, 346-7, 350; Ibn al-Faḳīh, 187-8; Yāḳūt, Buldān, i, 642, 6-9; cf. al-Ṭabarī, i, 2377, 14 ff.). It was similar in Kūfa, which was founded in 17/638 by Saʿd b. Abī Waḳḳās. In the centre was the mosque, and beside it the Dār al-Imāra was laid out. The mosque at first was simply an open quadrangle, ṣaḥn, marked off by a trench round it. The space was large enough for 40,000 persons. It seems that reeds were also used for building the walls here and later Saʿd used labin. On the south side (and only here) there was an arbour, ẓulla, built (cf. al-Balād̲h̲ūrī, 348, i: ṣuffa). The Dār al-Imāra beside the mosque was later by ʿUmar’s orders combined with the mosque (al-Ṭabarī, i, 2481, 12 ff., 2485, 16, 2487 ff., 2494, 14; Yāḳūt, Muʿd̲j̲am, iv, 323, 10 ff.; al-Balād̲h̲uri, 275 ff., cf. Annali dell’ Islām, iii, 846 ff.). The plan was therefore an exact reproduction of that of the mosque in Medina (as is expressly emphasised in al-Ṭabarī, i, 2489, 4 ff.); the importance of the mosque was also expressed in its position, and the commander lived close beside it. There was no difference in al-Fusṭāṭ, which, although there was already an older town here, was laid out as an entirely new camp. In the year 21/642, after the conquest of Alexandria, the mosque was laid out in a garden where ʿAmr had planted his standard. It was 50 d̲h̲irāʿ s long and 30 broad. Eighty men fixed its ḳibla, which, however, was turned too far to the east, and was therefore altered later by Ḳurra b. S̲h̲arīk [q.v.]. The court was quite simple, surrounded by a wall and had trees growing on it; a simple roof is mentioned; it must be identical with the above-mentioned ẓulla or ṣuffa. ʿAmr b. al-ʿĀṣ lived just beside the mosque and around it the Ahl al-Rāya. Like the house of the Prophet, the general’s house lay on the east side with only a road between them. There were two doors in each wall except the southern one (Yāḳūt, Buldān, iii, 898-9; al-Maḳrīzī, K̲h̲iṭaṭ, iv, 4 ff.; Ibn Duḳmāḳ, K. al-Intiṣār, Cairo 1893, 59 ff.; al-Suyūṭī, Ḥusn al-muḥāḍara, i, 63-4; ii, 135-6; cf. Annali dell’ Islām, iv, 554, 557, 563 ff.). We find similar arrangements made in al-Mawṣil in 20/641 (al-Balād̲h̲ūrī, 331-2).

In other cases, the Muslims established themselves in old towns either conquered or surrendered by treaty; by the treaty, they received a site for their mosque (e.g. al-Balād̲h̲ūrī, 116, 14, 147, 2). But the distinction between towns which were conquered and those which were surrendered soon disappeared, and the position is as a rule not clear. Examples of old towns in which the Muslims established themselves are al-Madāʾin, Damascus and Jerusalem. — In Madāʾin, Saʿd b. Abī Waḳḳās after the conquest in 16/637 distributed the houses among the Muslims, and Kisrā’s Īwān was made into a mosque, after Saʿd had conducted the ṣalāt al-fatḥ in it (al-Ṭabarī, i, 2443, 15 f.; 2451, 7 ff.). In Damascus, which was occupied in 14/635 or 15/636 by capitulation, according to tradition, the Church of St. John was divided so that the eastern half became Muslim, from which Muslim tradition created the legend that the city was taken partly by conquest and partly by agreement (al-Balād̲h̲ūrī, 125; Yāḳūt, Buldān, ii, 591; Ibn D̲j̲ubayr, Riḥla, 262; JA, ser. 9, vii, 376, 381, 404). As a matter of fact, however, the Muslims seem to have laid out their own mosque here just beside the church [see dimas̲h̲ḳ ]; and close beside it again was the K̲h̲aḍrāʾ, the commander-in-chief’s palace, from which a direct entrance to the maḳṣūra was later made (al-Muḳaddasī, 159, 4). Conditions here were therefore once more the same as in Medina. But the possibility of an arrangement such as is recorded by tradition cannot be rejected, for there is good evidence of it elsewhere; in Ḥimṣ, for example, the Muslims and Christians shared a building in common as a mosque and church, and it is evident from al-Iṣṭak̲h̲rī and Ibn Ḥawḳal that this was still the case in the time of their common authority, al-Balk̲h̲ī (309/921) (al-Iṣṭak̲h̲rī1, 61, 7 f.; Ibn Ḥawḳal1, 117, 5; al-Muḳaddasī, 156, 15), and a similar arrangement is recorded for Dabīl in Armenia (al-Iṣṭak̲h̲rī1, 188, 3 f.; Ibn Ḥawḳal1, 244, 21; cf. al-Muḳaddasī, 377, 3 f.).

There were special conditions in Jerusalem. The Muslims recognised the sanctuary there, as is evident from the earlier ḳibla and from sūra XVII, 1 (in the traditional interpretation). It must therefore have been natural for the conquerors, when the town capitulated, to seek out the recognised holy place. Indeed, we are told that ʿUmar in the year 17/638 built a mosque in Jerusalem on the site of the temple of Solomon (F. Baethgen, Fragmente syr. u. arab. Hist., 17, 110, following Īs̲h̲ōʿdenaḥ, metropolitan of Baṣra after 700 A.D.; cf. for the 2nd/8th century Theophanes, quoted by Le Strange, Palestine under the Moslems, London 1890, 91 n.). That the Ḳubbat al-Ṣak̲h̲ra [q.v.], which the Mosque of ʿUmar replaced, stands on the old site of the Temple is undoubted. How he found the site is variously recorded [see al-ḳuds ]. The building was, like other mosques of the time of ʿUmar, very simple. Arculf, who visited Jerusalem about 670, says “The Saracens attend a quadrangular house of prayer (domus orationis, i.e. masd̲j̲id) which they have built with little art with boards and large beams on the remains of some ruins, on the famous site where the Temple was once built in all its splendour” (Itinera Hierosolymitana, ed. P. Geyer, 1898, 226-7, tr. P. Mickley, in Das Land der Bibel, ii/2, 1917, 19-20). It is of interest to note that this simple mosque, like the others, was in the form of a rectangle; in spite of its simple character it could hold 3,000 people, according to Arculf.

As late as the reign of Muʿāwiya, we find a new town, al-Ḳayrawān, being laid out on the old plan as a military camp with a mosque and Dār al-Imāra in the centre (Yāḳūt, Muʿd̲j̲am, iv, 213, 10 ff.). As al-Balād̲h̲ūrī, for example, shows, the Muslim conquerors even at a later date always built a mosque in the centre of a newly-conquered town, at first a simple one in each town, and it was a direct reproduction of the simple mosque of the Prophet in Medina. It was the exception to adapt already existing buildings in towns. But soon many additional mosques were added.

2. Tribal mosques and sectarian mosques. There were mosques not only in the towns. When the tribes pledged themselves to the Prophet to adopt Islam, they had also to perform the ṣalāt. It is not clear how far they took part in Muslim worship, but if they concerned themselves with Islam at all, they must have had a Muslim place of meeting. Probably even before Islam they had, like the Meccans, their mad̲j̲lis or nādī or dār s̲h̲ūrā, where they discussed matters of general importance (cf. Lammens, Moʿâwia, 205; Ziād b. Abīhi, 30 ff., 90-1; Le Berceau de l’Islam, 222 ff.). As the mosque was only distinguished from such places by the fact that it was also used for the common ṣalāt, it was natural for tribal mosques to come into existence. Thus we are told that as early as the year 5/626-7 the tribe of Saʿd b. Bakr founded mosques and used an ad̲h̲ān (Ibn Saʿd, i/2, 44, 7, not mentioned in Ibn His̲h̲ām, 943-4; al-Ṭabarī, i, 1722); it is also recorded of the Banū D̲j̲ad̲h̲īma, who lived near Mecca, that they built mosques in the year 8/629-30 and introduced the ad̲h̲ān (al-Wāḳidī-Wellhausen, 351). How far one can rely on such stories in a particular case is however uncertain. A later writer like al-Diyārbakrī says of the Banu ’l-Muṣṭaliḳ that they aslamū wa-banaw masād̲j̲ida (Taʾrīk̲h̲ al-K̲h̲amīs, ii, 132, 20; cf. Annali dell’ Islām, ii, 221); in the early sources, this is not found. Nor is the story told by Ibn Saʿd at all probable, that envoys from the Banū Ḥanīfa received orders to destroy their churches, sprinkle the ground with water and build a mosque (Ibn Saʿd, i/2, 56, 11 ff., while Ibn His̲h̲ām, 945-6, al-Ṭabarī, i, 1737 ff., and al-Balād̲h̲ūrī, 86-7, say nothing about it). But that there were tribal mosques at a very early date is nevertheless quite certain. The mosque at Ḳubāʾ was the mosque of the tribe of ʿAmr b. ʿAwf (Ibn Saʿd, i/1, 6, 6 and cf. above) and according to one tradition, the Banū G̲h̲anm b. ʿAwf were jealous of it and built an opposition mosque (al-Balād̲h̲urī, 3; al-Ṭabarī, Tafsīr, i, 21). A Companion who had taken part in the battle of Badr, ʿItbān b. Malik, complained to the Prophet that he could not reach the masd̲j̲id of his tribe in the rainy season and wanted to build a mosque for himself (al-Buk̲h̲ārī, Ṣalāt, bāb 46; Muslim, Masād̲j̲id, bāb 47). The Prophet himself is said to have visited the masd̲j̲id of the Banū Zurayḳ (al-Buk̲h̲ārī, D̲j̲ihād, bābs 56-8) and in the masd̲j̲id of the Banū Salima during the prayer, there was revealed to him sūra II, 139, which ordered the new ḳibla, wherefore it was called Masd̲j̲id al-Ḳiblatayn (Wüstenfeld, Medina, 62).

The tribal mosque was a sign that the independence of the tribe was still retained under Islam. Indeed, we hear everywhere of tribal mosques, for example, around Medina that of the Banū Ḳurayẓa, of the Banū Ḥārit̲h̲a, of the Banū Ẓafar, of the Banū Wāʾil, of the Banū Ḥarām, of the Banū Zurayḳ (said to have been the first in which the Ḳurʾān was publicly read), that of the Banū Salima, etc. (see Wüstenfeld, Gesch. d. Stadt Medina, 29, 37 ff., 44, 50, 57, 136 ff.); the “mosque of the two ḳibla s” belonged to the Banū Sawād b. G̲h̲anm b. Kaʿb b. Salima (Wüstenfeld, Medina, 41). This then was the position in Medina: the tribes usually had their own mosques, and one mosque was the chief mosque. This was probably the position within the Prophet’s lifetime, for in the earliest campaigns of conquest, mosques were built on this principle. ʿUmar is said to have written to Abū Mūsā in Baṣra telling him to build a mosque li ’l-d̲j̲amāʿa and mosques for the tribes, and on Fridays the people were to come to the chief mosque. Similarly, he wrote to Saʿd b. Abī Waḳḳāṣ in Kūfa and to ʿAmr b. al-ʿĀṣ in Miṣr. On the other hand in Syria, where they had settled in old towns, they were not to build tribal mosques (al-Maḳrīzī, K̲h̲iṭaṭ, iv, 4 below). It is actually recorded that the tribes in each k̲h̲iṭṭa had their own mosques around the mosque of ʿAmr in Fusṭāṭ (cf. Ibn Duḳmāḳ, 62 below 67), and even much later, a tribal mosque like that of the Rās̲h̲ida was still in existence (al-Maḳrīzī, K̲h̲iṭaṭ, 64, 4 ff.). Even in the chief mosque, the tribes had their own places (ibid., 9, 12-10). We have similar evidence from ʿIrāḳ. In Baṣra, for example, there was a Masd̲j̲id Banī ʿUbād (al-Balād̲h̲ūrī, 356, 2), one of the Banū Rifāʿa (Ibn Rusta, 201, 16), one of the Banū ʿAdī (Ibn al-Faḳīh, 191, 4) and one of the Anṣār (cf. Goldziher, Muhammedanische Studien, i, 77, n. 5); in Kūfa we find quite a number, such as that of the Anṣār (al-Ṭabarī, ii, 284, 13 f.), of the ʿAbd al-Ḳays (ibid., ii, 657, 2, 9), of the Banū Duhmān (ibid., 670, 4), of the Banū Mak̲h̲zūm (ibid., 734, 19), of the Banū Hilāl (ibid., 1687,8), of the Banū ʿAdī (ibid., 1703, 4), of the Banū D̲h̲uhl and Banū Ḥud̲j̲r (ibid., 532, 8 f.), of the D̲j̲uhayna (ibid., 533, 8), of the Banū Ḥarām (ibid., iii, 2509, 10), and the ʿAbsīs even had several masād̲j̲id (al-Balād̲h̲ūrī, 278, 12 f., see also 285, and Goldziher, loc. cit.).

During the wars, these tribal mosques were the natural rallying points for the various tribes, the mosque was a mad̲j̲lis, where councils were held (al-Ṭabarī, ii, 532, 6 ff.) and the people were taught from its minbar (ibid., 284); battles often centred for this reason round these mosques (e.g. al-Ṭabarī, ii, 130, 148, 6, 960). “The people of your mosque” ahl masd̲j̲idikum (ibid., 532, 19) became identical with “your party”. Gradually, as new sects arose, they naturally had mosques of their own, just as Musaylima before them is said to have had his own mosque (al-Balād̲h̲ūrī, 90, 4 from below; Ibn Ḥanbal, Musnad, i, 404 below). Thus we read later of the mosques of the Ḥanbalīs in Bag̲h̲dād, in which there was continual riot and confusion (Hilāl al-Ṣābī, Kitāb al-Wuzarāʾ, ed. Amedroz, 335). It sometimes happened that different parties in a town shared the chief mosque (al-Muḳaddasī, 102, 5), but as a rule it was otherwise. In particular, the Sunnīs and S̲h̲īʿīs as a rule had separate mosques (cf. Mez, Die Renaissance des Islâms, 63). It sometimes even happened that Ḥanafīs and S̲h̲āfiʿīs had separate mosques (Yāḳūt, Buldān, iv, 509, 9; al-Muḳaddasī, 323, 11). These special mosques were a great source of disruption in Islam, and we can understand that a time came when the learned discussed whether such mosques should be permitted at all. But the question whether one might talk of the Masd̲j̲id Banī Fulān was answered by saying that in the time of the Prophet, the Masd̲j̲id Banī Zurayḳ was recognised (al-Buk̲h̲ārī, Ṣalāt, bāb 41; cf. D̲j̲ihād, bābs 56-8, and al-Ṭabarī, Tafsīr, xi, 20, after the middle of the page).

3. Adaptation to Islam of older sanctuaries; memorial mosques. According to the early historians, the towns which made treaties with the Muslims received permission to retain their churches (al-Balad̲h̲ūrī, 121, in the middle; al-Ṭabarī, i, 2405, 2407), while in the conquered towns the churches fell to the Muslims without any preamble (cf. al-Balād̲h̲ūrī, 120 below). Sometimes also it is recorded that a certain number of churches were received from the Christians, e.g. fifteen in Damascus according to one tradition (ibid., 124, 8, otherwise on 121; cf. JA, Ser. 9, vii, 403). It is rather doubtful whether the process was such a regular one; in any case, the Muslims in course of time appropriated many churches to themselves. With the mass conversions to Islam, this was a natural result. The churches taken over by the Muslims were occasionally used as dwellings (cf. al-Ṭabarī, i, 2405, 2407); at a later date, it also happened that they were used as government offices, as in Egypt in 146/763 (al-Maḳrīzī, iv, 35; cf. for Kūfa, al-Balād̲h̲ūrī, 286). The obvious thing, however, was to transform the churches taken into mosques. It is related of ʿAmr b. al-ʿĀṣ that he performed the ṣalāt in a church (al-Maḳrīzī, iv, 6) and Zayd b. ʿAlī says regarding churches and synagogues, “Perform thy ṣalāt in them; it will not harm thee” (Corpus iuris di Zaid b. ʿAlī, ed. Griffini, no. 364). It is not clear whether the reference in these cases is to conquered sanctuaries; it is evident, in any case, that the saying is intended to remove any misgivings about the use of captured churches and synagogues as mosques. The most important example of this kind was in Damascus, where al-Walīd b. ʿAbd al-Malik in 86/705 took the church of St. John from the Christians and had it rebuilt; he is said to have offered the Christians another church in its stead (see the references above, in I. B. 1; and also JA, 9 Ser., vii, 369 ff.; Quatremère, Hist. Suit. Maml., ii/1, 262 ff. and art. dimas̲h̲ḳ). He is said to have transformed into mosques a total often churches in Damascus. It must have been particularly in the villages, with the gradual conversion of the people to Islam, that the churches were turned into mosques. In the Egyptian villages there were no mosques in the earlier generations of Islam (al-Maḳrīzī, iv, 28-9, 30). But when al-Maʾmūn was fighting the Copts, many churches were turned into mosques in (ibid., 30). It is also recorded of mosques in Cairo that they were converted churches. According to one tradition, the Rās̲h̲ida mosque was an unfinished Jacobite church, which was surrounded by Jewish and Christian graves (al-Maḳrīzī, iv, 63, 64), and in the immediate vicinity al-Ḥākirn turned a Jacobite and a Nestorian Church into mosques (ibid., 65). When D̲j̲awhar built a palace in al-Ḳāhira, a dayr or monastery was taken in and transformed into a mosque (ibid., 269); similar changes took place at later dates (ibid., 240) and synagogues also were transformed in this way (Masd̲j̲id Ibn al-Bannāʾ, ibid., 265). The chief mosque in Palermo was previously a church (Yāḳūt, Buldān, i, 719). After the Crusades, several churches were turned into mosques in Palestine (Sauvaire, Hist. de Jérus. et d’Hébron, 1876, 7; Quatremère, Hist. Sult. Maml., i/2, 40).

Other sanctuaries than those of the “people of the scripture” were turned into mosques. For example a Masd̲j̲id al-S̲h̲ams between al-Ḥilla and Karbalāʾ was the successor of an old temple of S̲h̲amas̲h̲ (see Goldziher, Muh. Stud., ii, 331). Not far from Iṣṭak̲h̲r was a Masd̲j̲id Sulaymān which was an old fire-temple, the pictures on the walls of which could still be seen in the time of al-Masʿūdī and al-Muḳaddasī (4th/10th century) (al-Masʿūdī, Murūd̲j̲, iv, 77 = § 1403; al-Muḳaddasī, 444). In Iṣṭak̲h̲r itself there was a d̲j̲āmiʿ, which was a converted fire temple (ibid., 436). In Masṣīṣa, the ancient Mopsuestia, al-Manṣūr in 140/797-8 built a mosque on the site of an ancient temple (al-Balād̲h̲urī, 165-6) and the chief mosque in Dihlī was originally a temple (Ibn Baṭṭūṭa, iii, 151); as to Ṭāʾif, cf. Abū Dāwūd, Ṣalāt, bāb 10. Thus in Islam also, the old rule holds that sacred places survive changes of religion. It was especially easy in cases where Christian sanctuaries were associated with Biblical personalities who were also recognised by Islam: e.g., the Church of St. John in Damascus and many holy places in Palestine. One example is the mosque of Job in S̲h̲ayk̲h̲ Saʿd, associated with sūra XXI, 83, XXXVIII, 40; here in Silvia’s time (4th century) there was a church of Job (al-Masʿūdī, i, 91 = § 84; Baedeker, Paläst. u. Syrien 7, 1910, 147).

But Islam itself had created historical associations which were bound soon to lead to the building of new mosques. Even in the lifetime of the Prophet, the Banū Sālim are said to have asked him to perform the ṣalāt in their masd̲j̲id to give it his authority (see above, in I. A. 3). At the request of ʿItbān b. Malik, the Prophet performed the ṣalāt along with Abū Bakr in his house and thereby consecrated it as a muṣallā, because he could not reach the tribal mosque in the rainy season (al-Buk̲h̲ārī, Ṣalāt, bāb 47; Tahad̲j̲d̲j̲ud, bāb 36; Muslim, Masād̲j̲id, tr. 46; a similar story in al-Buk̲h̲ārī, Ad̲h̲ān, bāb 47, Tahad̲j̲d̲j̲ud, bāb 33, is perhaps identical in origin). After the death of the Prophet, his memory became so precious that the places where he had prayed obtained a special importance and his followers, who liked to imitate him in everything, preferred to perform their ṣalāt in such places. But this tendency was only an intensification of what had existed in his lifetime; and so it is not easy to decide how far the above stories reflect later conditions. Mosques very quickly arose on the road between Mecca and Medina at places where, according to the testimony of his Companions, the Prophet had prayed (al-Buk̲h̲ārī, Ṣalāt, bāb 89; al-Wāḳidī-Wellhausen, 421 ff.); the same was the case with the road which the Prophet had taken to Tabūk in the year 9/630-1 (Ibn His̲h̲ām, 907; al-Wāḳidī-Wellhausen, 394; there were 19 in all, which are listed in Annali dell’ Islām, ii-246-7). Indeed, wherever he had taken the field, mosques were built; for example, on the road to Badr, where according to tradition Abū Bakr had built a mosque (al-Wāḳidī-Wellhausen, 39, also Wüstenfeld, Medina, 135). The mosque of al-Faḍīk̲h̲ was built on the spot where the Prophet had prayed in a leather tent during the war with the Banu ’l-Naḍīr in the year 4/625-6 (al-Wāḳidī-Wellhausen, 163; Wüstenfeld, Medina, 132). He is said to have himself built a little mosque in K̲h̲aybar during the campaign of the year 7/628-9 (al-Diyārbakrī, ii, 49-50; cf. Annali dell’ Islām, ii, 19). Outside Ṭāʾif, a mosque was built on a hillock, because the Prophet had performed the ṣalāt there during the siege in the year 8/629-30, between the tents of his two wives, Umm Salama and Zaynab (Ibn His̲h̲ām, 872-3; al-Wāḳidī-Wellhausen, 369); in Liyya, the Prophet is said to have himself built a mosque while on the campaign against Ṭāʾif (Ibn His̲h̲ām, 872; al-Wāḳidī-Wellhausen, 368-9). Mosques arose in and around Medina, “because Muḥammad prayed here” (Wüstenfeld, Gesch. d. Stadt Medina, 31, 38, 132 ff.). It is obvious that in most of these cases, later conditions are put back to the time of the Prophet; in connection with the “Campaign of the Trench” we are told that “he prayed everywhere where mosques now stand” (al-Wāḳidī-Wellhausen, 208). Since, for example, the Masd̲j̲id al-Faḍīk̲h̲ is also called Masd̲j̲id al-S̲h̲ams (Wüstenfeld, Medina, 132), we have perhaps here actually an ancient sanctuary.

Mosques became associated with the Prophet in many ways. In Medina, for example, there was the Masd̲j̲id al-Bag̲h̲la where footprints of the Prophet’s mule were shown in a stone, the Masd̲j̲id al-Id̲j̲āba where the Prophet’s appeal was answered, the Masd̲j̲id al-Fatḥ which recalls the victory over the Meccans, etc. (see Wüstenfeld, Medina, 136 ff.). In Mecca, there was naturally a large number of places sacred through associations with the Prophet and therefore used as places of prayer. The most honoured site, next to the chief mosque, is said to have been the house of K̲h̲adīd̲j̲a, also called Mawlid al-Sayyida Fāṭima, because the daughter of the Prophet was born there. This house, in which the Prophet lived till the hid̲j̲ra, was taken over by ʿAḳīl, ʿAlī’s brother, and bought by him through Muʿāwiya and turned into a mosque (Chroniken d. Stadt Mekka, ed. Wüstenfeld, i, 423; iii, 438, 440). Next comes the house in which the Prophet held his first secret meetings. This was bought by al-K̲h̲ayzurān [q.v.], mother of Hārūn al-Ras̲h̲īd, on her pilgrimage in 171/788 and turned into a mosque (Chron. Mekka, iii, 112, 440). She also purchased the Prophet’s birthplace, Mawlid al-Nabī, and made it into a mosque (ibid., i, 422; iii, 439). If Muʿāwiya really bought the Prophet’s house from his cousin, it was probably the right one; but the demand for places associated with the Prophet became stronger and stronger, and we therefore find more and more places referred not only to the Prophet, but also to his Companions. Such are the birthplaces of Ḥamza, ʿUmar and ʿAlī (Chron. Mekka, iii, 445), and the house of Māriya, the mother of the Prophet’s son, Ibrāhīm (ibid., i, 447, 466), who also had a mosque at Medina (Wüstenfeld, Medina, 133). There were also a Masd̲j̲id K̲h̲adīd̲j̲a (ibid., i, 324) and a Masd̲j̲id ʿĀʾis̲h̲a (ibid., iii, 454), a Masd̲j̲id of the “granted appeal” in a narrow valley near Mecca, where the Prophet performed the ṣalāt (ibid., 453), a Masd̲j̲id al-D̲j̲inn, where the D̲j̲inn overheard his preaching (ibid., i, 424; iii, 453), a Masd̲j̲id al-Raʾya, where he planted his standard at the conquest (ibid., ii, 68 below and 71 above; iii, 13, 453), a Masd̲j̲id al-Bayʿa where the first homage of the Medinans was received (ibid., i, 428; iii, 441). In the Masd̲j̲id al-K̲h̲ayf in Minā is shown the mark of the Prophet’s head in a stone into which visitors also put their heads (ibid., iii, 438). Persons in the Bible are also connected with mosques, Adam, Abraham and Ismāʿīl with the Kaʿba, beside which the Maḳām Ibrāhim is shown, and in ʿArafa there is still a Masd̲j̲id Ibrāhīm (ibid., i, 415, 425) and another in al-Zāhir near Mecca (Ibn D̲j̲ubayr. Riḥla, Leiden-London 1907, 112). To these memorial mosques others were later added, e.g. the Masd̲j̲id Abī Bakr, Masd̲j̲id Bilāl, the Mosque of the Splitting of the Moon (by the Prophet), etc. (see Ibn D̲j̲ubayr, Riḥla, 114 ff.; al-Muḳaddasī, 102-3; Snouck Hurgronje, Mekka, ii, 27; al-Batanūnī, al-Riḥla al-Ḥid̲j̲āziyya 2, Cairo 1329/1911, 52 ff.).

In al-Ḥid̲j̲āz, the Muslims thus acquired a series of mosques which became important from their association with the Prophet, his family and his Companions, and made Muslim history live. On the other hand, in lands formerly Christian, they took over sanctuaries which were associated with the Biblical history which they had assimilated (see Le Strange, Palestine, passim). Other mosques soon became associated with Biblical and Muslim story. The mosque founded by ʿUmar on the site of the Temple in Jerusalem was, as already pointed out, identified as al-Masd̲j̲id al-Aḳṣā mentioned in sūra XVII, 1, and therefore connected with the Prophet’s night journey and the journey to Paradise. The rock is said to have greeted the Prophet on this occasion, and marks in a stone covering a hole are explained as Muḥammad’s footprints (sometimes also as those of Idrīs; cf. Le Strange, Palestine, 136; al-Batanūnī, Riḥla, 165; Baedeker, Palästina, 1910, 52-3; cf. al-Yaʿḳūbī, Taʾrīk̲h̲, ii, 311). The name al-Masd̲j̲id al-Aḳṣā was used throughout the early period for the whole Ḥarām area in Jerusalem, later partly for it, and partly for the building in its southern part (Ibn al-Faḳīh, 100; Sauvaire, Hist. Jérus. et Hébron, 95, 121; cf. Le Strange, Palestine, 96-7). Then there were the mosques which had specifically Muslim associations, like the Masd̲j̲id of ʿUmar on the Mount of Olives where he encamped at the conquest (al-Muḳaddasī, 172).

In Egypt not only was an old Christian sanctuary called Maʿbad Mūsā (al-Maḳrīzī, iv, 269), but we are also told, for example, that the Mosque of Ibn Ṭūlūn was built where Mūsā talked with his Lord (al-Maḳrīzī, iv, 36); according to al-Ḳurdāʿī, there were in Egypt four Masd̲j̲ids of Mūsā (Ibn Duḳmāḳ, ed. Vollers, 92); there was a Masd̲j̲id Yaʿḳūb wa-Yūsuf (al-Muḳaddasī, 200) and a Joseph’s prison, certainly dating from the Christian period (al-Maḳrīzī, iv, 315). There was also a Mosque of Abraham in Munyat Ibn al-K̲h̲aṣīb (Ibn D̲j̲ubayr, 58). The chief mosque of Ṣanʿāʾwas built by Shem, son of Noah (Ibn Rusta, 110). The old temple near Iṣṭak̲h̲r mentioned above was connected with Sulaymān (al-Masʿūdī, Murūd̲j̲, iv, 76-7 =§ 1403; Yāḳūt, i, 299). In the mosque of Kūfa, not only Ibrāhīm but one thousand other prophets and one thousand saints, described as waṣī, are said to have offered their prayers; here was the tree Yaḳṭīn (sūra XXXVII, 146); here died Yag̲h̲ūt̲h̲ and Yaʿūḳ, etc. (Yāḳūt, iv, 325; also Ibn D̲j̲ubayr, 211-12), and in this mosque there was a chapel of Abraham, Noah and Idrīs (Ibn D̲j̲ubayr, 212); a large number of mosques were associated with Companions of the Prophet. What emphasis was laid on such an association is seen, for example, from the story according to which ʿUmar declined to perform the ṣalāt in the Church of the Resurrection in Jerusalem, lest the Church should afterwards be claimed as a mosque.

4. Tomb-mosques. A special class of memorial mosques consisted of those which were associated with a tomb. The graves of ancestors and of saints had been sanctuaries from ancient times and they were gradually adopted into Islam. In addition, there were the saints of Islam itself. The general tendency to distinguish places associated with the founders of Islam naturally concentrated itself round the graves in which they rested. In the Ḳurʾān, a tomh-masd̲j̲id is mentioned in connection with the Seven Sleepers (sūra XVIII, 20) but it is not clear if it was recognised. As early as the year 6/627-8 the companions of Abū Baṣīr are said to have built a mosque at the place where he died and was buried (al-Wāḳidī-Wellhausen, 262). The Prophet is also said to have visited regularly at al-Baḳīʿ in Medina the tombs of martyrs who fell at Uḥud and paid reverence to them (ibid., 143). Whatever the exact amount of truth in the story, there is no doubt that the story of the tomb-mosque of Abū Baṣīr is antedated. The accounts of the death of the Prophet and of the period immediately following reveal no special interest in his tomb. But very soon the general trend of development stimulated an interest in graves, which led to the erection of sanctuaries at them. The progress of this tendency is more marked in al-Wāḳidī, who died in 207/823, than in Ibn Isḥāḳ, who died in 151/768.

The collections of Ḥadīt̲h̲ made in the 3rd/9th century contain discussions on this fact which show that the problem was whether the tombs could be used as places of worship and in this connection whether mosques could be built over the tombs. The ḥadīt̲h̲ s answer both questions in the negative, which certainly was in the spirit of the Prophet. It is said that “ Ṣalāt at the graves (fi ’l-maḳābir) is makrūh ” (al-Buk̲h̲ārī, Ṣalāt, bāb 52); “sit not upon graves and perform not ṣalāt towards them” (Muslim, D̲j̲anāʾiz, tr. 33); “hold the ṣalāt in your houses, but do not use them as tombs” (Muslim, Ṣalāt al-musāfirīn, tr. 28). On the other hand, it is acknowledged that Anas performed the ṣalāt at the cemetery (al-Buk̲h̲ārī, Ṣalāt, bāb 48). We are also told that tombs cannot be used as masād̲j̲id (al-Buk̲h̲ārī, Ṣalāt, bāb 48; D̲j̲anāʾiz, bāb 62). On his deathbed the Prophet is said to have cursed the Jews and the Christians because they used the tombs of their prophets as masād̲j̲id. Ḥadīt̲h̲ explains this by saying that the tomb of the Prophet was not at first accessible (al-Buk̲h̲ārī, Ṣalāt, bāb 48, 55; D̲j̲anāʾiz, bāb 62; Anbiyāʾ, bāb 50; Muslim, Masād̲j̲id, tr. 3); as a matter of fact, its precise location was not exactly known (D̲j̲anāʾiz, bāb 96). The attacks in Ḥadīt̲h̲ insist that tomb-mosques are a reprehensible Jewish practice: “When a pious man dies, they built a masd̲j̲id on his tomb”, etc. (al-Buk̲h̲ārī, Ṣalāt, bāb 48, 54; Muslim, D̲j̲anāʾiz, bāb 71). Although this view of tomb-mosques is still held in certain limited circles (cf. Ibn Taymiyya and the Wahhābis), the old pre-Islamic custom soon also became a Muslim one. The expositors of Ḥadīt̲h̲ like al-Nawawī (on Muslim, Masād̲j̲id, tr. 3, lith. Dihlī 1319, i, 201) and al-ʿAsḳalānī, (Cairo 1329, i, 354) explain the above passages to mean that only an exaggerated taʿ ẓ īm of the dead is forbidden so that tombs should not be used as a ḳibla ; otherwise, it is quite commendable to spend time in a mosque in proximity to a devout man.

The name given to a tomb-mosque is often ḳubba [q.v.] a word which is used of a tent (al-Buk̲h̲ārī, D̲j̲anāʾiz, bāb 62; Ḥad̲j̲d̲j̲, bāb 64; Farḍ al-k̲h̲ums, bāb 19; al-D̲j̲izya, bāb 15; Ṭarafa, Dīwān, vii, 1), but later came to mean the dome which usually covers tombs and thus became the general name for the sanctuary of a saint (cf. Ibn D̲j̲ubayr, Riḥla, 114, 115; cf. Dozy, Supplément, s.v.). Maḳām also means a little chapel and a saint’s tomb (van Berchem, CIA, i, no. 72, etc.; cf. index). The custom of making a ḳubba at the tomb of a saint was firmly rooted in Byzantine territory, where sepulchral churches always had a dome (Herzog-Hauch, Realenzyclopädie 3, x, 784). The usual name however for a tomb-sanctuary was mas̲h̲had ; this is applied to places where saints are worshipped, among Muslim tombs particularly to those of the friends and relations of the Prophet (van Berchem, CIA, i, nos. 32, 63, 417, 544; al-Maḳrīzī, iv, 265, 309 ff.), but also to tombs of other recognised saints, e.g. Mas̲h̲had D̲j̲ird̲j̲is in Mawṣil (Ibn D̲j̲ubayr, Riḥla, 236), etc.

The transformation of the tombs of the Prophet and his near relatives into sanctuaries seems to have been a gradual process. Muḥammad, Abū Bakr andʿUmar are said to have been buried in the house of ʿĀʾis̲h̲a; Fāṭima and ʿAlī lived beside it. ʿĀʾis̲h̲a had a wall built between her room and the tombs to prevent visitors carrying off earth from the tomb of the Prophet. The houses of the Prophet’s wives remained as they were until al-Walīd rebuilt them. He thought it scandalous that Ḥasan b. ʿAlī should live in Fāṭima’s house and ʿUmar’s family close beside ʿĀʾis̲h̲a’s home in the house of Ḥafṣa. He acquired the houses, had all the houses of the Prophet’s wives torn down and erected new buildings. The tombs were enclosed by a pentagonal wall; the whole area was called al-Rawḍa “the garden”; it was not till later that a dome was built over it (Wüstenfeld, Medina, 66 ff., 72-3, 78 ff., 89). In the cemetery of Medina, al-Baḳīʿ [see baḳīʿ al-g̲h̲arḳad ], a whole series of mas̲h̲āhid came to be built where tombs of the family and of the Companions of the Prophet were located (ibid., 140 ff.; Ibn D̲j̲ubayr, Riḥla, 195 ff.). It is often disputed whether a tomb belonged to one or the other (e.g. al-Ṭabarī, iii, 2436, 2). Such tomb-mosques were sacred (muḳaddas ; Ibn D̲j̲ubayr, Riḥla, 114, 13, 17), and they were visited li ’l-baraka. The name al-Rawḍa of the Prophet’s tomb became later applied to other sanctuaries (ibid., 46, 16; 52, 11). Separate limbs were revered in some mosques, like the head of al-Ḥusayn in Cairo, which was brought there in 491/1098 from ʿAsḳalān (ʿAlī Pās̲h̲a Mubārak, al-K̲h̲iṭaṭ al-d̲j̲adīda, iv, 91 ff.; cf. Sauvaire, Hist. Jérus. et Hébr., 16); his head was also revered for some time in the Mas̲h̲had al-Raʾs in Damascus (according to Ibn S̲h̲ākir, JA, ser. 9, vii, 385).

Gradually, a vast number of Muslim tombs of saints came into existence; and to these were added all the pre-Islamic sanctuaries which were adopted by Islam. No distinction can therefore be drawn between tomb-mosques and other memorial mosques. It was often impossible to prove that the tomb in question ever really existed. In the Mas̲h̲had ʿAlī, for example, ʿAlī’s tomb is honoured, but Ibn D̲j̲ubayr leaves it in doubt whether he is really buried there (Riḥla, 212) and many located his grave in the mosque at Kūfa and elsewhere (al-Masʿūdī, Murūd̲j̲, iv, 289, v, 68 = §§ 1612, 1825; Ibn Ḥawḳal 1, 163). In ʿAyn al-Baḳar near ʿAkkā there was also a Mas̲h̲had ʿAlī (Yāḳūt, iii, 759) and also in the Mosque of the Umayyads (Ibn D̲j̲ubayr, 267); on this question, cf. al-Muḳaddasī, 46. Names frequently become confused and transferred. In Mecca, between Ṣafā and Marwa there was a ḳubba, which was associated with ʿUmar b. al-K̲h̲aṭṭāb; but Ibn D̲j̲ubayr says that it should be connected with ʿUmar b. ʿAbd al-ʿAzīz (Riḥla, 115, 11 ff.). In D̲j̲īza there was a Mas̲h̲had Abī Hurayra, where the memory of this Companion of the Prophet was honoured; it is said to have been originally the grave of another Abū Hurayra (Maḳrīzī, i, 335, 19). Wherever S̲h̲īʿīs ruled, there arose numerous tomb-mosques of the Ahl al-Bayt. In Egypt, Ibn D̲j̲ubayr gives a list of 14 men and five women of the Prophet’s family, who were honoured there (Riḥla, 46-7). Islam was always creating new tombs of saints who had been distinguished for learning or asceticism or miracle-working, e.g. the tomb of al-S̲h̲āfiʿī in Cairo and Aḥmad al-Badawī in Ṭanṭa. There were mosques, chiefly old-established sanctuaries, of Biblical and semi-Biblical personages like Rūbīl (Reuben) and Āsiya the wife of Pharaoh (ibid., 46). In and around Damascus were a number of mosques, which were built on the tombs of prophets and unnamed saints (Ibn D̲j̲ubayr, Riḥla, 273 ff.). In Palestine could be seen a vast number of tombs of Biblical personages (cf. Le Strange, Palestine under the Moslems, index, and Conder, in Palestine Explor. Fund, Quarterly Statement, 1871, 89 ff.), usually mosques with a ḳubba.

After the sanctuaries of persons mentioned in the Bible came those of people mentioned in the Ḳurʾān. For example, outside the D̲j̲āmiʿ in ʿAkkā was shown the tomb-mosque of the prophet Ṣāliḥ (Nāṣir-i K̲h̲usraw, Safar-nāma, ed. Schefer, 15, 1, tr. 49), and in Syria that of his son (Ibn D̲j̲ubayr, 46); that of Hūd was also shown near ʿAkkā (Nāṣir-i K̲h̲usraw. 16, 5, tr. 52), farther east, that of S̲h̲uʿayb and of his daughter (ibid., 16, 12, tr. 53); the tomb of Hūd was also pointed out in Damascus and in Ḥadramawt (Yāḳūt ii, 596, 16); then we have peculiarly Muslim saints like D̲h̲u ’l-Kifl, the son of Job (Nāṣir-i K̲h̲usraw, 16, 4, tr. 52). Then there are the sanctuaries of saints who are only superficially Muslim but really have their origins in old popular superstitions, like al-K̲h̲aḍir, who had a mas̲h̲had in Damascus (Yāḳūt, ii, 596, 9), or a saint like ʿAkk, founder of the town of ʿAkkā, whose tomb Nāṣir-i K̲h̲usraw visited outside the town (15, 6 from below, tr. 51). Such tombs were much visited by pious travellers and are therefore frequently mentioned in literature (on mas̲h̲āhid of the kinds mentioned here in ʿIrāḳ, see al-Muḳaddasī, 130; for Mawṣil, etc., ibid., 146). In this way, ancient sanctuaries were turned into mosques, and it is often quite a matter of chance under what names they are adopted by Islam (cf. Goldziher, Muh. Studien, ii, 325 ff.). It therefore sometimes happens that the same saint is honoured in several mosques. Abū Hurayra, who is buried in Medina, is honoured not only in the above-mentioned tomb-mosque in D̲j̲īza but also at various places in Palestine, in al-Ramla and in Yubnā south of Ṭabariyya (K̲h̲alīl ed-Ḍāhiry, Zoubdat Kachf el-Mamâlik ed. P. Ravaisse, 42, 1 from below; Nāṣir-i K̲h̲usraw, 17, 1 from below, no. 59; Yāḳūt, iii, 512, 20; iv, 1007, 12; cf. Symbolae Osloenses Fasc. Supplet., ii [1928], 31). The tomb of the Prophet Jonah is revered not only in the ancient Niniveh but also in Palestine.

Just as the ḳubba, under which the saint lay and the mosque adjoining it were sanctified by him, so vice-versa a ḳubba, and a mosque could cause a deceased person to become considered a saint. It was therefore the custom for the mighty not only to give this distinction to their fathers but also to prepare such buildings for themselves even in their own lifetime. This was particularly the custom of the Mamlūk sultans, perhaps stimulated by the fact that they did not found dynasties in which power passed from father to son. Such buildings are called ḳubba, (van Berchem, CIA, i, nos. 82, 95, 96, 126, 138, etc.), exceptionally zāwiya (ibid., no. 98), frequently turba (ibid., no. 58, 66, 88, 106, 107, 116, etc.); the formula is also found: “this ḳubba, is a turba” (no. 67); the latter word acquired the same meaning as masd̲j̲id, mas̲h̲had, partly saint’s grave and partly sacred site (cf. Ibn D̲j̲ubayr, Riḥla, 114, 196); but this word does not seem to be used of ordinary tomb-mosques, although the distinction between these and mosques in honour of saints often disappeared. In these ḳubba s, the regular recitation of the Ḳurʾān was often arranged and the tomb was provided with a kiswa. The mausoleum might be built in connection with a great mosque and be separated from it by a grille (Yāḳūt, iv, 509, 6 ff.).

5. Mosques deliberately founded. In the early period, the building of mosques was a social obligation of the ruler as representative of the community and the tribes. Very soon a number of mosques came into existence, provided by individuals. In addition to tribal mosques, as already mentioned, there were also sectarian mosques, and prominent leaders built mosques which were the centres of their activity, for example the Masd̲j̲id ʿAdī b. Ḥātim (al-Ṭabarī, ii, 130), the Masd̲j̲id Simāk in Kūfa (ibid., i, 2653), the Masd̲j̲id al-As̲h̲ʿat̲h̲, etc. As old sanctuaries became Islamised, the mosque received more of the character of a sanctuary and the building of a mosque became a pious work; there arose a ḥadīt̲h̲, according to which the Prophet said: “for him who builds a mosque, God will build a home in Paradise”; some add “if he desire to see the face of God” (Corpus iuris di Zaid b. ʿAlī, ed. Griffini, no. 276; al-Buk̲h̲ārī, Salāt, bāb 65; Muslim, Masād̲j̲id, tr. 4; Zuhd, tr. 3; al-Maḳrīzī, iv, 36). Like other sanctuaries, mosques were sometimes built as a result of a revelation in a dream. A story of this kind of the year 557/1162 is given by al-Samhūdī for Medina (Wüstenfeld, Medina, 91); and a similar one of a mosque in Damascus (JA, ser. 9, vii, 384); a mosque was also built out of gratitude for seeing the Prophet (al-Madrasa al-S̲h̲arīfiyya, al-Makrīzī, iv, 209). It was of course particularly an obligation on the mighty to build mosques. Even in the earliest period, the governors took care that new mosques were built to keep pace with the spread of Islam (cf. al-Balād̲h̲urī, 178-9). About the year 390/1000 the governor of D̲j̲ibāl, Badr b. Ḥasanawayh, is said to have built 3,000 mosques and hostels (Mez, Die Renaissance des Islâms, 24, Eng. tr. 27). The collections of inscriptions, as well as the geographical and topographical works, reveal how the number of mosques increased in this way.

In Egypt, al-Ḥākim in the year 403/1012-13 had a census taken of the mosques of Cairo, and these were found to amount to 800 (al-Maḳrīzī, iv, 264); al-Ḳuḍāʿī (d. 454/1062) also counted the mosques, and his figure is put at 30,000 or 36,000 (Yāḳūt, iii, 901; Ibn Duḳmāḳ, ed. Vollers, 92; al-Maḳrīzī, iv, 264), which seems a quite fantastic figure (there is probably a wa-lacking before alf, i.e. 1,036). Ibn al-Mutawwad̲j̲ (d. 730/1330) according to al-Maḳrīzī counted 480, and Ibn Duḳmaḳ (about 800/1398) gives in addition to the incomplete list of d̲j̲āmiʿ s a list of 472 mosques, not including madāris, k̲h̲ānaḳāh s, etc.; the figure given by al-Maḳrīzī is smaller. The fantastic figure of 30,000 for Bag̲h̲dād is found as early as al-Yaʿḳūbī (Buldān, 250). It is also an exaggeration when Ibn D̲j̲ubayr was told in Alexandria that there were 12,000 or 8,000 mosques there (43). In Baṣra, where Ziyād built 7 mosques (Ibn al-Faḳīh, 191), the number also increased rapidly, but here again an exaggerated figure (7,000) is given (al-Yaʿḳūbī, op. cit., 361). In Damascus, Ibn ʿAsākir (d. 571/1176) counted 241 within and 148 outside the city (JA, ser. 9, vii, 383). In Palermo, Ibn Ḥawḳal counted over 300, and in a village above it 200 mosques. In some streets there were as many as 20 mosques within a bowshot of one another; this multiplicity is condemned: everyone wanted to build a mosque for himself (Yāḳūt, i, 719; iii, 409, 410). As a matter of fact, one can almost say that things tended this way; al-Yaʿḳūbī mentions in Bag̲h̲dād a mosque for the Anbārī officials of the tax-office (Buldān, 245), and several distinguished scholars practically had their own mosques. It occasionally happened that devout private individuals founded mosques. In 672/1273-4 Tād̲j̲ al-Dīn built a mosque and a separate chamber in which he performed the ṣalāt alone and meditated (al-Maḳrīzī, iv, 90). The mosques thus founded were very often called after their founders, and memorial and tomb-mosques after the person to be commemorated. Sometimes a mosque is called after some devout man who lived in it (al-Maḳrīzī, iv, 97, 265 ff.) and a madrasa might be called after its head or a teacher (ibid., iv, 235; Yāḳūt, Udabāʾ, vii, 82). Lastly, a mosque might take its name from its situation or from some feature of the building.

6. Al-Muṣallā. In addition to the mosques proper, al-Maḳrīzī mentions for Cairo eight places for prayer (muṣallā) mainly at the cemetery (iv, 334-5). The word muṣallā may mean any place of prayer, therefore also mosque (cf. sūra II, 119; cf. al-Maḳrīzī K̲h̲iṭāṭ, iv, 25, 16; idem, Ittiʿāẓ, ed. Bunz. 91, 17; Yāḳūt, Buldān, iv, 326, 3-5) or a particular place of prayer within a mosque (al-Ṭabarī, i, 2408, 16; al-Buk̲h̲ārī, G̲h̲usl, bāb 17; Ṣalāt, bāb 91). In Palestine, there were many open places of prayer, provided only with a miḥrāb and marked off, but quite in the open (cf. for Tiberias, Nāṣir-i K̲h̲usraw. ed. Schefer, 36). It is recorded of the Prophet that he used to go out at the two festivals (al-Fiṭr and al-Aḍḥā) to the place of prayer (al-muṣallā) of the Banū Sālima. A lance which the Negus of Ethiopia had presented to al-Zubayr was carried in front of him and planted before the Prophet as sutra. Standing in front of it, he conducted the ṣalāt, and then preached a k̲h̲uṭba without a minbar to the rows in front of him (al-Ṭabarī, i, 1281, 14 ff.; al-Buk̲h̲ārī, Ḥayḍ, bāb 6; Ṣalāt, bāb 90; ʿĪdayn, bāb 6). He also went out to the muṣallā for the ṣalāt al-istisḳāʾ (Muslim, Istisḳāʾ, tr. 1). This muṣallā was an open space, and Muḥammad is even said to have forbidden a building on it (Wüstenfeld, Medina, 127 ff.). This custom of performing the ṣalāt on a muṣallā outside the town on the two festivals became sunna. There is evidence of the custom for several towns. In Medina, however, a mosque was later built on the muṣallā (ibid., 128) which also happened in other places. An early innovation was the introduction of a minbar by Marwān (ibid., 128; al-Buk̲h̲ārī, ʿĪdayn, bāb 6). When Saʿd b. Abī Waḳḳāṣ built a mosque in Kisrā’s Īwān in al-Madāʾ in, at the festival in the year 16/637, it was expressly stated that it was sunna to go out to it; Saʿd, however, thought it was a matter of indifference (al-Ṭabarī, i, 2451). Shortly after 300/912-13 a muṣallā outside of Hamad̲h̲ān is mentioned (al-Masʿūdī, Murūd̲j̲, ix, 23 = § 3595). There was al-Muṣallā al-ʿAtīḳ in Bag̲h̲dād; here a dakka was erected for the execution of the Ḳarmaṭian prisoners (al-Ṭabarī, iii, 2244-5; cf. 1659, 18); in Kūfa, several are mentioned (ibid., ii, 628, 16; 1704, 8; iii, 367, 8-368) two in Marw (ibid., ii, 1931, 2; 1964, 19; cf. Nāṣir-i K̲h̲usraw, tr. 274), one in Farg̲h̲āna (Ibn Ḥawḳal1, 393, 11). In Tirmid̲h̲, the muṣallā was within the walls (Ibn Ḥawḳal1, 349, 18) which also happened elsewhere (ibid., 378, 6-377). In Cairo, the two festivals were celebrated on the Muṣallā K̲h̲awlān (a Yemeni tribe) with the k̲h̲aṭīb of the Mosque of ʿAmr as leader: according to al-Ḳuḍāʿī, the festivals were to be celebrated on a muṣallā opposite the hill Yaḥmūm, then on al-Muṣallā al-Ḳadīm where Aḥmad b. Ṭūlūn erected a building in 256/870. The site was several times changed (al-Maḳrīzī, iv, 334-5; cf. al-Muḳaddasī, 200, 14-20). In 302, 306 and 308 the ṣalāt al-ʿīd was performed for the first time in the Mosque of ʿAmr (al-Maḳrīzī, iv, 20, 8 ff.; al-Suyūṭī, Ḥusn al-muḥāḍara, ii, 137 below; Ibn Tag̲h̲rībirdī, ii, 194, 9 ff.). Ibn Baṭṭūṭa notes the custom in Spain (i, 20) and Tunis (i, 22) and also in India (iii, 154). Ibn al-Ḥād̲j̲d̲j̲ (d. 737/1336-7) says that in his time the ceremonies still took place on the muṣallā but condemns the bidaʿ associated with them (K. al-Madk̲h̲al, Cairo 1320, ii, 82). It is also laid down in Muslim law, although not always definitely (see Juynboll, Handbuch d. Islām. Ges., 1910, 127; I. Guidi, Il Muḫtaṣar, i, 1919, 136). The custom seems in time to have become generally abandoned. In the 9th/15th century the Masd̲j̲id Āḳsunḳur was expressly built for the k̲h̲uṭba at the Friday services and at festivals (al-Maḳrīzī, iv, 107, 17).

C. The mosque as the centre for divine worship.

1. Sanctity of the mosque. The history of the mosques in the early centuries of Islam shows an increase in its sanctity, which was intensified by the adoption of the traditions of the church and especially by the permeation of the cult of saints. The sanctity already associated with tombs taken over by Islam was naturally very soon transferred to the larger and more imposing mosques. The expression Bayt Allāh “house of God”, which at first was only used of the Kaʿba came now be applied to any mosque (see Corpus iuris di Zaid b. ʿAlī, no. 48, cf. 156, 983; Chron. Mekka, ed. Wüstenfeld, iv, 164; van Berchem, CIA, i, no. 10, 1. 18; Ibn al-Ḥād̲j̲d̲j̲, K. al-Madk̲h̲al, i, 20, 23; ii, 64, 68; cf. Bayt Rabbihi, ibid., i, 23, 73; ii, 56). The alteration in the original conception is illustrated by the fact that the Mamlūk al-Malik al-Ẓāhir Baybars declined to build a mosque on a place for tethering camels because it was unseemly, while the mosque of the Prophet had actually been built on such a place (al-Maḳrīzī, iv, 91; Abū Dāwūd, Ṣalāt, bāb 22).

In the house of God, the miḥrāb and the minbar enjoyed particular sanctity, as did the tomb, especially in Medina (al-Buk̲h̲ārī, Faḍl al-ṣalāt fī masd̲j̲id Makka wa ’l-Madīna, bāb 5). The visitors sought baraka, partly by touching the tomb or the railing round it, partly by praying in its vicinity; at such places “prayer is heard” (Chron. Mekka, iii, 441, 442). In the Masd̲j̲id al-K̲h̲ayf in Minā, the visitor laid his head on the print of the Prophet’s head and thus obtained baraka (ibid., iii, 438). A mosque could be built on a site, the sanctity of which had been shown by the finding of hidden treasure (al-Maḳrīzī, iv, 75). There were often places of particular sanctity in mosques. In the mosques at Ḳubāʾ and Medina, the spots where the Prophet used to stand at prayer were held to be particularly blessed (al-Balād̲h̲urī, 5; al-Buk̲h̲ārī, Ṣalāt, bāb 91; Wüstenfeld, Medina, 65, cf. 82, 109). In other mosques, places where a saint had sat or where a divine phenomenon had taken place, e.g. in the Mosque of ʿAmr and in the Azhar Mosque (al-Maḳrīzī, iii, 19, 52) or the Mosque in Jerusalem (al-Muḳaddasī, 170), were specially visited. Pious visitors made ṭawāf [see ḥad̲j̲d̲j̲ ] between such places in the mosque (al-Maḳrīzī, iv, 20). Just as in other religions, we find parents dedicating their children to the service of a sanctuary, so we find a Muslim woman vowing her child or child yet unborn to the mosque (al-Buk̲h̲ārī, Ṣalāt, bāb 74; al-Maḳrīzī, iv, 20). The fact that mosques, like other sanctuaries, were sometimes founded after a revelation received in a dream has already been mentioned (see 1. B. 5).

This increase in sanctity had as a natural result that one could no longer enter a mosque at random as had been the case in the time of the Prophet. In the early Umayyad period, Christians were still allowed to enter the mosque without molestation (cf. Lammens, Moʿâwia, 13-14; Goldziher, in WZKM, vi [1892], 100-1). Muʿāwiya used to sit with his Christian physician, Ibn Ut̲h̲āl, in the mosque of Damascus (Ibn Abī Uṣaybiʿa, i, 117). According to Aḥmad b. Ḥanbal, the Ahl al-Kitāb (or Ahl al-ʿAhd) and their servants, but not polytheists, were allowed to enter the mosque of Medina (Musnad, iii, 339, 392). At a later date, entrance was forbidden to Christians and this regulation is credited to ʿUmar (Lammens, op. cit., 13, n. 6). A strict teacher of morality like Ibn al-Ḥād̲j̲d̲j̲ thought it unseemly that the monks who wove the mats for the mosques should be allowed to lay them in the mosque (Madk̲h̲al, ii, 57). Conditions were not always the same. In Hebron, Jews and Christians were admitted on payment to the sanctuary of Abraham until in 664/1265 Baybars forbade it (Quatremère, Hist. Sult. Maml, i/2, 27).

According to some traditions, a person in a state of ritual impurity could not enter the mosque (Abū Dāwūd, Ṭahāra, bāb 92; Ibn Mād̲j̲a, Ṭahāra, bāb 123). In any case, only the pure could acquire merit by visiting the mosque (Muslim, Masād̲j̲id, tr. 49; Corpus iuris di Zaid b. ʿAlī, no. 48), and in a later period it is specially mentioned that the wuḍūʾ cannot be undertaken in the mosque itself (Madk̲h̲al, ii, 47 below) nor could shaving (ibid., 58-9).

It is always necessary to be careful not to spit in a mosque, although some traditions which are obviously closer to the old state of affairs say, “not in the direction of the ḳibla, only to the left!” (al-Buk̲h̲ārī, Ṣalāt, bābs 33-4). The custom of taking off one’s sandals in the mosque is found as early as the time of Abū ʿUbayd (2nd/8th century) (Yāḳūt, Udabāʾ, v, 272, 13-237) and according to Ibn al-Ḥād̲j̲d̲j̲’s Madk̲h̲al (see below) is also mentioned by Abū Dāwūd. Al-Ṭabarī puts the custom back to the time of ʿUmar (i, 2408). That it is based on an old custom observed in sanctuaries is obvious (cf. on the history of the custom, F. Cumont, Fouilles de Doura-Europos, 1926, 60-1). The custom, however, seems not to have been always observed. In the 2nd/8th century in the Mosque of the Umayyads, the shoes were taken off only in the maḳṣūra, because the floor was covered with mats; but in 212/827 an Egyptian superintendent ordered that the mosque should only be entered with bare feet (JA, ser. 9, vii, 211, 217). The visitor on entering should place his right foot first and utter certain prayers with blessings on the Prophet and his family (which Muḥammad is said to have done!) and when he is inside perform two rakʿas (al-Buk̲h̲ārī, Ṣalāt, bāb 47; Tahad̲j̲d̲j̲ud, bāb 25; Muslim, Ṣalāt al-musāfirīn, trs. 12-13; al-Ṭabarī, iii, 2464, 2532). Certain regulations for decent conduct came into being, the object of which was to preserve the dignity of the house of divine service. Public announcements about strayed animals were not to be made, as the Bedouins did in their houses of assembly, and one should not call out aloud and thereby disturb the meditations of the worshippers (al-Buk̲h̲ārī, Ṣalāt, bāb 83; Muslim, Masād̲j̲id, tr. 18; more fully in Madk̲h̲al, i, 19 ff.). One should put on fine clothes for the Friday service, rub oneself with oil and perfume oneself (al-Buk̲h̲ārī D̲j̲umʿa, bābs 3, 6, 7, 19) as was also done with. ṭīb for the Ḥad̲j̲d̲j̲ (al-Buk̲h̲ārī, Ḥad̲j̲d̲j̲. bāb 143).

A question which interested the teachers of morality was that of the admission of women to the mosques. That many did not desire their presence is evident from the ḥadīt̲h̲ that one cannot prevent them as there is no fitna connected with it, but they must not be perfumed (Muslim, Ṣalāt, bāb 29; al-Buk̲h̲ārī, D̲j̲umʿa, bāb 13; cf. Chron. Mekka, iv, 168). Other ḥadīt̲h̲s say they should leave the mosques before the men (al-Nasāʾī, Sahw, bāb 77; cf. Abū Dāwūd, Ṣalāt, bābs 14, 48). Sometimes a special part of the mosque was railed off for them; for example, the governor of Mecca in 256/870 had ropes tied between the columns to make a separate place for women (Chron. Mekka, ii, 197 below). According to some, women must not enter the mosque during their menstruation (Abū Dāwūd, Ṭahāra, bābs 92, 103; Ibn Mād̲j̲a, Ṭahāra, bābs 117, 123). In Medina at the present day, a wooden grille shuts off a place for women (al-Batanūnī, al-Riḥla al-Ḥid̲j̲āziyya, 240). At one time, the women stood at the back of the mosque here (Yāḳūt, Udabāʾ, vi, 400). In Jerusalem there were special maḳsūras for them (Ibn al-Faḳih, 100). Ibn al-Ḥād̲j̲d̲j̲ would prefer to exclude them altogether and gives ʿĀʾis̲h̲a as his authority for this.

Although the mosque became sacred, it could not quite cast off its old character as a place of public assembly, and in consequence, the mosque was visited for many other purposes than that of divine worship. Not only in the time of the Umayyads was considerable business done in the mosques (al-Ṭabarī, ii, 1118; cf. Lammens, Ziād, 98) which is quite in keeping with the ḥadīt̲h̲ (al-Buk̲h̲ārī, Ṣalāt, bābs 70-1) which actually found it necessary to forbid the sale of wine in the mosque (ibid., bāb 73), but Ibn al-Ḥād̲j̲d̲j̲ records with disapproval that business was done in the mosques: women sit in the mosques and sell thread, in Mecca hawkers even call their wares in the mosques. The list given by this author gives one the impression of a regular market-place (Madk̲h̲al, ii, 54). Strangers could always sit down in a mosque and talk with one another (see al-Muḳaddasī, 205); they had the right to spend the night in the mosque; according to some, however, only if there was no other shelter available (Madk̲h̲al, ii, 43 below, 49 above; see below I.D. 1b). It naturally came about that people also ate in the mosque; this was quite common, and regular banquets were even given in them (e.g. al-Maḳrīzī, iv, 67, 121-2; cf. in Ḥadīt̲h̲: Ibn Mād̲j̲a, Aṭʿima, bābs 24, 29; Aḥmad b. Ḥanbal, ii, 106, 10 from below). Ibn al-Ḥād̲j̲d̲j̲ laments that in the Masd̲j̲id al-Aḳṣā people even threw the remains of their repast down in the mosque; animals were brought in, and beggars and water-carriers called aloud in them, etc. (Madk̲h̲al, ii, 53 ff.). It is even mentioned as a sign of the special piety of al-S̲h̲īrāzī (d. 476/1083) that he often brought food into the mosque and consumed it there with his pupils (Wüstenfeld, Der Imâm Schâfiʿī, iii, 298). Gradually, the mosques acquired greater numbers of residents (see below, I.D. 2b). In the Azhar Mosque, it was the custom with many to spend the summer nights there because it was cool and pleasant (al-Maḳrīzī, iv, 54). This was the state of affairs about 800/1398. Similar conditions still prevail in the mosques.

2. The mosque as a place of prayer. Friday mosques. As places for divine worship, the mosques are primarily “houses of which God has permitted that they be erected and that His name be mentioned in them” (sūra XXIV, 36), i.e. for His service demanded by the law, for ceremonies of worship (manāsik), for assemblies for prayer (d̲j̲amāʿat) and other religious duties (cf. Chron. Mekka, iv, 164). The mosques were maʿābid (al-Maḳrīzī, iv, 117, 140). In Medina after a journey, the Prophet went at once to the mosque and performed two rakʿas, a custom which was imitated by others and became the rule (al-Buk̲h̲ārī, Ṣalāt, bābs 59-60; Muslim, Ṣalāt al-musāfirīn, tr. 11; al-Wāḳidī-Wellhausen, 412, 436). In this respect, the mosque played a part in public worship similar to that of the Kaʿba in Mecca at an earlier date and the Rabba sanctuary in Ṭāʾif. The daily ṣalāts, which in themselves could be performed anywhere, became especially meritorious when they were performed in mosques, because they expressed adherence to the community. A ṣalāt al-d̲j̲amāʿa, we are told, is twenty or twenty-five times as meritorious as the ṣalāt of an individual at home or in his shop (Muslim, Masād̲j̲id, tr. 42; Buk̲h̲ārī, Ṣalāt, bāb 87; Buyūʿ, bāb 49). There are even ḥadīt̲h̲s which condemn private ṣalāts: “Those who perform the ṣalāt in their houses abandon the sunna of their Prophet” (Muslim, Masād̲j̲id, tr. 44; but cf. 48 and al-Buk̲h̲ārī, Ṣalāt, bāb 52). If much rain falls, the believers may, however, worship in their houses (al-Buk̲h̲ārī, D̲j̲umʿa. bāb 14). In this connection, a blind man was given a special ruk̲h̲ṣa ; it is particularly bad to leave the mosque after the ad̲h̲ān (Muslim, Masād̲j̲id, tr. 45). It is therefore very meritorious to go to the mosque; for every step a man advances into the mosque, he receives forgiveness of sins, God protects him at the last judgment and the angels also assist him (Muslim, Masād̲j̲id, bābs 49-51; al-Buk̲h̲ārī, Ṣalāt, bāb 87; Ad̲h̲ān, bābs 36, 37; D̲j̲umʿa. bābs 4, 18, 31; Corpus iuris di Zaid b. ʿAlī, nos. 48, 156, 983).

This holds especially of the Friday ṣalāt (ṣalāt al-d̲j̲umʿa), which can only be performed in the mosque and is obligatory upon every free male Muslim who has reached years of discretion (cf. Juynboll, Handbuch, 86; Guidi, Sommario del diritto Malechita, i, 125-6. According to Ibn His̲h̲ām (290), this ṣalāt, which is distinguished by the k̲h̲uṭba, was observed in Medina even before the hid̲j̲ra. This is hardly probable and besides is not in agreement with other ḥadīt̲h̲s (see al-Buk̲h̲ārī, D̲j̲umʿa. bāb 11) but the origin of this divine worship, referred to in sūra LXII, 9, is obscure. The assemblies of the Jews and Christians on a particular day must have formed the model (cf. al-Buk̲h̲ārī. D̲j̲umʿa. bāb 1). Its importance in the earlier period lay in the fact that all elements of the Muslim camp, who usually went to the tribal and particular mosques, assembled for it in the chief mosque under the leadership of the general. The chief mosque, which for this reason was particularly large, was given a significant name. They talk of al-masd̲j̲id al-aʿẓam (al-Ṭabarī, i, 2494; ii. 734, 1701, 1702, Kūfa; al-Balād̲h̲urī, 5; al-Ṭabarī, Tafsīr, xi, 21, centre; ibid, also al-masd̲j̲id al-akbar, Medina; cf. al-masd̲j̲id al-kabīr, al-Yaʿḳūbī, Buldān, 245) or masd̲j̲id al-d̲j̲amāʿa (Yāḳūt, iii, 896, Fusṭāt; also al-Ṭabarī, ii, 1119; Ibn Ḳutayba, Maʿārif, ed. Wüstenfeld, 106). masd̲j̲id li ’l-d̲j̲amāʿa (al-Maḳrīzī, iv, 4); masd̲j̲id d̲j̲āmiʿ (al-Balād̲h̲urī, 289, Madāʾ in; Yāḳūt, i, 643, 647, Baṣra); then masd̲j̲id al-d̲j̲amiʿ (Yāḳūt, iii, 899; iv, 885; Ibn Ḥawḳal 1, 298, 315, 387; al-Yaʿḳūbī, 110, etc.). As an abbreviation we find also al-d̲j̲amāʿa (Yāḳūt, i, 400; Ibn Baṭṭūṭa, iv, 343; cf. masd̲j̲id al-d̲j̲amāʿa, al-Balād̲h̲urī, 348) and especially d̲j̲āmiʿ. As the k̲h̲uṭba was the distinguishing feature, we also find masd̲j̲id al-k̲h̲uṭba (al-Maḳrīzī, iv, 44, 64, 87), d̲j̲āmiʿ al-k̲h̲uṭba (ibid., iv. 55) or masd̲j̲id al-minbar, al-Muḳaddasī, 316, for d̲j̲āmiʿ, 1. 8).

Linguistic usage varied somewhat in course of time with conditions. In the time of ʿUmar there was properly in every town only one masd̲j̲id d̲j̲āmiʿ for the Friday service. But when the community became no longer a military camp and Islam replaced the previous religion of the people, a need for a number of mosques for the Friday service was bound to arise. This demanded mosques for the Friday service in the country, in the villages on the one hand and several Friday mosques in the town on the other. This meant in both cases an innovation, compared with old conditions, and thus there arose some degree of uncertainty. The Friday service had to be conducted by the ruler of the community, but there was only one governor in each province; on the other hand, the demands of the time could hardly be resisted and, besides, the Christian converts to Islam had been used to a solemn weekly service.

As to the villages (al-ḳurā), ʿAmr b. al-ʿĀṣ in Egypt forbade their inhabitants to celebrate the Friday service for the reason just mentioned (al-Maḳrīzī, iv, 7). At a later period, then, the k̲h̲uṭba was delivered exceptionally, without minbar and only with staff, until Marwān b. Muḥammad in 132/749-50 introduced the minbar into the Egyptian ḳurā also (ibid., 8). Of a mosque in which a minbar had been placed, we are told d̲j̲uʿila masd̲j̲idan li ’l-aʿyān (al-Ṭabarī, i, 2451) and a village with a minbar is called ḳarya d̲j̲āmiʿa (al-Buk̲h̲ārī, D̲j̲umʿa. bāb 15; cf. madīna d̲j̲āmiʿa, Ibn Ḥawḳal 1, 321), an idea which was regarded by al-Buk̲h̲ārī (d. 256/870) as quite obvious. In introducing the minbars into the Egyptian villages, Marwān was apparently following the example of other regions. In the 4th/10th century, Ibn Ḥawḳal mentions a number of manābir in the district of Iṣṭak̲h̲r (1st edn., 182 ff.) and a few in the vicinity of Marw (ibid., 316) and in Transoxania (ibid., 378; cf. 384), and al-Muḳaddasī does the same for other districts of Persia (309, 317) and he definitely says that the ḳurā of Palestine are d̲h̲āt manābir (ibid., 176; cf. al-Iṣṭak̲h̲rī 1. 58); al-Balād̲h̲urī (331) also uses the name minbar for a village mosque built in 239/853-4; in general, when speaking of the ḳurā, one talks of manābir and not of d̲j̲awāmiʿ (cf. al-Iṣṭak̲h̲rī1, 63). Later, however, the term masd̲j̲id d̲j̲āmiʿ is used for a Friday mosque (Ibn D̲j̲ubayr, 217). The conditions of primitive Islam are reflected in the teaching of the Ḥanafīs, who only permit the Friday service in large towns (cf. al-Māwardī, al-Aḥkām al-ṣulṭāniyya, ed. Enger, 177).

As to the towns, the S̲h̲āfiʿīs on the other hand have retained the original conditions, since they permit the Friday service in only one mosque in each town (see d̲j̲umʿa and op. cit., 178-8), but with the reservation that the mosque is able to hold the community. The distinction between the two rites was of importance in Egypt. When in 569/1173-4 Ṣalāḥ al-Dīn became supreme in Egypt, he appointed a S̲h̲āfiʿī chief ḳāḍī and the Friday service was therefore held only in the al-Ḥākim mosque, as the largest; but in 665/1266, al-Malik al-Ẓāhir Baybars gave the Ḥanafīs preference, and many mosques were therefore used as Friday mosques (al-Maḳrīzī, iv, 52 ff.; al-Suyūṭī, Ḥusn al-muḥādara, ii, 140; Quatremère, Hist. Sult. Maml. i/2, 39 ff). During the Umayyad period, the number of d̲j̲awāmiʿ in the towns were still very small. The geographers of the 3rd/9th and 4th/10th centuries in their descriptions of towns as a rule mention only “the d̲j̲āmiʿ ”. Ibn al-Faḳīh, ca. 290/903, sometimes says masd̲j̲id d̲j̲āmiʿ wa-minbar (304-6, also minbar simply, 305). In keeping with the oldest scheme of town planning, it was very often in the middle of the town surrounded by the business quarters (Ibn Ḥawḳal1, 298, 325; al-Muḳaddasī, 274-5, 278, 298, 314, 316, 375, 376, 413, 426, 427, etc.; Nāṣir-i K̲h̲usraw, ed. Schefer, 35, 41, 56) and the dār al-imāra was still frequently in the immediate vicinity of the chief mosque (Ibn Ḥawḳal1, 298, 314; al-Muḳaddasī, 426).

Al-Iṣṭak̲h̲rī mentions as an innovation in Islam that al-Ḥad̲j̲d̲j̲ād̲j̲ built a d̲j̲āmiʿ in al-Wāsiṭ on the west bank, although there was already one on the east bank (al-Iṣṭak̲h̲rī,1 82-3; cf. al-Yaʿḳubī, Buldān, 322). Ibn D̲j̲ubayr (Riḥla, 211) mentions only one d̲j̲āmiʿ in Kūfa, called Masd̲j̲id al-Kūfa by Ibn al-Faḳīh, although he also mentions other mosques (173; cf. 174, 183 and al-Muḳaddasī, 116). In Baṣra, where al-Yaʿḳūbī (278/891) already mentions 7,000 mosques (Buldān, 361), al-Muḳaddasī (375/985) gives 3 d̲j̲awāmiʿ (117). In Sāmarrāʾ, among many mosques, there was one d̲j̲āmiʿ (al-Yaʿḳūbī, Buldān, 258, 259), which was later replaced by another (ibid., 260-1); al-Mutawakkil also built one outside the original town (ibid., 265; see also P. Schwarz, Die ʿAbbāsiden-Residenz Sāmarrā, 1909, 32). In Bag̲h̲dād, al-Yaʿḳūbī mentions only one d̲j̲āmiʿ for the eastern town and for the western (Buldān, 240, 245, 251, 253; the almost contemporary Ibn Rusta just mentions the old western town and its d̲j̲āmiʿ, 109) although he gives the fantastic figures of 15,000 mosques in the east town (ibid., 254) and 30,000 in the west (or in the whole town?, ibid, 250). After 280/893-4 there was added the d̲j̲āmiʿ of the eastern palace of the caliph (Mez, Renaissance, 388, Eng. tr. 410, quoting al-K̲h̲aṭīb al-Bag̲h̲dādī, Taʾrīk̲h̲ Bag̲h̲dād a private d̲j̲āmiʿ of Hārūn al-Ras̲h̲īd in the Bustān Umm Mūsā is mentioned by Ibn al-Ḳifṭī, Taʾrīk̲h̲ al-Ḥukamāʾ, ed. Lippert, 433 below). These three d̲j̲awāmiʿ are mentioned about 340/951 by al-Iṣṭak̲h̲rī (84), who also mentions one in the suburb of Kalwād̲h̲ā. Ibn Ḥawḳal in 367/977 mentions the latter and also the D̲j̲āmiʿ al-Barāt̲h̲ā (164-5, of 329/940-1; Mez, loc. cit.), a fifth was added in 379/989, a sixth in 383/993 (Mez, 389, Eng. tr. 410-11); thus al-K̲h̲aṭīb al-Bag̲h̲dādī in 460 (1058 gives 4 for West Bag̲h̲dād, 2 for the east town (cf. Le Strange, Baghdad, 324). Ibn D̲j̲ubayr in 581/1185 gives in the east town 3, and 11 d̲j̲awāmiʿ (Riḥla, 228-9) for the whole of Bag̲h̲dād. For Cairo, al-Iṣṭak̲h̲rī gives two d̲j̲āmiʿ s: the ʿAmr and Ṭūlūn Mosques (49) besides that in al-Ḳarāfa, which was regarded as a separate town (cf. Ibn Rusta [ca. 290/903], 116-17). Al-Muḳaddasī, who wrote (375/985) shortly after the Fāṭimid conquest, mentions the ʿAmr mosque (al-Azhar), also one in al-D̲j̲azīra, in D̲j̲īza and in al-Ḳarāfa (198-200, 209; the d̲j̲āmiʿ in al-D̲j̲azīra, also D̲j̲āmiʿ Miḳyās [cf. al-Maḳrīzī, iv, 75] is mentioned in an inscription of the year 485/1092; see van Berchem, CIA, i, no. 39). As these places were all originally separate towns, the principle was not abandoned that each town had only one d̲j̲āmiʿ. The Fāṭimids, however, extended the use of Friday mosques and, in addition to those already mentioned, used the D̲j̲āmiʿ al-Ḥākim, al-Maḳs and Rās̲h̲ida (al-Maḳrīzī, iv, 2-3). Nāṣir-i K̲h̲usraw in 439/1047 mentions in one passage the d̲j̲awāmiʿ of Cairo, in another seven for Miṣr and fifteen in all (ed. Schefer, 134-5, 147). This was altered in 569/1173-4 by Ṣalāḥ al-Dīn (see above), but the quarters, being still regarded as separate towns, retained their own Friday mosques (cf. for the year 607/1210-11 in al-Ḳarāfa, al-Maḳrīzī, iv, 86).

After the Friday worship in Egypt and Syria was freed from restriction, the number of d̲j̲awāmiʿ increased very much. Ibn Duḳmāḳ (ca. 800/1397-8) gives a list of only eight d̲j̲awāmiʿ in Cairo (ed. Vollers, 59-78), but this list is apparently only a fragment (in all, he mentions something over twenty in the part of his book that has survived); al-Maḳrīzī (d. 845/1442) gives 130 d̲j̲awāmiʿ (iv, 2 ff.). In Damascus, where Ibn D̲j̲ubayr still spoke of “the d̲j̲āmiʿ al-Nuʿaymī (d. 927/1521) gives twenty d̲j̲awāmiʿ (JA, ser. 9, vii, 231 ff.), and according to Ibn Baṭṭūṭa, there were in all the villages in the region of Damascus masād̲j̲id d̲j̲āmiʿa (i, 236). The word d̲j̲āmiʿ in al-Maḳrīzī always means a mosque in which the Friday worship was held (vi, 76, 115 ff.), but by his time this meant any mosque of some size. He himself criticises the fact that since 799/1396-7 the ṣalāt al-d̲j̲umʿa was performed in al-Aḳmar, although another d̲j̲āmiʿ stood close beside it (iv, 76; cf. also 86).

The great spread of Friday mosques was reflected in the language. While inscriptions of the 8th/14th century still call quite large mosques masd̲j̲id, in the 9th/15th most of them are called d̲j̲āmiʿ (cf. on the whole question, van Berchem, CIA, i, 173-4); and while now the madrasa [q.v.] begins to predominate and is occasionally also called d̲j̲āmiʿ, the use of the word masd̲j̲id becomes limited. While, generally speaking, it can mean any mosque (e.g. al-Maḳrīzī, iv, 137, of the Muʾayyad mosque), it is more especially used of the smaller unimportant mosques. While Ibn Duḳmāḳ gives 472 masād̲j̲id in addition to the d̲j̲awāmiʿ, madāris, etc., al-Maḳrīzī only gives nineteen, not counting al-Ḳarāfa, which probably only means that they were of little interest to him. D̲j̲āmiʿ is now on the way to become the regular name for a mosque of any size, as is now the usage, in Egypt and Turkey at least. In Ibn al-Ḥād̲j̲d̲j̲ (d. 737/1336-7), al-d̲j̲awāmiʿ is occasionally used in this general meaning in place of al-masād̲j̲id (Madk̲h̲al, ii, 50). Among the many Friday mosques, one was usually distinguished as the chief mosque; we therefore find the expression al-d̲j̲āmiʿ al-aʿẓam (Ibn Baṭṭūṭa, ii, 54, 94; cf. the older expression al-masd̲j̲id al-aʿẓam, in ibid., ii, 53). The principal d̲j̲āmiʿ decided on such questions as the beginning and ending of the fast of Ramaḍān (Madk̲h̲al, ii, 68).

3. Other religious activities in the mosque. “The mentioning of the name of God” in the mosques, was not confined only in the official ritual ceremonies. Even in the time of the Prophet, we are told that he lodged T̲h̲aḳafī delegates in the mosque so that they could see the rows of worshippers and hear the nightly recitation (al-Wāḳidī-Wellhausen, 382). Although this story (which is not given in Ibn His̲h̲ām, 916) may simply be a reflection of later conditions, the recitation of the Ḳurʾān must have come to be considered an edifying and pious work at quite an early date. In the time of al-Muḳaddasī, the ḳurraʾ of Naysābūr used to assemble on Fridays in the d̲j̲āmiʿ in the early morning and recite till the ḍuḥā, (328), and the same author tells us that in the Mosque of ʿAmr in Egypt the aʾimmat al-ḳurrāʾ sat in circles every evening and recited (205). In the time of Ibn D̲j̲ubayr, there were recitations of the Ḳurʾān in the Umayyad mosque after the ṣalāt al-ṣubḥ and every afternoon after the ṣalāt al-ʿaṣr (Riḥla, 271-2). Besides the recitation of the Ḳurʾān, there were praises of God, etc., all that which is classed as d̲h̲iḳr, and which was particularly cultivated by Ṣūfism. This form of worship also took place in the mosque. The ahl al-tawḥīd wa ’l-maʿrifa formed mad̲j̲ālis al-d̲h̲ikr, and assembled in the mosques (al-Makkī, Ḳūt al-ḳulūb, i, 152). In the Mosque of the Umayyads and other mosques of Damascus, d̲h̲ikr was held during the morning on Friday (al-Maḳrīzī, iv, 49). In the Masd̲j̲īd al-Aḳṣā the Ḥanafīs held d̲h̲ikr, and recited at the same time from a book (al-Muḳaddasī, 182). In Egypt, Aḥmad b. Ṭūlūn and K̲h̲umāwarayh allowed twelve men quarters in a chamber near the minaret in order to praise God, and during the night, four of them took turns to praise God with recitations of the Ḳurʾān and with pious ḳaṣīdas. From the time of Ṣalāḥ al-Dīn, an orthodox ʿaḳīda was recited by the muʾad̲h̲d̲h̲ins in the night (al-Maḳrīzī, iv, 48). Ibn al-Ḥād̲j̲d̲j̲ demands that the recitation of the Ḳurʾān aloud should take place in a mosque for the special purpose (masd̲j̲id mad̲j̲hūr), as otherwise pious visitors are disturbed (Madk̲h̲al, ii, 53, 67). Mosques and, in particular, mausoleums, had as a rule regularly-appointed reciters of the Ḳurʾān. In addition there was, e.g. in Hebron and in a mosque in Damascus, a s̲h̲ayk̲h̲ who had to read al-Buk̲h̲ārī (or also Muslim) for three months (Sauvaire, Hist. Jérus. et Hébr., 17; JA, ser. 9, iii. 261). In Tunis, al-Buk̲h̲ārī was read daily in a hospital (al-Zarkas̲h̲ī, tr. Fagnan, Rec. Soc. Arch. Constantine [1894], 188).

Sermons were not only delivered at the ṣalāt al-d̲j̲umʿa. In ʿIrāḳ, even in al-Muḳaddasī’s time, one was preached every morning, according to the sunna of Ibn ʿAbbās (130), it was said. Ibn D̲j̲ubayr, in the Niẓāmiyya in Bag̲h̲dād, heard the S̲h̲āfiʿī raʾīs preach from the minbar on Friday after the ʿaṣr. His sermon was accompanied by the skilled recitations of the ḳurrāʾ who sat on chairs; these were over twenty in number (Ibn D̲j̲ubayr, 219-22). In the same way, the calls of the muʾad̲h̲d̲h̲ins to prayer to the Friday k̲h̲uṭba were delivered to a musical accompaniment (see below, I. H. 4). The unofficial sermons, which moreover were not delivered in mosques alone, were usually delivered by a special class, the ḳuṣṣāṣ (pl. of ḳāṣṣ) (on these, cf. Goldziher, Muh. Stud., ii, 161 ff.; Mez, Die Renaissance des Islâms, 314 ff.; and ḳāṣṣ). The ḳuṣṣāṣ, who delivered edifying addresses and told popular stories, were early admitted to the mosques.

Tamīm al-Dārī is said to have been the first of these; in Medina in the caliphate of ʿUmar before the latter’s decease, he used to deliver his orations at the Friday ṣalāt, and under ʿUmar he was allowed to talk twice a week in the mosque; in the reign of ʿAlī and of Muʿāwiya the ḳuṣṣāṣ were employed to curse the other side (al-Maḳrīzī, iv, 16-7). In the Mosque of ʿAmr in Cairo, by the year 38/658-9 or 39/659-60 a ḳāṣṣ was appointed, named Sulaym b. ʿItr al-Tud̲j̲ībī, who was also ḳāḍī (ibid., iv, 17, wrongly: Sulaymān; al-Kindī, Governors and judges, ed. Guest, 303-4). There are other occurrences of the combination of the two offices (Ibn Ḥud̲j̲ayra [d. 83/702], al-Kindī, 317; K̲h̲ayr b. Nuʿaym in the year 120/738, ibid., 348; cf. al-Suyūṭī, Ḥusn al-muḥāḍara, i, 131, D̲j̲abr, according to T̲h̲awba b. Nimr, Ḥusn, i, 130 below; Ibrāhīm b. Isḥāḳ al-Ḳārī [d. 204], Kindī, 427; see also al-Maḳrīzī, iv, 18), which shows that the office of ḳāṣṣ was quite an official one. There is also evidence of the employment of ḳuṣṣāṣ in the mosques of ʿIrāḳ in the ʿAbbāsid period (Yāḳūt, Udabāʾ iv, 268, v, 446). The ḳāṣṣ read from the Ḳurʾān standing and then delivered an explanatory and edifying discourse, the object of which was to instil the fear of God into the people (al-Maḳrīzī, iv, 18). Under the Fāṭimids also, the ḳuṣṣāṣ were appointed to the mosques; for example in 403/1012-13 the imām undertook the office in the Mosque of ʿAmr (al-Maḳrīzī, iv, 18, below) and the rulers had also a ḳāṣṣ in the palace. The ḳuṣṣāṣ were called aṣḥāb al-karāsī, because they delivered their discourses on the kursī (al-Makkī, Ḳūt al-ḳulūb, i, 152; Ibn al-Ḥād̲j̲d̲j̲, Madk̲h̲al, i, 159; cf. al-Maḳrīzī, iv, 121). Their discourse was called d̲h̲ikr or waʿẓ or mawʿiẓa, whence the ḳāṣṣ was also called mud̲h̲akkir (al-Muḳaddasī, 205) or wāʿiẓ. Specimens of their discourses are given by Ibn ʿAbd Rabbihi (al-ʿIḳd al-farīd, Cairo 1321/1903, i, 294 ff.). It was not only the appointed officials who delivered such discourses in the mosque. Ascetics made public appearances in various mosques and collected interested hearers around them (cf. e.g. al-Maḳrīzī, iv, 135). In the D̲j̲āmiʿ al-Ḳarāfa, a whole society, the Banū D̲j̲awharī, delivered waʿẓ discourses from a kursī for three months on end; their servant collected money in a begging-bowl during the discourse, and the s̲h̲ayk̲h̲ distributed some of it among the poor (ibid., iv, 121).

The ḳaṣaṣ was completely taken over by popular Ṣūfism and later writers would hardly reckon, as al-Makkī does, the “story-tellers” among the mutakallimūn (Ḳūt al-ḳulūb, i, 152). The whole system degenerated to trickery and charlatanry of all kinds, as may be seen in the Maḳāma [q.v.] literature (cf. thereon Yāḳūt, Udabāʾ, vi, 167-8, and see also Mez and Goldziher, op. cit.). Al-Maḳrīzī therefore distinguishes between al-ḳaṣaṣ al-k̲h̲āṣṣa, the regular and seemly edifying discourse in the mosque, and al-ḳaṣaṣ al-ʿāmma, which consisted in the people gathering round all kinds of speakers, which is makrūh (iv, 17). Others also have recorded their objections to the ḳuṣṣāṣ. Ibn al-Ḥād̲j̲d̲j̲ utters a warning against them and wants to forbid their activities in the mosque completely, because they deliver “weak” narratives (Madk̲h̲al, i, 158-9; ii, 13-14, 50). He says that Ibn ʿUmar, Malik and Abū Dāwūd rejected them and ʿAlī ejected them from the masd̲j̲id of Baṣra. It is of little significance that al-Muʿtaḍid in 284/897 forbade people to gather round them, for he issued a similar interdict against the fuḳahāʾ and the reasons were evidently political (al-Ṭabarī, iii, 2165); it was for political reasons also, but with a very different motive, that ʿAḍud al-Dawla forbade their appearing publicly in Bag̲h̲dād because they increased the tension between Sunnīs and S̲h̲īʿis (Mez, op. cit., 319). As late as 580/1184, the wuʿʿāẓ still flourished in the mosques of Bag̲h̲dād, as is evident from the Riḥla of Ibn D̲j̲ubayr (219 ff., 224), and in the 9th/15th century there was in the Azhar mosque a mad̲j̲lis al-waʿẓ as well as a ḥalaḳ al-d̲h̲ikr (al-Maḳrīzī, iv, 54).

When Ibn al-Ḥād̲j̲d̲j̲ denounces speaking aloud in the mosque, it is in the interest of the pious visitors who are engaged in religious works and meditation. Iʿtikāf [q.v.], retirement to a mosque for a period, was adopted into Islam from the older religions.

The word ʿaḳf means in the Ḳurʾān the ceremonial worship of the object of the cult (sūra VII, 134; XX, 93, 97; XXI, 53; XXVI, 71; cf. al-Kumayt, Hās̲h̲imiyyāt, ed. Horovitz, 86, 15) and also the ritual stay in the sanctuary, which was done for example in the Meccan temple (sūra II, 119; XXII, 25). In this connection, it is laid down in the Ḳurʾān that in the month of Ramaḍān believers must not touch their wives “while ye pass the time in the mosques” (ʿākifūn fi ’l-masād̲j̲id, sūra II, 183), an expression which shows, firstly, that there were already a number of mosques in the lifetime of the Prophet, and secondly, that these had already to some extent taken over the character of the temple. The connection with the early period is evident from a ḥadīt̲h̲, according to which the Prophet decides that ʿUmar must carry out a vow of iʿtikāf for one night in the Masd̲j̲id al-Ḥarām made in the D̲j̲āhiliyya (al-Buk̲h̲ārī, Iʿtikāf, bāb 5, 15-16; Farḍ al-k̲h̲ums, bāb 19; Mag̲h̲āzī, bāb 54; Aymān wa ’l-nud̲h̲ār, bāb 29). It is completely in keeping with this that the Prophet, according to the ḥadīt̲h̲, used to spend ten days of the month of Ramaḍān in iʿtikāf in the mosque of Medina (al-Buk̲h̲ārī. Iʿtikāf, bābʿ ; Faḍl Laylat al-ḳadar, bāb 3), and in the year in which he died, as many as twenty days (ibid., Iʿtikāf, bāb 17). During this period, the mosque was full of booths of palm branches and leaves in which the ʿākifūn lived (ibid., bāb 13; cf. 6, 7). The Prophet only went to his house for some very special reason (ibid., bāb 3). This custom was associated with the ascetism of the monks. The faithful were vexed, when on one occasion he received Ṣafiyya in his booth and chatted for an hour with her (al-Buk̲h̲ārī, Farḍ al-k̲h̲ums, bāb 4; Iʿtikāf, bāb 8, 11, 12). According to another tradition, his iʿtikāf was broken on another occasion by his wives putting up their tents beside him, and he postponed his Iʿtikāf till S̲h̲awwāl (al-Buk̲h̲ārī, Iʿtikāf, bābs 6, 7, 14, 18). According to Zayd b. ʿAlī, the Iʿtikāf can only be observed in a chief mosque (d̲j̲āmiʿ) (Corpus iuris di Zaid b. ʿAlī, no. 447). During the early period, it was one of the initiatory rites for new converts. In the year 14/1635 ʿUmar ordered the retreat (al-ḳiyām) in the mosques during the month of Ramaḍān for the people of Medina and the provinces (al-Ṭabarī, i, 2377). The custom persisted and has always been an important one among ascetics. “The man who retires for a time to the mosque devotes himself in turn to ṣalāt, recitation of the Ḳurʾān, meditation, d̲h̲ikr, etc.” says Ibn al-Ḥād̲j̲d̲j̲ (Madk̲h̲al, ii, 50). There were pious people who spent their whole time in a mosque (aḳāmū fīhi ; al-Maḳrīzī, iv, 87, 97); of one we were told that he spent his time in the manāra of the Mosque of ʿAmr (iʿtakafa, ibid., 44). Al-Samhūdī says that during the month of Ramaḍān, he spent day and night in the mosque (Wüstenfeld, Medina, 95). Saʿd al-Dīn (d. 644/1246-7) spent the month of Ramaḍān in the Mosque of the Umayyads without speaking (Ibn Abī Uṣaybiʿa, ii, 192). Nocturnal vigils in the mosque very early became an established practice in Islam. According to Ḥadīt̲h̲, the Prophet frequently held nocturnal ṣalāts in the mosque with the believers (al-Buk̲h̲ārī, D̲j̲umʿa. bāb 29), and by his orders ʿAbd Allāh b. Unays al-Anṣārī came from the desert for twenty-three successive nights to pass the night in his mosque in rites of worship (Ibn Ḳutayba, Maʿārif ed. Wüstenfeld, 142-3). Out of this developed the tahad̲j̲d̲j̲ud [q.v.] ṣalāt, particularly recommended in the law and notably the tarāwīḥ ṣalāts [q.v.]. In Dihlī on these occasions, women singers actually took part (Ibn Baṭṭūṭa, iii, 155).

During the nights of the month of Ramaḍān, there were festivals in the mosques, and on other occasions also, such as the New Year, sometimes at the new moon, and in the middle of the month. The mosque on these occasions was illuminated: there was eating and drinking; incense was burned and d̲h̲ikr and ḳirāʾa performed.

The Friday ṣalāt was particularly solemn in Ramaḍān, and in the Fāṭimid period, the caliph himself delivered the k̲h̲uṭba (see al-Maḳrīzī, ii, 345 ff.; Ibn Tag̲h̲rībirdī, ii/1, ed. Juynboll, 482-6, ii/2, ed. Popper, 331-3). The mosques associated with a saint had and still have their special festivals on his mawlid [q.v.]; they also are celebrated with d̲h̲ikr, ḳirāʾa, etc. (cf. Lane, Manners and customs, chs. xxiv ff.). The saint’s festivals are usually local and there are generally differences in the local customs. In the Mag̲h̲rib, for example, in certain places the month of Ramaḍān is opened with a blast of trumpets from the manābir (Madk̲h̲al, ii, 69).

The mosque thus on the whole took over the role of the temple. The rulers from ʿUmar onwards dedicated gifts to the Kaʿba (Ibn al-Faḳīh, 20-1, and BGA, iv, Indices, glossarium, s.v. s̲h̲amsa), and, as in other sanctuaries, we find women vowing children to the service of the mosque (al-Buk̲h̲ārī, Ṣalāt, bāb 74; al-Maḳrīzī, iv, 20). Ṭawāf was performed, as at the Kaʿba, in mosques with saints’ tombs as is still done, e.g. in Hebron; Mud̲j̲īr al-Dīn sees a pre-Islamic custom in this (Sauvaire, Hist. Jérus. et Hébron, 5). Especially important business was done here. In times of trouble, the people go to the mosque to pray for help, for example during drought, for which there is a special ṣalāt (which however usually takes place on the muṣallā) [see istisḳāʾ ], in misfortunes of all kinds (e.g. Wüstenfeld, Medina, 19-20; al-Maḳrīzī, iv, 57); in time of plague and pestilence, processions, weeping and praying with Ḳurʾāns uplifted, were held in the mosques or on the muṣallā, in which even Jews and Christians sometimes took part (Ibn Tag̲h̲rībirdī, ii/2, ed. Popper, 67; Ibn Baṭṭūṭa, i, 243-4, cf. Quatremère, Hist. Sult. Maml., ii/1, 35, 40; ii/2, 199) or for a period a sacred book like al-Buk̲h̲ārī’s Ṣaḥīḥ was recited (Quatremère, op. cit. ii/2, 35; al-D̲j̲abartī, Merveilles biographiques, French tr., vi, 13). In the courtyards of the mosques in Jerusalem and Damascus in the time of Ibn Baṭṭūṭa, solemn penance was done on the day of ʿArafa (i, 243-4), an ancient custom which had already been introduced into Egypt in the year 27/647-8 by ʿAbd al-ʿAzīz b. Marwān (ḳuʿūd after the ʿaṣr ; cf. al-Kindī, Wulāt, 50). Certain mosques were visited by barren women (Wüstenfeld, Medina, 133). An oath is particularly binding if it is taken in a mosque (cf. J. Pedersen, Der Eid bei den Semiten, 144); this is particularly true of the Kaʿba, where written covenants were also drawn up to make them more binding (ibid., 143-4, Chron. Mekka, i, 160-1). It is in keeping with this idea of an oath that Jews who had adopted Islam in Cairo had to take oaths in a synagogue which had become a mosque (al-Maḳrīzī, iv, 265). The contract of matrimony (ʿaḳd al-nikāḥ) also is often concluded in a mosque (Santillana, Il Muḫtaṣar, ii, 548; Madk̲h̲al, ii, 72 below; Snouck Hurgronje, Mekka, ii, 163-4), and the particular form of divorce which is completed by the liʿān [q.v.] takes place in the mosque (al-Buk̲h̲ārī, Ṣalāt, bāb 44; cf. Pedersen, Der Eid, 114).

It is disputed whether a corpse may be brought into the mosque and the ṣalāt al-d̲j̲ināza performed there. According to one ḥadīt̲h̲, the bier of Saʿd b. Abī Waḳḳāṣ was taken into the mosque at the request of the Prophet’s widow and the ṣalāt held there. Many disapproved of this, but ʿĀʾis̲h̲a pointed out that the Prophet had done this with the body of Suhayl b. Bayḍāʾ (Muslim, D̲j̲anāʾiz, tr. 34; cf. also Ibn Saʿd, i/1, 14-15). The discussion on this point is not unconnected with the discussions regarding the worship of tombs. In theory, this is permitted by al-S̲h̲āfiʿī, while the others forbid it (see Juynboll, Handbuch, 170; I. Guidi, Il Muḫtaṣar, i, 151). The matter does not seem to be quite clear, for Ḳuṭb al-Dīn says that only Abū Ḥanīfa forbids it, but he himself thought that it might be allowable on the authority of a statement by Abū Yūsuf (Chron. Mekka, hi, 208-10). In any case, it was a very general practice to allow it, as Ḳuṭb al-Dīn also points out. ʿUmar conducted the funeral ṣalāt for Abū Bakr in the Mosque of the Prophet and ʿUmar’s own dead body was brought there; later it became a general custom to perform the ceremony in Medina close to the Prophet’s tomb and in Mecca at the door of the Kaʿba; some even made a sevenfold ṭawāf with the corpse around the Kaʿba. This was for a time forbidden by Marwān b. al-Ḥakam and later by ʿUmar b. ʿAbd al-Azīz (Ḳuṭb al-Dīn, loc. cit., Wüstenfeld, Medina, 77). The custom was very early introduced into the Mosque of ʿAmr (al-Maḳrīzī, iv, 7, 1 ff.). That later scholars often went wrong about the prohibition is not at all remarkable; for it is not at all in keeping with the ever-increasing tendency to found mosques at tombs. Even Ibn al-Ḥād̲j̲d̲j̲, who was anxious to maintain the prohibition, is not quite sure and really only forbids the loud calling of the ḳurrāʾ, d̲h̲ākirūn, mukabbirūn and murīdūn on such occasions (Madk̲h̲al, ii, 50-1, 64, 81). When a son of Sultan al-Muʾayyad died and was buried in the eastern ḳubba, of the Muʾayyad mosque, the k̲h̲aṭīb delivered a k̲h̲uṭba and conducted the ṣalāt thereafter and the ḳurrāʾ recited for a week at the grave, while the amīrs paid their visits to the grave (al-Maḳrīzī, iv, 240, 2 ff.). In Persia, it was the custom for the family of the deceased to sit in the mosque for three days after the death and receive visits of condolence (al-Muḳaddasī, 440 below).

4. Mosques as objects of pilgrimage. As soon as the mosque became a regular sanctuary, it became the object of pious visits. This holds especially true of the memorial mosques associated with the Prophet and other saints. Among them, three soon became special objects of pilgrimage. In a ḥadīt̲h̲ the Prophet says “One should only mount into the saddle to visit three mosques: al-Masd̲j̲id al-Ḥarām, the Mosque of the Prophet and al-Masd̲j̲id al-Aḳṣā” (al-Buk̲h̲ārī, Faḍl al-ṣalāt fī masd̲j̲id Makka wa ’l-Madīna, bāb 16; D̲j̲azāʾ al-ṣayd, bāb 26; Ṣawm, bāb 67; Muslim, Ḥad̲j̲d̲j̲, tr. 93; Chron. Mekka, i, 303). This ḥadīt̲h̲ reflects a practice which only became established at the end of the ʿUmayyad period. The pilgrimage to Mecca had been made a duty by the prescription of the ḥad̲j̲d̲j̲ in the Ḳurʾān. The pilgrimage to Jerusalem was a Christian custom which could very easily be continued, on account of the significance of al-Masd̲j̲id al-Aḳṣā in the Ḳurʾān. This custom became particularly important when ʿAbd al-Malik made it a substitute for the pilgrimage to Mecca (al-Yaʿḳūbī, Taʾrīk̲h̲, ii, 311). Although this competition did not last long, the significance of Jerusalem was thereby greatly increased. Pilgrimage to Medina developed out of the increasing veneration for the Prophet. In the year 140/757-8, Abū D̲j̲aʿfar al-Manṣūr on his ḥad̲j̲d̲j̲ visited the three sanctuaries (al-Ṭabarī, iii, 129) and this became a very usual custom. Mecca and Medina, however, still held the preference. Although those of Mecca and Jerusalem were recognised as the two oldest (the one is said to be 40 years older than the other; Muslim, Masād̲j̲id, tr. 1; Chron. Mekka, i, 301), the Prophet is however reputed to have said “A ṣalāt in this mosque is more meritorious than 1,000 ṣalāts in others, even in al-Masd̲j̲id al-Ḥarām” (al-Buk̲h̲ārī, Faḍl al-ṣalāt jī masd̲j̲id Makka wa ’l-Madīna, bāb 1; Muslim, Ḥad̲j̲d̲j̲, tr. 89; Chron. Mekka, i, 303). The ḥadīt̲h̲ is aimed directly against Jerusalem and therefore probably dates from the Umayyad period. According to some, it was pronounced because someone had commended performing the ṣalāt in Jerusalem, which the Prophet was against (Muslim, loc. cit.; al-Wāḳidī-Wellhausen, 349). The three mosques, however, retained their pride of place (Ibn K̲h̲aldūn, Muḳaddima, faṣls 4, 6; Ibn al-Ḥād̲j̲d̲j̲, Madk̲h̲al, ii, 55), and as late as 662/1264 we find Baybars founding awḳāf for pilgrims who wished to go on foot to Jerusalem (Quatremère, Hist. Suit. Maml., i/1, 248).

Although these three mosques officially hold a special position, others also are highly recommended, e.g. the mosque in Ḳubāʾ [see al-madīna ]. A ṣalāt in this mosque is said to be as valuable as an ʿumra or two visits to the mosque in Jerusalem (al-Diyarbakrī, K̲h̲amīs, i, 381-2). Attempts were also made to raise the mosque of Kūfa to the level of the three. ʿAlī is said to have told someone who wanted to make a pilgrimage from Kūfa to Jerusalem that he should stick by the mosque of his native town, it was “one of the four mosques” and two rakʿas in it were equal to ten in others (Ibn al-Faḳīh, 173-4; Yāḳūt, Muʿd̲j̲am, iv, 325); in another tradition, ṣalāts in the provincial mosques are said to be generally worth as much as the pilgrimage (al-Maḳrīzī, iv, 4), and traditions arose about the special blessings associated at definite times with different holy places of Islam (al-Muḳaddasī, 183) and especially about their superior merits (Ibn al-Faḳīh, 174). The Meccan sanctuary, however, always retained first place, which was marked by the ḥad̲j̲d̲j̲. It was imitated by al-Mutawakkil in Sāmarrāʾ: he built a Kaʿba as well as a Minā and an ʿArafa there and made his amīrs perform their ḥad̲j̲d̲j̲ there (al-Muḳaddasī, 122).

D. The component parts and furnishings of the mosque.

1. The development of the edifice. Except in the case of Mecca the earliest mosques as described above (B. 1) were at first simply open spaces marked off by a ẓulla. The space was sometimes, as in al-Fusṭāṭ, planted with trees and usually covered with pebbles, e.g. in Medina (Muslim, Ḥad̲j̲d̲j̲, tr. 95; al-Balād̲h̲urī, 6) and Fusṭāṭ (al-Maḳrīzī, iv, 8; Ibn Duḳmāḳ, iv, 62; Ibn Tag̲h̲rībirdī, i, 77), which was later introduced in Baṣra and Kūfa, the courtyards of which were otherwise dusty (al-Balād̲h̲urī, 277, 348). These conditions could only last so long as the Arabs retained their ancient customs as a closed group in their simple camps. The utilisation of churches was the first sign of a change and was rapidly followed by a mingling with the rest of the population and the resulting assimilation with older cultures.

ʿUmar made alterations in the mosques in Medina and in Mecca also. He extended the Mosque of the Prophet by taking in the house of ʿAbbās; but like the Prophet, he still built with labin, palm trunks and leaves and extended the booths (al-Buk̲h̲ārī, Ṣalāt, bāb 62; al-Balād̲h̲urī. 6). In Mecca also, his work was confined to extending the area occupied by the mosque. He bought the surrounding houses and took them down and then surrounded the area with a wall to the height of a man; the Kaʿba was thus given its fināʾ like the mosque in Medina (al-Balād̲h̲urī, 46; Chron. Mekka, i, 306; Wüstenfeld, Medina, 68). ʿUt̲h̲mān also extended these two mosques, but introduced an important innovation in using hewn stone and plaster (d̲j̲aṣṣ) for the walls and pillars. For the roof he used teak (sād̲j̲). The booths, which had been extended by ʿUmar, were replaced by him by pillared halls (arwiḳa, sing, riwāḳ) and the walls were covered with plaster (al-Buk̲h̲ārī, Ṣalāt, bāb 62; al-Balād̲h̲urī, 46; Wüstenfeld, Medina, 70). Saʿd b. Abī Waḳḳāṣ is said to have already taken similar steps to relieve the old simplicity of the barely-equipped mosque in Kūfa. The ẓulla consisted of pillars of marble adorned in the style of Byzantine churches (al-Ṭabarī, i, 2489; Yāḳūt, iv, 324).

This was little in keeping with the simple architecture of the original town, for Baṣra and Kūfa had originally been built of reeds and only after several great fires were they built of labin (see above, LB. 1; cf. Ibn Ḳutayba, Maʿārif, ed. Wüstenfeld, 279). As to Kūfa, Saʿd by ʿUmar’s orders extended the mosque so that it became joined up with the Dār al-Imāra. A Persian named Rūzbih b. Buzurd̲j̲mihr was the architect for this. He used fired bricks (ād̲j̲urr) for the building, which he brought from Persian buildings, and in the mosque he used pillars which had been taken from churches in the region of Ḥīra belonging to the Persian kings; these columns were not erected at the sides but only against the ḳibla wall. The original plan of the mosque was therefore still retained, although the pillared hall, which is identical with the ẓulla already mentioned (200 d̲h̲irāʿs broad), replaced the simple booth, and the materials were better in every way (al-Ṭabarī, i, 2491-2, 2494). Already under the early caliphs we can therefore note the beginnings of the adoption of a more advanced architecture.

These tendencies were very much developed under the Umayyads. Even as early as the reign of Muʿāwiya, the mosque of Kūfa was rebuilt by his governor Ziyād. He commissioned a pagan architect, who had worked for Kisrā, to do the work. The latter had pillars brought from al-Ahwāz, bound them together with lead and iron clamps to a height of 30 d̲h̲irāʿs and put a roof on them. Similar halls, built of columns (here like the old booth in Medina called ṣuffa : al-Ṭabarī, i, 2492, 14; but also ẓulla, plur. ẓilāl : al-Ṭabarī, ii, 259-60) were added by him on the north, east and western wall. Each pillar cost him 18,000 dirhams. The mosque could now hold 60,000 instead of 40,000 (idem, i, 2492, 6 ff., cf. 2494, 7; Yāḳūt, iv, 324, 1 ff.; al-Balād̲h̲urī, 276). Al-Ḥad̲j̲d̲j̲ād̲j̲ also added to the mosque (Yāḳūt, iv, 325-6). Ziyād did similar work in Baṣra. Here also he extended the mosque and built it of stone (or brick) and plaster and with pillars from al-Ahwāz, which were roofed with teak. We are told that he made al-ṣuffa al-muḳaddima, i.e. the ḳibla hall, with 5 columns. This seems to show that the other sides also—as in Kūfa—had pillared halls. He erected the Dār al-Imāra close to the ḳibla side. This was taken down by al-Ḥad̲j̲d̲j̲ād̲j̲, rebuilt by others, and finally taken into the mosque by Hārūn al-Ras̲h̲īd (al-Balād̲h̲urī. 347, 348 above, 349; Yāḳūt, i, 642, 643). In Mecca also in the same period similar buildings were erected. Ibn al-Zubayr and al-Ḥad̲j̲d̲j̲ād̲j̲ both extended the mosque, and Ibn al-Zubayr was the first to put a roof on the walls; the columns were gilded by ʿAbd al-Malik and he made a roof of teak (Chron. Mekka, i, 307, 309). The Mosque of ʿAmr was extended in 53/673 with Muʿāwiya’s permission by his governor Maslama b. Muk̲h̲allad to the east and north; the walls were covered with plaster (nūra) and the roofs decorated; it is evident from this that here also the original booth of the south side was altered to a covered hall during the early Umayyad period. A further extension was made in 79/698 in the reign of ʿAbd al-Malik (al-Maḳrīzī, iv, 7, 8; Ibn Duḳmāḳ, iv, 62). Thus we find that during the early Umayyad period, and in part even earlier, the original simple and primitive mosques were in some cases extended, in other cases altered. The alteration consisted in the old simple booth of the Mosque of the Prophet being gradually enlarged and transformed into a pillared hall with the assistance of the arts of countries possessing a higher degree of civilisation. In this way, what had originally been an open place of assembly developed imperceptibly into a court, surrounded by pillared halls. Very soon a fountain was put in the centre of the court, and we now have the usual type of mosque. The same plan is found in the peristyle of the houses and in the aithrion of a basilica like that of Tyre (Herzog-Hauch, Realencyclopädie 3, x, 780).

The great builders of the Umayyads, ʿAbd al-Malik and his son al-Walīd I, made even more radical progress. The former entirely removed the original mosque in Jerusalem, and his Byzantine architects erected the Dome of the Rock as a Byzantine building (cf. Sauvaire, Jérus. et Hébron, 48 ff.). Al-Walīd likewise paid equally little attention to the oldest form of mosque, when, in Damascus, he had the church of St. John transformed by Byzantine architects into the Mosque of the Umayyads. As al-Muḳaddasī distinctly states, they wanted to rival the splendours of the Christian churches (159). The new mosques, which were founded in this period, were therefore not only no longer simple, but they were built with the help of Christians and other trained craftsmen with the use of material already existing in older buildings. Al-Ḥad̲j̲d̲j̲ād̲j̲, for example, used materials from the surrounding towns when building his foundation of Wāsiṭ (al-Ṭabarī, iii, 321; al-Balād̲h̲urī, 290). Columns from churches were now used quite regularly (e.g. in Damascus: al-Masʿūdī, Murūd̲j̲, iii, 408 = § 1292; Ramla: al-Muḳaddasī, 165; cf. al-Balād̲h̲urī, 143 ff.; for Egypt, see al-Maḳrīzī, iv, 36, 124-5). Sometimes, remains of the older style remained alongside the new. In Īrāns̲h̲ahr, al-Muḳaddasī found in the chief mosque wooden columns of the time of Abū Muslim along with round columns of brick of the time of ʿAmr b. al-Layt̲h̲ (316). The building activities of al-Walīd extended to Fusṭāṭ, Mecca and Medina (cf. Ibn al-Faḳīh, 106-7) where no fundamental alterations were made, but complete renovations were carried out. With these rulers, the building of mosques reaches the level of older architecture and gains a place in the history of art. There is also literary evidence for the transfer of a style from one region to another. In Iṣṭak̲h̲r, for example, there was a d̲j̲āmiʿ in the style of the Syrian mosques with round columns, on which was a baḳara (al-Muḳaddasī, iii, 436-7; cf. for S̲h̲īrāz, 430). Al-Walīd also rebuilt the Mosque of the Prophet, in part in the Damascus style (ibid., 80; al-Ḳazwīnī, ed. Wüstenfeld, ii, 71).

This revolution naturally did not take place without opposition, any more than the other innovations, which Islam adopted in the countries with a higher culture which it conquered. After the Mosque of the Prophet had been beautified by Christian architects with marble, mosaics, shells, gold, etc. and al-Walīd in 93/712 was inspecting the work, an old man said: “We used to build in the style of mosques; you build in the style of churches” (Wüstenfeld, Medina, 74). The disccusions on this point are reflected in ḥadīt̲h̲s. When ʿUmar enlarged the Mosque of the Prophet, he is reported to have said: “Give the people shelter from the rain, but take care not to make them red or yellow lest you lead the people astray”, while Ibn ʿAbbās said: “You shall adorn them with gold as the Jews and Christians do” (al-Buk̲h̲ārī, Ṣalāt, bāb 62). Ibn ʿAbbās here takes up the Umayyad attitude and ʿUmar that of old-fashioned people, according to whom any extension or improvement of the ẓulla was only permissible for strictly practical reasons. The conservative point of view is predominant in Ḥadīt̲h̲. It is said that extravagant adornment of the mosques is a sign of the end of the world; the works of al-Walīd were only tolerated from fear of the fitna (Ibn Ḥanbal, Musnad, iii, 134, 145, 152, 230, 283; al-Nasāʾī, Masād̲j̲id, bāb 2; Ibn Mād̲j̲a, Masād̲j̲id, bāb 2). The lack of confidence of pious conservatives in the great mosques finds expression in a ḥadīt̲h̲, according to which the Prophet (according to Anas) said: “A time will come over my umma when they will vie with one another in the beauty of their mosques; then they will visit them but little” (al-ʿAsḳalānī, Fatḥ al-Bārī, i, 362). In fiḳh, we even find divergence from the oldest quadrangular form of the mosque condemned (Guidi, Il Muḫtaṣar, i, 71). Among the types which arose later was the “suspended” (muʿallaḳ), i.e. a mosque situated in an upper storey (e.g. in Damascus, JA, ser. 9, vol. v, 409, 415, 422, 424, 427, 430).

2. Details of the component parts and equipment of the mosque. — a. The Minaret [see on this, manāra ]. — b. The Chambers. The old mosque consisted of the courtyard and the open halls running along the walls: these were called al-mug̲h̲aṭṭā (al-Muḳaddasī, 82, 158, 165, 182) because they were roofed over. When we are told that in Palestine, except in Jericho, towers were placed between the mug̲h̲aṭṭā and the courtyard (ibid., 182), this seems to suggest that the halls were closed, which would be quite in keeping with the winter climate of this region. The halls were particularly extensive on the ḳibla side, because assemblies were held here. The space between two rows of pillars was called riwāḳ, pl. arwiḳa or riwāḳāt (ibid., 158, 159; al-Maḳrīzī, iv, 10, 11, 12, 49). Extension often took the form of increasing the number of the arwiḳa. In some districts, a sail-cloth was spread over the open space as a protection from the sun at the time of the worship (al-Muḳaddasī, 205, 430).

The courtyard was called ṣaḥn. The open space around the Kaʿba is called Fināʾ al-Kaʿba (Chron. Mekka, i, 307; Ibn His̲h̲ām, 822; cf. Fināʾ Zamzam : Yāḳūt, Udabāʾ, vi, 376). Fināʾ is also the name given to the open space around the mosque (al-Maḳrīzī, iv, 6). Trees were often planted in the courtyard: e.g. in the mosque of ʿAmr (see above, I. B. 1; when we read in al-Maḳrīzī, iv, 6, that it had no ṣaḥn, this probably means that this space, planted with trees, between the covered halls was very narrow). In Medina, at the present day, there are still trees in the Rawḍa (al-Batanūnī, Riḥla, 249); in Ibn D̲j̲ubayr’s time there were 15 palms there (Riḥla, 194). Other mosques in Cairo had trees growing in them (al-Maḳrīzī, iv, 54, 64, 65, 120; in al-Masd̲j̲id al-Kāfūrī, there were as many as 516 trees: ibid., 266), as is still the case today. In other cases the court was covered with pebbles (see above, I. D. 1); but this was altered with a more refined style of architecture. Al-Muḳaddasī mentions that this was only found in Tiberias, out of all the mosques in Palestine (182). Frequently, as in Ramla, the halls were covered with marble and the courtyard with flat stone (ibid., 165). In the halls also, the ground was originally bare or covered with little stones; for example in the mosques of ʿAmr until Maslama b. Muk̲h̲allad covered it with mats (see below). The floor of the Mosque of ʿAmr was entirely covered with marble in the Mamlūk period (al-Maḳrīzī, iv, 13-14, cf. in S̲h̲īrāz, Ibn Baṭṭūṭa, ii, 53). But in the mosque of Mecca, the ṣaḥn is still covered with little stones (al-Batanūnī, Riḥla, 99 below); 400 dīnārs used to be spent annually on this (Chron. Mekka, ii, 10-11). In Medina also, little pebbles were used (Ibn D̲j̲ubayr, Riḥla, 190; Ibn Baṭṭūṭa, i, 263).

There were not at first enclosed chambers in the halls. A change in this respect came with the introduction of the maḳṣūra (on this word, cf. Quatremère, Hist. Sult. Maml., i/1, 164, n. 46). This was a box or compartment for the ruler built near the miḥrāb. Al-Samhūdī gives the history of the maḳṣūra in Medina (Wüstenfeld, Medīna, 71-2, 89). The traditions all agree that the maḳṣūra was introduced to protect the ruler from hostile attacks. According to some authorities, ʿUt̲h̲mān built a maḳṣūra of labin with windows, so that the people could see the imām of the community (ibid., and al-Maḳrīzī, iv, 7). According to another tradition, Marwān b. al-Ḥakam, governor of Medina after an attempt had been made on him by a Yamanī in the year 44/664, was the first to build a maḳṣūra of dressed stone with a window (al-Balād̲h̲urī. 6 below; al-Ṭabarī, ii, 70). Muʿāwiya is then said to have followed his example. Others, again, say that Muʿāwiya was the first to introduce this innovation. He is said to have introduced the maḳṣūrāt with the accompanying guard as early as the year 40/660-1 or not till 44/664-5 after the K̲h̲ārid̲j̲ī attempt (al-Ṭabarī, i, 3465, 9; Ibn al-Faḳīh, 109, 3; al-Maḳrīzī, iv, 12, 11 ff.); according to one story because had had seen a dog on the minbar (al-Bayhaḳī, ed. Schwally, 393 below; cf. on the whole question, H. Lammens, Moʿâwiya, 202 ff.). This much seems to be certain, that the maḳṣūra was at any rate introduced at the beginning of the Umayyad period, and it was an arrangement so much in keeping with the increasing dignity of the ruler that, as Ibn K̲h̲aldūn says, it spread throughout all the lands of Islam (Muḳaddima, Cairo 1322/1904-5, 212-13, faṣl 37). The governors built themselves compartments in the principal mosques of the provinces, e.g. Ziyād in Kūfa and Baṣra (al-Balād̲h̲urī, 277, 348) and probably Ḳurra b. S̲h̲arīk in Fusṭāṭ (al-Maḳrīzī, iv, 12). In Medina, we are told that ʿUmar b. ʿAbd al-ʿAzīz as governor (86-93/705-12) raised the maḳṣūra and built it of teak, but al-Mahdī had it taken down in 160/777 and a new one built on the level of the ground (ibid., 7; Wüstenfeld, op. cit.; al-Balād̲h̲urī, 7 centre). We are further told that in 161/778, al-Mahdī prohibited the maḳāṣīr of the provinces, and al-Maʾmūn even wanted to clear all the boxes out of the masād̲j̲id d̲j̲āmiʿa, because their use was a sunna introduced by Muʿāwiya (al-Maḳrīzī, iv, 12; al-Yaʿḳūbī, Taʾrīk̲h̲, ii, 571). But this attempt did not succeed. On the contrary, their numbers rapidly increased. In Cairo, for example, the D̲j̲āmiʿ al-ʿAskar built in 169/785-6 had a maḳṣūra (al-Maḳrīzī, iv, 33 ff.) and the mosque of Ibn Ṭūlūn had a maḳṣūra beside the miḥrāb which was accessible from the Dār al-Imāra (ibid., 36, 37, 42; Ibn Tag̲h̲rībirdī, ii, 8, 14). The maḳṣūra was found in the larger mosques. In the D̲j̲āmiʿ al-Ḳalʿa, Muḥammad b. Ḳalāwūn in 718/1318 built a maḳṣūra of iron for the sultan’s ṣalāt (al-Maḳrīzī, iv, 132). According to Ibn K̲h̲aldūn, the maḳṣūra was an innovation peculiar to the Islamic world. The question must however be left open, whether in its introduction and development there may not be some connection with the boxes of the Byzantine court, at least, for example, when the Turks in the Yes̲h̲il D̲j̲āmiʿ in Bursa put the sultan’s box over the door (R. Hartmann, Im neuen Anatolien, 27).

Although the maḳṣūra was introduced with the object of segregating the ruler and was therefore condemned by the strict as contrary to the spirit of Islam (e.g. Madk̲h̲al, ii, 43-4), maḳāṣīr were probably introduced for other purposes. Ibn D̲j̲ubayr mentions three in the Mosque of the Umayyads: the old one built by Muʿāwiya in the eastern part of the mosque, one in the centre, which contained the minbar, and one in the west where the Ḥanafīs taught and performed the ṣalāt. There were also other small rooms shut off by wooden lattices, which could be sometimes called maḳṣūra and sometimes zāwiya. As a rule, there were quite a number of zāwiyas connected with the mosque which were used by students (Riḥla, 265-6). We find the same state of affairs in other mosques.

While the groups of the ḳurrāʾ, the students, the lawyers, etc. had originally to sit together in a common room, gradually the attempt was made to introduce separate rooms for some of them. Small compartments were either cut off in the main chamber or new rooms were built in subsidiary buildings. In the former case, we get the already mentioned maḳāṣīr or zawāyā. Ibn al-Ḥād̲j̲d̲j̲ says that a madrasa was often made by the simple process of cutting off a part of the mosque by a balustrade (darbazīn) (Madk̲h̲al, ii, 44). Thus in the halls of the Mosque of ʿAmr there were several compartments for teaching, which were called maḳṣūra and zāwiya, in which studies were prosecuted (al-Maḳrīzī, iv, 20, 16, 25). In the Azhar Mosque, a maḳṣūrat Fāṭima was made in the time of the Fāṭimids, where she had appeared, and the amīrs in the following period made a large number of such maḳāṣīr (ibid., 52, 53). In the Aḳṣā Mosque about 300/912-13, there were three maḳṣūras for women (Ibn al-Faḳīh, 100). These divisions might be a nuisance at the great Friday assemblies, and this is why al-Mahdī wanted to remove them in 161/778 from the masād̲j̲id al-d̲j̲amāʿāt (al-Ṭabarī, iii, 486), and Ibn al-Ḥād̲j̲d̲j̲ condemned them as works of the mulk and numbers them like other embellishments with the as̲h̲rāṭ al-sāʿa (Madk̲h̲al, ii, 43-4).

The muʾad̲h̲d̲h̲ins not only lived in the minarets, where, at any rate in the Ṭūlūnid period, they held vigils (al-Maḳrīzī, iv, 48). They had rooms (g̲h̲uraf, sing, g̲h̲urfa) on the roof and these rooms in time came to be numerous (ibid., 13, 14). All kinds of rooms were put in subsidiary buildings, for the k̲h̲aṭīb (ibid., 13), for judges, for studies, etc. In addition, there were dwelling-houses, not only for the staff but also for others. As already mentioned, devout men used to take up their residence in the mosque for a considerable period for iʿtikāf and any one at any time could take up his quarters in the mosque; he could sleep there and make himself at home. It therefore came quite natural to the devout to reside permanently in the mosque. Ascetics often lived in the minaret (see above), a zāhid lived on the roof of the Azhar mosque, others made themselves cells in the mosque, as a s̲h̲ayk̲h̲ in Naṣībīn did (Ibn D̲j̲ubayr, Riḥla, 240; cf. in Ḥarrān, 245) and as happened in Ṣalāḥ al-Dīn’s time in the Mosque of the Umayyads (Ibn Abī Uṣaybiʿa, ii, 182). It was, however, very usual for them to live in the side rooms of the mosque, as was the case for example, in the Mosque of the Umayyads (Ibn D̲j̲ubayr, 269; Ibn Baṭṭūṭa, i, 206). In particularly holy mosques like that in Hebron, houses for al-muʿtakifūn were built around the sacred place (Sauvaire, Hist. Jérus. et Hébron, 11-12) and also beside the Masd̲j̲id Yūnis at the ancient Niniveh (al-Muḳaddasī, 146). Kitchens were therefore erected with the necessary mills and ovens and cooked food (d̲j̲as̲h̲īs̲h̲a) and 14-15,000 loaves (rag̲h̲īf) were daily distributed to those who stayed there and to visitors (Sauvaire, 20; cf. Quatremère, Hist. Sult. Maml., i/1, 231). Bread was also baked in the mosque of Ibn Ṭulūn (Quatremère, op. cit., i/1, 233) and kitchens were often found in the mosques (for al-Azhar, see al-D̲j̲abartī, Merveilles, iii, 238-9; Sulaymān Raṣad, Kanz al-d̲j̲awhar fī taʾrīk̲h̲ al-Azhar, 71 ff., 107 ff.). Those who lived in and beside the mosque were called mud̲j̲āwirūn (cf. al-Muḳaddasī, 146; for Jerusalem, Nāṣir-i K̲h̲usraw, 82, 91; for Mecca, Ibn D̲j̲ubayr. 149; for Medina, Ibn Baṭṭūṭa, i, 279, where we learn that they were organised under a ḳadīm, like the North Africans under an amīn in Damascus; Ibn D̲j̲ubayr, 277-8). They were pious ascetics, students and sometimes travellers. The students generally found accommodation in the madāris, but large mosques like that of the Umayyads or al-Azhar had always many students, who lived in them. The name of the halls, riwāḳ, was later used for these students’ lodgings (cf. van Berchem, CIA, i, 43, n. 1; perhaps al-Maḳrīzī, iv, 54, 23). Strangers always found accommodation in the mosques (cf. above, I. C. 1). In smaller towns, it was the natural thing for the traveller to spend the night in the mosque and to get food there (Yāḳūt, iii, 385; al-Ḳifṭī, Taʾīk̲h̲ al-Ḥukamāʾ ed. Lippert, 252). Travellers like Nāṣir-i K̲h̲usraw. Ibn D̲j̲ubayr, Ibn Baṭṭūṭa and al-ʿAbdarī (JA, ser. 5, iv [1854], 174) were able to travel throughout the whole Muslim world from one mosque (or madrasa or ribāṭ) to the other. The traveller could even leave his money for safe keeping in a mosque (Safar-nāma, 51). Large endowments were bequeathed for those who lived in the mosques (Ibn D̲j̲ubayr, op. cit.; Ibn Tag̲h̲rībirdī, ii/2, 105 f.).

In later times, the rulers often built a lodge or pavilion (manẓara) in or near the mosque (al-Maḳrīzī, ii, 345; iv, 13; cf. on the word, Quatremère, Hist. Sult. Maml, ii/2, 15).

There was often a special room with a clock in the mosques; this also is probably an inheritance from the church, for Ibn Rusta talks of similar arrangements in Constantinople (126 above). Ibn D̲j̲ubayr (270) describes very fully the clock in the Mosque of the Umayyads (cf. JA, ser. 9, vii, 205-6). It was made in the reign of Nūr al-Dīn by Fak̲h̲r al-Dīn b. al-Sāʿātī (Ibn Abī Uṣaybiʿa, ii, 183-4; an expert was kept to look after it, ibid., 191). There was a clock in the Mustanṣiriyya in Bag̲h̲dād (Sarre and Herzfeld, Arch. Reise, ii, 170), and the Mosque of ʿAmr also a g̲h̲urfat al-sāʿāt (al-Maḳrīzī, iv, 13, 15). In the Mosque of Ibn Ṭūlūn is still kept a sundial of the year 696/1296-7; cf. van Berchem, CIA, i, no. 415), but the clocks were usually mechanical (see also Dozy, Supplément, s.v. mind̲j̲āna, and on the clock generally, E. Wiedemann, in Nova Acta der K. Leop. Carol. Akad., c [Halle 1915]). In the Mag̲h̲rib also we find mosqueclocks, e.g. in the Bū ʿInāniyya (JA, ser. 11, xii, 357 ff.).

The very varied uses to which the mosques were put resulted in their becoming storehouses for all sorts of things. In 668/1269-70, the Mosque of the Umayyads was cleared of all such things; in the courtyard there were, for example, stores for machines of war, and the zāwiya of Zayn al-ʿĀbidīn was a regular k̲h̲ān (JA, ser. 9, vii, 225-6).

c. The prayer-niche or Miḥrāb [see for this, miḥrāb ].

d. The pulpit or Minbar [see for this, minbar ].

e. The platform or Dakka. In the larger mosques, there is usually found near the minbar a platform to which a staircase leads up. This platform (dakka, popularly often dikka) is used as a seat for the muʾad̲h̲d̲h̲ins when pronouncing the call to prayer in the mosque at the Friday service. This part of the equipment of a mosque is connected with the development of the service (cf. below, under I. H. 4, and C. H. Becker, Zur Geschichte des islamischen Kultus, in Isl., iii [1912], 374-99 = Islamstudien, i, 472-500; E. Mittwoch, Zur Entstehungsgeschichte des islamischen Gebets und Kultus, in Abh. Pr. Ak. W. [1913], Phil.-Hist. Cl., no. 2). The first ad̲h̲ān call is pronounced from the minaret, the second (when the k̲h̲aṭīb mounts the minbar) and the third (before the ṣalāt, iḳāma) in the mosque itself. These calls were at first pronounced by the muʾad̲h̲d̲h̲in standing in the mosque. At a later date, raised seats were made for him.

Al-Ḥalabī records that Maslama, Muʿāwiya’s governor in Egypt, was the first to build platforms (here called manābir) for the calls to prayer in the mosques (Sīra Ḥalabiyya, ii, 111 below). This story, however, given without any reference to older authorities, is not at all reliable. It seems that a uniform practice did not come into existence at once. In Mecca, the muʾad̲h̲d̲h̲ins for a time uttered the second call (when the preacher mounted the minbar) from the roof. As the sun in summer was too strong for them, the amīr of Mecca, in the reign of Hārūn al-Ras̲h̲īd, made a little hut (ẓulla) for them on the roof. This was enlarged and more strongly built by al-Mutawakkil in 240/854-5, as his contemporary al-Azraḳī relates (Chron. Mekka, i, 332-3). The position in the mosque of ʿAmr in Cairo was similar. Here also the ad̲h̲ān was uttered in a chamber (g̲h̲urfa) on the roof, and in 336/947-8 there is a reference to its enlargement (al-Maḳrīzī, iv, 11). As late as the time of Baybars, when the many chambers were removed from the roof of the Mosque of ʿAmr, the old g̲h̲urfa of the muʾad̲h̲d̲h̲in was left intact (ibid., 14; cf. al-Kindī, Wulāt, ed. Guest, 469, n. 2). In the Mosque of Ibn Ṭūlūn, the ad̲h̲ān was pronounced from the cupola in the centre of the ṣaḥn (al-Maḳrīzī, iv, 40). Al-Muḳaddasī records in the 4th/10th century as a notable thing about K̲h̲urāsān that the muʾad̲h̲d̲h̲ins there pronounced the ad̲h̲ān on a sarīr placed in front of the minbar (327). The dukkān “platform” in front of the minbar in the mosques of Shahrastān must have had the same purpose (ibid., 357).

In the 8th/14th century, Ibn al-Ḥād̲j̲d̲j̲ mentions the dakka as a bidʿa in general use, which should be condemned as it unnecessarily prevents freedom of movement within the mosque (Madk̲h̲al, ii, 45 above). In the year 827/1424 a dakka in the mosque of al-Ḥākim is mentioned (al-Maḳrīzī, iv, 61); the dakkas mentioned in inscriptions from Cairo all date from the period before and after 900/1495. Ibn al-Ḥād̲j̲d̲j̲ mentions that, in addition to the large dakka used for the Friday worship, there was sometimes a lower one for ordinary ṣalāts (Madk̲h̲al, ii, 46-7) and says that in the larger mosques there were several dakkas on which muʾad̲h̲d̲h̲ins pronounced the ad̲h̲ān in succession so that the whole community could hear it (tablīg̲h̲ ; ibid., 45-6). Lane also mentions several muballig̲h̲s in the Azhar Mosque (Manners and customs, Everyman’s Library edn., 87, 2).

f. The reading-stand or Kursī Ḳurʾāns and relics. In the mosques there is usually a kursī [q.v.], that is, a wooden stand with a seat and a desk. The desk is for the Ḳurʾān, the seat for the ḳāṣṣ, or reader, ḳāriʾ. Ibn D̲j̲ubayr attended the worship in Bag̲h̲dād at which a celebrated preacher spoke from the minbar, but only after the ḳurrāʾ, sitting on karāsī had recited portions of the Ḳurʾān (Riḥla, 219, 222). The wāʿiẓ, often identical with the ḳāṣṣ, sat on a kursī made of teak (Ibn D̲j̲ubayr. 200. Yāḳūt, Udabāʾ, ii, 319; al-Maḳrīzī, iv, 121); sometimes he spoke from the minbar to which the wāʿiẓ often had access (cf. Ibn D̲j̲ubayr; see Mez, Renaissance des Islâms, 320, Eng. tr. 332). The ḳuṣṣāṣ are called by al-Makkī aṣḥāb al-karāsī, which is in keeping with this (Ḳūt al-ḳulāb, i, 152, quoting K. al-Madk̲h̲al, i, 159). Several karāsī are often mentioned in one mosque (cf. for the Mosque of ʿAmr, al-Maḳrīzī, iv, 19). Whether the karāsī mentioned for the earlier period always had a desk cannot be definitely ascertained. The karāsī with dated inscriptions given by van Berchem in his Corpus all belong to the 9th/15th century (nos. 264, 302, 338, 359bis, 491). According to Lane, at the Friday service, while the people are assembling, a ḳāriʾ on the kursī recites sūra XVIII up to the ad̲h̲ān (Manners and customs, 86). The same custom is recorded by Ibn al-Ḥād̲j̲d̲j̲ and condemned because it has a disturbing effect (Madk̲h̲al, ii, 44, middle).

The Ḳurʾān very soon received its definite place in the mosque, like the Bible in the church (cf. al-Buk̲h̲ārī. Ṣalāt, bāb 91: they prayed at a pillar beside al-muṣḥaf). According to one tradition, ʿUt̲h̲mān had several copies of his Ḳurʾān sent to the provinces (e.g. Nöldeke-Schwally, Gesch. d. Qor., ii, 112-13); al-Ḥad̲j̲d̲j̲ād̲j̲, a little later, is said to have done the same thing (al-Maḳrīzī, iv, 17). The mosques had many other copies beside the one kept on the kursī. Al-Ḥākim put 814 maṣāḥif in the Mosque of Ibn Ṭūlūn, where the founder had already put boxes of Ḳurʾāns (al-Maḳrīzī, iv, 36, 40; cf. al-Suyūṭī, Ḥusn al-muḥāḍara, ii, 138) and in 403/1012-13, he presented 1,289 copies to the Mosque of ʿAmr, some of which were written in letters of gold (al-Maḳrīzī, iv, 12; al-Suyūṭī, ii, 136). Even earlier than this there were so many that the ḳāḍī al-Ḥārit̲h̲ b. Miskīn (237-45/851-9) appointed a special amīn to look after them (al-Kindī, Wulāt, 469); there are still a very large number in the Mosque of the Prophet (see al-Batanūnī, Riḥla, 241 above). Of particular value was the muṣḥaf Asmāʾ, belonging to the Mosque of ʿAmr, prepared by ʿAbd al-ʿAzīz b. Marwān, later bought by his son and afterwards by his daughter Asmāʾ; her brother left it in 128/746 to the mosque and it was used for public readings (see its whole history in al-Maḳrīzī, iv, 17-18). Besides it, another copy was for some time also used for reading, which was said to have lain beside ʿUt̲h̲mān, when he was killed and to have been stained with his blood, but this one was removed by the Fāṭimids (ibid., 19). In the time of Ibn Baṭṭūṭa, a Ḳurʾān for which the same claims were made was kept in Baṣra (ii, 10). On New Year’s Day, when the Fāṭimid caliphs used to go in procession through the town, the caliph at the entrance to the Mosque of ʿAmr took up in his hands a muṣḥaf said to have been written by ʿAlī and kissed it (Ibn Tag̲h̲rībirdī, ii/1, 472 middle); it was perhaps the muṣḥaf Asmāʾ. In Syria, Egypt, and the Ḥid̲j̲āz, in the 4th/10th century, there were Ḳurʾāns which were traced back to ʿUt̲h̲mān (al-Muḳaddasī, 143; cf. Ibn Ḥawḳal 1, 117). One of the Ḳurʾāns made for ʿUt̲h̲mān was shown in the Mosque of the Umayyads in Damascus in the time of Ibn D̲j̲ubayr. It was produced after the daily ṣalāts and the people touched and kissed it (Riḥla, 268). It was brought there in the year 507/1113-14 from Tiberias (al-D̲h̲ahabī, Taʾrīk̲h̲, Ḥaydarābād, 1337, ii, 25). Other Ḳurʾāns of ʿUt̲h̲mān were shown in Bag̲h̲dād and Cordova (see Mez, Renaissance des Islâms, 327, Eng. tr., 338-9) and Ibn D̲j̲ubayr saw another in the Mosque of the Prophet; it lay in a desk on a large stand, here called miḥmal (Riḥla, 193; cf. thereon Dozy, Supplément, s.v.). The Fāḍiliyya madrasa also had a muṣḥaf ʿUt̲h̲mān, bought by the Ḳāḍī al-Fāḍil for 30,000 dīnārs (al-Maḳrīzī, iv, 197) and there is one in Fās (Archives Marocaines, xviii [1922], 361). Valuable Ḳurʾāns like these had the character of relics and belonged to the k̲h̲izāna of the mosque. They were often kept in a chest (ṣandūḳ) (Ibn D̲j̲ubayr, op. cit.; for al-muṣḥaf, al-Buk̲h̲ārī, Ṣalāt, bāb 95, Muslim has al-ṣandūḳ see al-ʿAṣkalānī, Fatḥ al-Bārī, i, 385), also called tābūt (Ibn D̲j̲ubayr, 104). In the Kaʿba, Ibn D̲j̲ubayr saw two chests with Ḳurʾāns (84, 3). Ibn al-Faḳīh mentions 16 chests with Ḳurʾāns in the Jerusalem mosque (100). In the mosques there were also ṣanādīḳ for other things, such as lamps (al-Maḳrīzī, iv, 53; Wüstenfeld, Medina, 82 = Ibn D̲j̲ubayr. 194), a tābūt for alms Madk̲h̲al, ii, 44, below), for the bayt al-māl or the property of the mosque (see below). There were also chests for rose-wreaths (Madk̲h̲al, ii, 50) which were in charge of a special officer. In the Mosque of ʿAmr there was a whole series of tawābīt (al-Maḳrīzī, iv, 9).

The Ḳurʾāns were not the only relics to be kept in the mosques. Bodies or parts of the bodies of saints (cf. above, B. 4, C. 1) and other āt̲h̲ār were kept and revered in mosques: the rod of Moses (in Kūfa, Yāḳūt, iv, 325, previously in Mecca, see Goldziher, Muh. Stud., ii, 361), the Prophet’s sandals (in Hebron, Ibn al-Faḳīh, 101, also in Damascus, where the Madrasa As̲h̲rafiyya had his left and the Dammāg̲h̲iyya his right sandal; JA, ser. 9, iii, 271-2, 402), his cloak (in Ad̲h̲ruḥ, al-Muḳaddasī, 178), hair from his beard (in Jerusalem among other places, al-Batanūnī, Riḥla, 165) and many other things (see Goldziher, Muh. Stud., ii, 358 ff.; Mez, Renaissance des Islâms, 325-6, Eng. tr., 337-9). These relics were often kept in valuable reliquaries. The head of Ḥusayn was buried in a tābūt in his mosque in Cairo (Ibn D̲j̲ubayr, 45). There was a black stone like that in the Kaʿba in a mosque in S̲h̲ahrastān (al-Muḳaddasī, 433).

On the other hand, pictures and images were excluded from the mosques, in deliberate contrast to the crucifixes and images of saints in churches, as is evident from Ḥadīt̲h̲ (al-Buk̲h̲ārī. Ṣalāt, bābs 48, 54; D̲j̲anāʾiz. bāb 71; Muslim, Masād̲j̲id, tr. 3; cf. on the question, Becker, Christliche Polemik und islamische Dogmenbildung, in ZA, xxvi [1911] = Islamstudien, i, 445 ff.). It is of interest to note that in the earliest period, Saʿd b. Abī Waḳḳāṣ had no scruples about leaving the wall-paintings in the Īwān of Kisrā at Madāʾin standing, when it was turned into a mosque (al-Ṭabarī, i, 2443, 2451). The case was somewhat different, when, before the chief mosque in Dihlī, which had been a Hindu temple, two old copper idols formed a kind of threshold (Ibn Baṭṭūṭa, iii, 151), although even this is remarkable (cf. Snouck Hurgronje, Verspreide Geschriften, ii, 451 ff. = ZDMG, lxi [1907], 186 ff.). In some circles the opposition to pictures extended to other relics also. Ibn Taymiyya condemned the reverence paid to the Prophet’s footprint, which was shown, as in Jerusalem, in a Damascus mosque also (Quatremère, Hist. Sult. Maml, ii/2, 246).

g. Carpets. Carpets [see on these, bisāṭ in Suppl.] were used to improve the appearance of the mosques. The custom of performing the ṣalāt upon a carpet is ascribed by Ḥadīt̲h̲ to the Prophet himself. Anas b. Malik performed the ṣalāt with him in his grandmother’s house and the Prophet used a cloth or mat (ḥaṣīr), which had become black through wear; as a rule, he used a mat woven of palm leaves, k̲h̲umra (al-Buk̲h̲ārī, Ṣalāt, bābs 19, 20, 21; Ḥayḍ, bāb 30; Muslim, Masād̲j̲id, tr. 47; Aḥmad b. Ḥanbal, Musnad, iii, 145). In any case, it is clear from al-Balād̲h̲urī that the ṣalāt was at first performed in the mosque simply in the dust and then on pebbles (al-Balād̲h̲urī, 277, 348; cf. al-Zurḳānī, S̲h̲arḥ ʿalā ’l-Muwaṭṭāʾ, i, 283-4). Later, when the halls were extended, the ground, or the paving, was covered with matting.

The first to cover the ground in the Mosque of ʿAmr with ḥuṣur instead of ḥaṣbāʾ was Muʿāwiya’s governor Maslama b. Muk̲h̲allad (al-Maḳrīzī, iv, 8; al-Suyūṭī, ii, 136; Ibn Tag̲h̲rībirdī. i, 77). The different groups which frequented the mosque (cf. above) had their places on particular mats: when a ḳāḍī (middle of the 3rd/9th century) ejected the S̲h̲āfiʿīs and Ḥanafīs from the mosque, he had their ḥuṣur torn up (al-Kindī, Wulāt, 469). Ibn Ṭūlūn covered his mosque floor with ʿAbbādānī and Sāmānī mats (al-Maḳrīzī, iv, 36, 38). For the mosque of al-Ḥākim in the year 403/1012-13, al-Ḥākim bought 1,036 d̲h̲irāʿs of carpeting for 5,000 dīnārs (al-Maḳrīzī, iv, 56; cf. for al-Azhar, ibid., 50). In the year 439/1047-8 in the Mosque of ʿAmr, there were ten layers of coloured carpets one above the other (Nāṣir-i K̲h̲usraw, ed. Schefer, text, 31, tr. 149). In the Mosque at Jerusalem, 800,000 d̲h̲irāʿs of carpets were used every year (Ibn al-Faḳīh, 100). In the Mosque in Mecca they were renewed every Ramaḍān (loc. cit.). On ceremonial occasions, the minbar was also draped with a carpet (sad̲j̲d̲j̲āda); in Medina, the minbar and the sacred tomb was always covered like the Kaʿba in Mecca (Wüstenfeld, Medina, 83; cf. Quatremère, Hist. Sult. Maml, ii/1, 91) and some, especially the teachers, had their skins (farwa), in some cases, also a cushion to lean upon. The doors were also covered with some material (al-Maḳrīzī, iv, 56). On feast-days, the mosques were adorned with carpets in a particularly luxurious fashion (see Ibn Tag̲h̲rībirdī. ii/1, 483). The puritanical rejected all this as bidʿa and preferred the bare ground (Madk̲h̲al, ii, 46, 49, 72, 74, 76), as the Wahhābīs still do.

h. Lighting. Where evening meetings and vigils were of regular occurrence, artificial lighting became necessary. Al-Azraḳī gives the history of the lighting of the Meccan Mosque. The first to illuminate the Kaʿba was ʿUḳba b. al-Azraḳ, whose house was next to the Mosque, just on the maḳām ; here he placed a large lamp (miṣbāḥ). ʿUmar, however, is said previously to have placed lamps upon the wall, which was the height of a man, with which he surrounded the mosque (al-Balād̲h̲urī, 46). The first to use oil and lamps (ḳanādīl) in the mosque itself was Muʿāwiya (cf. Ibn al-Faḳīh, 20). In the time of ʿAbd al-Malik, K̲h̲ālid b. ʿAbd Allāh al-Ḳasrī placed a lamp on a pillar of the Zamzam beside the Black Stone, and the lamp of the Azraḳ family disappeared. In the reign of al-Maʾmūn in 216/831, a new lamp-post was put up on the other side of the Kaʿba, and a little later two new lanterns were put up around the Kaʿba. Hārūn al-Ras̲h̲īd placed ten large lamps around the Kaʿba and hung two lanterns on each of the walls of the mosque (t̲h̲urayyāt ; cf. Ibn D̲j̲ubayr, Riḥla, 149, 150, 155, 271; van Berchem, CIA, i, no. 506). K̲h̲ālid al-Ḳasrī had the masʿā also illuminated during the pilgrimage, and in 119/737 the torches called nafāṭāt were placed here, and ʿUmar b. ʿAbd al-ʿAzīz ordered the people, who lived in the streets of Mecca, to put up lamps on 1 Muḥarram for the convenience of those visiting the Kaʿba (Chron. Mekka, i, 200-2, cf. 458-9). In 253/867 Muḥammad b. Aḥmad al-Manṣūrī erected a wooden pole in the centre of the ṣaḥn and ḳanādīl on ropes were hung from it. This was, however, very soon removed (ibid., ii, 196-7). About 100 years later, al-Muḳaddasī saw around the ṭawāf wooden poles on which hung lanterns (ḳanādīl), in which were placed candles for the kings of Egypt, Yemen, etc. (74). Ibn D̲j̲ubayr describes the glass ḳanādīl, which hung from hooks in the Meccan Ḥaram (Riḥla, 103) and lamps (mas̲h̲āʿil) which were lit in iron vessels (ibid., 103, cf. 143). Similar silver and gold ḳanādīl were seen by him in Medina (ibid., 192 at the top; see also Wüstenfeld, Medina, 83 ff.). According to Ibn al-Faḳīh (before 300/912), 1,600 lamps were lit every evening in Jerusalem (100), and in the next century al-Muḳaddasī says that the people of Palestine always burn ḳanādīl in their mosques, which were hung from chains as in Mecca (182). The illumination was thus very greatly increased. In the year 60/679-80, when ʿUbayd Allāh b. Ziyād was searching for his enemies in the mosque of Kūfa, the lamps were not sufficient, and large torches had to be used in searching the pillared halls (al-Ṭabarī, ii, 259-60). This, like what has already been said about Mecca, shows out of what modest beginnings this part of the mosque’s equipment developed.

In the time of the ʿAbbāsids, lamps and lanterns were part of the regular furniture of the mosque. Al-Maʾmūn is said to have taken a special interest in this. He ordered lamps to be put in all the mosques, partly to assist those who wanted to read and partly to prevent crime (al-Bayhaḳī, 473). For this purpose, the ḳanādīl, already mentioned, hung on chains were used, as at the building of the mosque of Ibn Ṭūlūn (al-Maḳrīzī, iv, 36, 38), in the Azhar Mosque and elsewhere; they were often of silver (ibid., 56, 63). Golden ḳanādīl were also used and were of course condemned by Ibn al-Ḥād̲j̲d̲j̲ (Madk̲h̲al, ii, 54) as ostentatious. At the same time, candles (s̲h̲amʿ or s̲h̲amaʿ) were used in large numbers, the candle-sticks (atwār, sing, tawr) often being of silver (Ibn D̲j̲ubayr, Riḥla, 45, 151, 194; cf. Wüstenfeld, Medina, 95, 100). About 400/1009-10, large candelabra were made in Egypt, which from their shapes were called tannūr, stoves. Al-Ḥākim presented the Mosque of ʿAmr with a tannūr made out of 100,000 dirhams of silver; the mosque doors had to be widened to admit it. He also gave it two other lamps (al-Suyūṭī, ii, 136 below; cf. Nāṣir-i K̲h̲usraw, text 51, tr. 148; Ibn Tag̲h̲rībirdī, ed. Popper, ii/2, 105). In the Mosque of al-Ḥākim, in addition to lamps and candle lanterns, he also put 4 silver tanānīr and he made similar gifts to the Azhar and other mosques: the lamps were of gold or silver (al-Maḳrīzī, iv, 51, 56, 63; cf. Ibn Tag̲h̲rībirdī, ii/2, 105). The tanānīr and other lanterns could also be made of copper (see van Berchem, CIA, i, nos. 502, 503, 506, 507, 511), as, for example, the celebrated candelabrum of the Mosque of Muʾayyad (al-Maḳrīzī, iv, 137) which was made for the mosque of Ḥasan but sold by it (ibid., 118).

This great interest in the lighting of the mosque was not entirely based on practical considerations. Light had a significance in the worship and Islam here, as elsewhere, was taking over something from the Christian Church. When, in 227/842 the caliph was on his deathbed, he asked that the ṣalāt should be performed over him with candles and incense (bi ’l-s̲h̲amʿ wa ’l-buk̲h̲ūr) exactly after the fashion of the Christians (Ibn Abī Uṣaybiʿa, i, 165; cf. ii, 89). The dependence of Islam on Christianity is also seen in the story that ʿUt̲h̲mān, when he was going to the evening ṣalāt in Medina, had a candle carried in front of him, which his enemies condemned as bidʿa (al-Yaʿḳūbī, Taʾrīk̲h̲, ii, 187). The S̲h̲īʿī bias does not affect the significance of this story. A light was used particularly in the miḥrāb, because it represented the holy cell, to which light belongs (cf. sūra XXIV, 35). Then, in Mecca, lamps were placed before the imāms in the miḥrābs and there were considerable endowments for such miḥrāb lamps (Ibn D̲j̲ubayr, Riḥla, 103, 144). Light, as was everywhere the custom in ancient times, was necessary in mausoleums, and the documents of endowment show that a large number of oil-lamps were used in this way (cf. e.g. the document for al-Malik al-As̲h̲raf’s mausoleum, van Berchem, CIA, i, no. 252). But in the mosque generally the use of lights had a devotional significance and lamps might be endowed for particular individuals (cf. al-Muḳaddasī, 74, quoted above). The lamps so given by al-Ḥākim were therefore placed in the mosques with great ceremony, with blasts of trumpets and beating of drums (Ibn Tag̲h̲rībirdī. ii/2, 105).

On ceremonial occasions a great illumination was therefore absolutely necessary. In the month of Ramaḍān, says Ibn D̲j̲ubayr, the carpets were renewed and the candles and lamps increased in number, so that the whole mosque was a blaze of light (Riḥla, 143); on certain evenings, trees of light were made with vast numbers of lamps and candles and the minarets were illuminated (ibid., 149-51, 154, 155). In the Mosque of the Prophet in the time of al-Samhūdī, forty wax candles burned around the sacred tomb, and three to four hundred lights in the whole mosque (Wüstenfeld, Medina, 100). On the mawlid al-nabī, says Ḳuṭb al-Dīn, a procession went from the Kaʿba in Mecca to the birthplace of the Prophet with candles, lanterns (fawānīs) and lamps (mas̲h̲āʿil) (see Chron. Mekka, iii, 439). In the ḥaram of Jerusalem, according to Mud̲j̲īr al-Dīn, 750 lamps were lit by night and over 20,000 at festivals (Sauvaire, Hist. Jérus. et Hébron, 138). In the dome of the Ḳubbat al-Ṣak̲h̲ra in 452/1060, a chandelier and 500 lamps fell down (ibid., 69); at the taking of the town in 492/1099, the Franks carried off 42 silver lamps, each of 3,600 dirhams, 23 lamps of gold and a tannūr of 40 raṭl s of silver (ibid., 71). It was similar, and still is, in Cairo and elsewhere in the Muslim world. For the laylat al-wuḳūd in the Mosque of ʿAmr, 18,000 candles were made for the Mosque of ʿAmr, and every night eleven-and-a-half ḳinṭārs of good oil were used (al-Maḳrīzī, iv, 21 and more fully, ii, 345-6). The four “nights of illumination” fell in the months of Rad̲j̲ab and S̲h̲aʿbān, especially niṣf S̲h̲aʿbān (Quatremère, Hist. Sult. Maml., ii/2, 131; cf. also Snouck-Hurgronje, Mekka, ii, 77). In 1908 electric light was introduced into the Mosque of the Prophet (al-Batanūnī, Riḥla, 245-6).

(On the question in general of illumination, see Clermont-Ganneau, La lampe et l’olivier dans le Coran, in Recueil d’Archéologie Orientale, viii [1924], 183-228; on the copper candelabra, see A. Wingham, Report on the analysis of various examples of oriental metal-work, etc. in the South Kensington Museum, etc., London 1892; F. R. Martin, Ältere Kupferarbeiten aus dem Orient, Stockholm 1902; on glass lamps, see G. Schmoranz, Altorientalische Glass-Gefässe, Vienna 1898; van Berchem, CIA, i, 678 ff.; M. Herz Bey, La Mosquée du Sultan Hasan (Comité de Conservation des Monuments de l’Art Arabe), 1899, 8 ff.; see also the Bibliography in Isl., xvii [1928], 217 ff.).

i. Incense. According to some traditions, even the Prophet had incense burned in the mosque (al-Tirmid̲h̲ī, i, 116; see Lammens, Moʿâwia, 367, n. 8) and in the time of ʿUmar, his client ʿAbd Allāh is said to have perfumed the mosque by burning incense while he sat on the minbar. The same client is said to have carried the censer (mid̲j̲mar : cf. Lammens, loc. cit.) brought by ʿUmar from Syria before ʿUmar when he went to the ṣalāt in the month of Ramaḍān (A. Fischer, Biographie von Gewährsmännern, etc., 55 n.). According to this tradition, the use of incense was adopted into Islam very early as a palpable imitation of the custom of the Church. In keeping with this is the tradition that, in Fusṭāṭ as early as the governorship of ʿAmr, the muʾad̲h̲d̲h̲in used to burn incense in the mosque (Ibn ʿAbd al-Ḥakam, 132; cf. Annali dell’ Islām, iv, 565). The Ḳubba, t al-Ṣak̲h̲ra Mosque had incense burned in it during the consecration ceremony (Sauvaire, Hist. Jérus. et Hébron, 53).

Under the Umayyads, incense was one of the regular requirements of the mosque (ṭīb al-masd̲j̲id : al-Ṭabarī, ii, 1234, 10). Muʿāwiya is named as the first to perfume the Kaʿba with perfume (k̲h̲alūḳ) and censer (ṭayyaba : Ibn al-Faḳīh, 20, 12). It became the custom to anoint the sacred tombs with musk and ṭīb (Chron. Mekka, i, 150, 10; Ibn D̲j̲ubayr, Riḥla, 191, 9). Baybars washed the Kaʿba with rose-water (al-Maḳrīzī, iv, 96, 14). Incense, as well as candles, was used at burials (cf. de Goeje, ZDMG, lix [1905], 403-4; Lammens, Moʿâwia, 436, n. 9). Al-Muʿtaṣim’s desire to be buried with candles and incense (buk̲h̲ūr) exactly like the Christians (Ibn Uṣaybiʿa, i, 165, 12 f., cf. above) shows that they were aware that the custom bore much the same relation to the Christian usage as the mosque building did to the church. The consumption of incense in the mosques gradually became very large, especially at festivals (see for the Fāṭimids, Ibn Tag̲h̲rībirdī, ii/1, 484, 12; ii/2, ed. Popper, 106, 3; al-Maḳrīzī, iv, 51; on vessels for holding incense, see the Bibliography in Isl, xvii [1928], 217-18, and maʿdin. 4. In Islamic art).

j. Water-supply. Nothing is said of a water-supply in connection with the oldest mosques. The Mosque of Mecca occupied a special position on account of the Zamzam well. In the early days of Islam, two basins (ḥawḍ) are said to have been supplied by it, one behind the well, i.e. just at the side of the mosque for wuḍūʾ and one between the well and the rukn for drinking purposes; the latter was moved nearer the well by Ibn al-Zubayr. In the time of Sulaymān b. ʿAbd al-Malik, a grandson of ʿAbd Allāh b. ʿAbbās for the first time built a ḳubba, in connection with Zamzam (Chron. Mekka, i, 299). At the same time, the governor K̲h̲ālid al-Ḳasrī laid down lead piping to bring water from the well of al-T̲h̲abīr to the mosque, to a marble basin (fisḳiyya) with a running fountain (fawwāra) between Zamzam and the rukn, probably on the site of the earlier ḥawḍ. It was intended to supply drinking-water in place of the brackish water of Zamzam, but a branch was led on to a birka at the Bāb al-Ṣafā, which was used for ritual ablutions. The people, however, would not give up the Zamzam water and immediately after the coming to power of the ʿAbbāsids, the provision for drinking-water was cut off, only the pipe leading to the birka being retained (ibid., i, 339-40). In Ibn D̲j̲ubayr’s time, there was, in addition to Zamzam, a supply of water in vessels and a bench for performing the wuḍūʾ (Riḥla, 89). K̲h̲alid’s plan, arrangements for ablutions at the entrance and a running fountain in the ṣaḥn, seems to have been a typically Umayyad one and to have been introduced from the north. Such fountains were usual in the north, not only in private houses, but also for example in the aithrion (atrium) surrounded by pillars, which, from Eusebius’s description, formed part of the church of Tyre (see Hauch, in Herzog-Hauch, Realenzyclop. f. prot. Theol. u. Kirche 3, x, 782).

The usual name for the basin, fisḳiyya (in Egypt now fasḳiyya), comes from piscina, which in the Mis̲h̲na and in Syriac takes the form piskīn (see Levy, Neuhebr. u. chald. Wörterbuch, iv, 81b; Fraenkel, Fremdwörter, 124; fisḳīna, found in al-Azraḳī, Chron. Mekka, i, 340 is probably due to a slip). At the same time, however, birka or siḳāya or ṣihrīd̲j̲, which probably comes from the Persian (cf. Fraenkel, op. cit., 287), or the old Arabic ḥawḍ, are also used. The arrangements for ablutions were called maṭāhir or mayāḍiʾ, sing. miʾḍaʾa (now usually mēḍā), “place for wuḍū ”. The accommodation in Mecca just mentioned was later extended. Ibn D̲j̲ubayr mentions a building at al-Ẓāhir, 1 mīl north of Mecca which contained maṭāhir and siḳāya for those performing the minor ʿumra (Riḥla, 111).

In Medina, Ibn D̲j̲ubayr mentions rooms for wuḍūʾ at the western entrance to the mosque (Riḥla, 197, 13 f.; cf. the plan in al-Batanūnī, Riḥla, facing p. 244). At the same time, Ibn Zabāla mentions seventeen receptacles for water in the ṣaḥn in the year 199/814-15, probably for drinking-water; later (8th/14th century) a large basin surrounded by a railing is mentioned in the centre of the court. It was intended for drinking purposes, but became used for bathing and was therefore removed. Baths and latrines were built anew by al-Nāṣir’s mother (Wüstenfeld, Medina, 99 ff.).

In Damascus, where every house, as is still the case, was amply supplied with water, Yāḳūt (d. 626/1229) found no mosque, madrasa or k̲h̲ānaḳāh which did not have water flowing into a birka in the ṣaḥn (Yāḳūt, ii, 590). Ibn D̲j̲ubayr describes the arrangements in the Mosque of the Umayyads. In the ṣaḥn, as is still the case, there were three ḳubbas. The centre one rested on four marble columns, and below it was a basin with a spring of drinking-water surrounded by an iron grille. This was called ḳafaṣ al-māʾ “water-cage”. North of the ṣaḥn was a Masd̲j̲id al-Kallāsa, in the ṣaḥn of which there was again a ṣihrīd̲j̲ of marble with a spring (Ibn D̲j̲ubayr, Riḥla, 267). There was also running water in an adjoining mas̲h̲had (269), in the k̲h̲ānaḳāh and madrasa (271), and in a hall beside the living apartments there was again a ḳubba, with a basin (ḥawḍ) and spring water (269). There were also siḳāyāt against the four outer walls of the mosque, whole houses fitted up with lavatories and closets (273); a century earlier, we are told that at each entrance to the mosque there was a miʾḍaʾa (159). The whole arrangements correspond exactly to those made by K̲h̲ālid al-Ḳasrī in Mecca in the Umayyad period and must therefore date from the Umayyads.

It was the same in other Syrian and Mesopotamian towns. In Sāmarrāʾ, al-Mutawakkil built in his new d̲j̲āmiʿ a fawwāra with constant running water (al-Yaʿkūbī, Buldān, 265). In Naṣībīn, the river was led through the ṣaḥn of the mosque into a ṣihrīd̲j̲, there was also a ṣihrīd̲j̲ at the eastern entrance with two siḳāyāt in front of the mosque (Ibn D̲j̲ubayr, Riḥla, 239). In Mawṣil in the mosque, which dated from the Umayyad period, there was a spring with a marble cupola over it (ibid., 235). In Ḥarrān, there were in the ṣaḥn three marble ḳubbas with a biʾr and drinking-water (ibid., 246), in Aleppo, two (ibid., 253). In Kūfa, there were three ḥawḍs with Euphrates water in front of the D̲j̲āmiʿ (ibid., 212), but in the mosque in a zāwiya, a domed building with running water (Yāḳūt, iv, 325, 326, here called tannūr ; cf. Ibn al-Faḳīh, 173, Ibn D̲j̲ubayr, 89, 267). It was the same in Amid (Nāṣir-i K̲h̲usraw, ed. Schefer, 28) and in Zarand̲j̲ in Sid̲j̲istān (Ibn Ḥawḳal 1, 298-9). The principal mosques of ʿIrāḳ had mayāḍiʾ at the entrances, for which, according to a remarkable note by al-Muḳaddasī, rents were paid (129, read karāsī ?; cf. maṣṭaba : Ibn D̲j̲ubayr. 89). In Palestine also, in al-Muḳaddasī’s time, there were conveniences for ablutions at the entrances to the d̲j̲awāmiʿ (maṭāhir : 182; mayāḍiʾ: al-Iṣṭak̲h̲rī 1, and in Ṣanʿāʾ in the 4th/10th century, beside each mosque, there was water for drinking and for wuḍūʾ (Ibn Rusta, 111). In Persia also, it was the custom to have a ḥawḍ in front of the mosque (al-Muḳaddasī, 318) and there was drinking-water in the mosque itself on a bench (kursī) in iron jars into which ice was put on Fridays (ibid., 327). Not only at the Zamzam well but also in the mosques of ʿIrāḳ, men were appointed whose duty it was to distribute drinking-water (al-Ṭabarī, iii, 2165). The regular custom, therefore, was to have at the entrance to, or in front of the mosque, conveniences for wuḍūʾ, and in the court of the mosque itself a fountain as the traditional ornament and for drinking water. It was the exception for the wuḍūʾ to take place in the mosque itself.

In Egypt, at first the Mosque of Ibn Ṭūlūn was arranged similarly to the Syrian mosques. In the centre of the ṣaḥn there was a gilt dome, supported by sixteen marble columns and surrounded by a railing. This upper storey was supported by nineteen marble columns and below was a marble basin (ḳaṣʿa) with a running fountain (fawwāra); the ad̲h̲ān was called from the dome (al-Maḳrīzī, iv, 37; the description is not quite clear). People complained that there were no arrangements for washing (miʾḍaʾa) there. Ibn Ṭūlūn replied that he had not made them because he had concluded the mosque would be polluted thereby. He therefore made a miʾḍaʾa with an apothecary’s shop behind the mosque (ibid., 38, 39; al-Suyūṭī, ii, 139; Ibn Tag̲h̲rībirdī, ii/10). This suggests that previously in Egypt, the washing arrangements had been directly connected with the mosque. After the fire of the year 376/986-7, the fawwāra was renovated by al-ʿAzīz (al-Maḳrīzī, iv, 40), and again in 696/1297 by Lād̲j̲īn, whose inscription still exists (CIA, i, no. 16). A new miʾḍaʾa was built in 792/1390 beside the old one on the north, outside the mosque (al-Maḳrīzī, iv, 42).

The Mosque of ʿAmr first got a fawwāra in the time of al-ʿAzīz. In 378-9/998-9 his vizier Yaʿḳūb b. Killis installed one in the cupola, already in existence for the bayt al-māl. Marble jars were put there for the water (probably drinking-water) (al-Maḳrīzī, iv, 9, 11; cf. al-Suyūṭī, ii, 136; Yāḳūt, iii, 899). A new water basin was installed by Ṣalāḥ al-Dīn beside his manẓara in the mosque. The water was led to the fawwārat al-fisḳiyya from the Nile. This was prohibited in the reign of Baybars al-Bunduḳdārī (658-76/1260-77) by the chief ḳāḍī, because the building was being affected by it (al-Maḳrīzī, iv, 14; al-Suyūṭī, ii, 137). The amīr, who restored it, brought the water for the fisḳiyya from a well in the street (al-Maḳrīzī, iv, 15).

Like Ibn Ṭūlūn, the Fāṭimids do not seem to have considered the miʾḍaʾa indispensable. For the Azhar Mosque had originally no miʾḍaʾa: as late as al-Ḥākim’s waḳf document for the provision of miʾḍaʾa, money is given only with the provision that something of the kind should be made (al-Maḳrīzī, iv, 51, 54). At a later date we hear of two miʾḍaʾa’s, one at the adjoining Āḳbug̲h̲āwiyya (ibid., 54). On the other hand, there was already a fisḳiyya in the centre of the court, but whether it had existed from the first is not known. It had disappeared, when traces of it were found in 827/1424 in laying-out a new ṣihrīd̲j̲ (ibid., 54). The fisḳiyya of the Mosque of al-Ḥākim was not erected by the founder. Like that of the Mosque of ʿAmr, it was removed in 660/1262 by the ḳāḍī Tād̲j̲ al-Dīn, but after the earthquake of 702/1302-3, it was again rebuilt and provided with drinking-water from the Nile (ibid., 56, 57) and again renovated after 780/1378 (ibid., 61). A small miʾḍaʾa, later replaced by another, was in the vicinity of the entrance (ibid., 61). Other Fāṭimid mosques had basins in the ṣaḥn, which were supplied from the Nile and from the K̲h̲alīd̲j̲ (ibid, 76, 81, 120).

The traditional plan was retained in the period following also. For example, we know that the amīr Ṭug̲h̲ān in 815/1412 placed a birka in the centre of the D̲j̲amiʿ of Āḳsunḳur which was covered by a roof supported by marble pillars and supplied by the same pipe as the already existing miʾḍaʾas (al-Maḳrīzī, iv, 107, cf. 124, 138, 139, etc.). At the ceremonial dedication of mosques, it was the custom for the patron to fill the birka in the ṣaḥn with sugar, lemonade or other sweet things (e.g. at the Muʾayyadī, in 822/1419, al-Maḳrīzī, iv, 139; at the Madrasa of D̲j̲amāl al-Dīn in 811/1408-9, ibid, 253; another in 757/1356, ibid, 256).

The importance of the birka of the mosque, as a drinking-place, diminished as pious founders erected drinking fountains everywhere (cf. for Mecca, Chron. Mekka, ii, 116-18; also BGA, Glossarium, 211, s.v. ḥubb ; 258, s.v. sabīl) and especially when it became the custom to build a sabīl with a boy’s school in part of the mosque (see below, I. E. 4, end). A ḥawḍ for watering animals was also sometimes built in the vicinity of the mosque (al-Maḳrīzī, iv, 76). Sometimes also the birka of the ṣaḥn was used for washing. In the year 799/1397 the amīr Yalbug̲h̲ā made arrangements for this in the Aḳmar mosque so that one could get water for wuḍūʾ from taps from a birka put up in the ṣaḥn (al-Maḳrīzī, iv, 76). Al-Maḳrīzī condemns this addition, but only because there was already a miʾḍaʾa at the entrance and the ṣaḥn was too small for the new one (ibid.), and not on grounds of principle; and it was only because the wall was damaged that the amīr’s gift was removed in 815/1412 (ibid., 77). The custom of using the water supply of the ṣaḥn for wuḍūʾ survived in many places in Egypt. The arrangements were therefore usually called miʾḍaʾa or rather mēḍā (which is not found in the inscriptions). If they had taps, they were called ḥanafiyya ; according to Lane’s suggestion, because the Ḥanafīs only permitted ablutions with running water or from a cistern 10 ells broad and deep (Lexicon, s.v.; cf. Manners and customs. Everyman’s Library, 69; cf. on the question M. Herz, Observations critiques sur les bassins dans les ṣaḥn s des mosquées, in BIE, iii/7 [1896], 47-51; idem, La mosquée du Sultan Ḥasan, 2; Herz wrongly dates the modern usage from the Turkish conquest in 923/1517). In quite recent times, the miʾḍaʾas have often been moved outside to special buildings. Ibn al-Ḥād̲j̲d̲j̲ condemns bringing water into the mosque, because the only object is for ablutions and ablutions in the mosque are forbidden by “our learned men” (Madk̲h̲al, ii, 47-8, 49); like shaving, ablutions should be performed outside the mosque in keeping with the Prophet’s saying id̲j̲ʿalū maṭāhirakum ʿalā abwāb i masād̲j̲idikum (ibid., ii, 58). It was in keeping with this principle that in earlier times the miʾḍaʾa was usually put at the entrance and the barbers took up their places before the entrance (cf. the name Bāb al-Muzayyinīn “The Barbers’ Gate” for the main entrance to the Azhar mosque). Miʾḍaʾas were also to be found in hospitals; thus the “lower hospital” was given two in 246/957, one of which was for washing corpses (Ibn Duḳmāḳ, 99 below).

E. The mosque as a state institution.

1. The mosque as a political centre. Its relation to the ruler. It was inherent in the character of Islam that religion and politics could not be separated. The same individual was ruler and chief administrator in the two fields, and the same building, the mosque, was the centre of gravity for both politics and religion. This relationship found expression in the fact that the mosque was placed in the centre of the camp, while the ruler’s abode was built immediately adjacent to it, as in Medina (and in Fusṭāṭ, Damascus, Baṣra, Kūfa). We can trace how this dār al-imāra or ḳaṣr (so for Kūfa: al-Ṭabarī, ii, 230-1; ḳaṣr al-imāra, ibid., 234) with the growth of the mosque gradually became incorporated in it at Fusṭāṭ and Damascus and was replaced by a new building. The tradition remained so strong that, in Cairo, when the new chief mosque D̲j̲āmiʿ al-ʿAskar was being planned in 169/785-6, a Dār Umarāʾ Miṣr was built beside it with direct access to the mosque (al-Maḳrīzī, iv, 33-4), and when Ibn Ṭūlūn built his mosque, a building called the Dār al-Imāra was erected on its south side, where the ruler, who now lived in another new palace, had rooms for changing his robes, etc., from which he could go straight into the maḳṣūra (ibid., 42).

The ʿAbbāsids at the foundation of Bag̲h̲dād introduced a characteristic innovation, when they made the palace the centre of the city; the case was similar with Fāṭimid Cairo; but Sulaymān b. ʿAbd al-Malik in Ramla had already built the palace in front of the mosque (al-Balād̲h̲urī, 143). Later rulers, who no longer lived just beside the mosque, had special balconies or something similar built for themselves in or beside the mosque. Ṣalāḥ al-Dīn built for himself a manẓara under the great minaret of the mosque of ʿAmr (al-Maḳrīzī, iv, 13; al-Suyūṭī, ii, 137) and just to the south of the Azhar mosque, the Fāṭimids had a manẓara from which they could overlook the mosque (al-Maḳrīzī, ii, 345).

The caliph was the appointed leader of the ṣalāt and the k̲h̲aṭīb of the Muslim community. The significance of the mosque for the state is therefore embodied in the minbar. The installation of the caliph consisted in his seating himself upon this, the seat of the Prophet in his sovereign capacity. When homage was first paid to Abū Bakr by those who had decided the choice of the Prophet’s successor, he sat on the minbar. ʿUmar delivered an address, the people paid homage to him and he delivered a k̲h̲uṭba, by which he assumed the leadership (Ibn His̲h̲ām, 1017; al-Ṭabarī, i, 1828-9; al-Diyārbakrī, ii, 75; al-Yaʿḳūbī, Taʾrīk̲h̲, ii, 142); it was the same with ʿUmar and ʿUt̲h̲mān (ibid., 157, 187).

The k̲h̲uṭba, after the glorification of God and the Prophet, contained a reference to the caliph’s predecessor and a kind of formal introduction of himself by the new caliph. It was the same in the period of the Umayyads and ʿAbbāsids (see for al-Walīd, al-Ṭabarī, ii, 1177 ff.; al-Amīn, ibid., iii, 764; al-Mahdī, ibid., iii, 398, 451, 457; cf. on this question also al-Buk̲h̲ārī, Aḥkām, bāb 43). The minbar and the k̲h̲uṭba associated with it was still more important than the imāmate at the ṣalāt, it was minbar al-mulk (Ḥamāsa, ed. Freytag, 656, v, 4). According to a ḥadīt̲h̲, the Prophet carried the little Ḥasan up to the minbar and said, “This my son is a chieftain”, etc. (al-Buk̲h̲ārī, Manāḳib, bāb 25). This reflects the later custom by which the ruler saw that homage was paid to his successor-designate; this also was done from the minbar (cf. k̲h̲uṭiba yawm al-d̲j̲umʿa li “l-Muʿtaḍid bi-wilāyat al-ʿahd, al-Ṭabarī, iii, 2131). The Fāṭimid caliph showed honour to a distinguished officer by allowing him to sit beside him on the minbar (al-Suyūṭī, ii, 91); in the same way, Muʿāwiya allowed Ibn ʿAbbās to sit beside him ʿalā sarīrihi (Ibn Abī Uṣaybiʿa, i, 119), but whether the reference is to the minbar is perhaps doubtful. The bayʿa could also be received by another on behalf of the caliph, but it had to be accepted on the minbar. Thus the governor of Mecca in 196/811-12 accepted on the minbar homage to ʿAbd Allāh al-Maʾmūn and the deposition of Muḥammad al-Amīn (al-Ṭabarī, iii, 861-2; cf. for al-Mahdī: ibid., 389). There are other cases in which the solemn deposition of a ruler took place on or beside the minbar (Ag̲h̲ānī 2, i, 12; Wüstenfeld, Medina, 15). Even at a much later date, when spontaneous acclamation by the populace was no longer of any importance, the ceremonial installation on the minbar was still of importance (al-Maḳrīzī, iv, 94). It had become only a formality but still an important one. Homage was paid to the ʿAbbāsid caliphs in Egypt in the great īwān of the palace or in a tent in which a minbar had been put up, and similarly to the sultans whose investiture was read out from the minbar (cf. Quatremère, Hist. Sult. Maml, i/1, 117, 149 ff., 183 ff.). If one dreamt that he was sitting on the minbar, it meant that he would become sultan (ibid., ii/2, 103). The ʿAbbāsid caliph had, however, long had his own throne after the old Persian fashion in his palace (Ps. al-D̲j̲āḥiẓ, al-Tād̲j̲ fī ak̲h̲lāḳ al-mulūk, ed. Aḥmad Zakī, Cairo 1914, 7 ff.; tr. Pellat, Le livre de la couronne, Paris 1954, 35 ff.) and so had the Fāṭimids (Ibn Tag̲h̲rībirdī, ii/1, 457) and the Mamlūks (Quatremère, op. cit., i/1, 87; cf. 147). When later we find mention of the kursī ’l-k̲h̲ilāfa (van Berchem, CIA, i, no. 33), sarīr al-mulk (Chron. Mekka, iii, 113), sarīr al-salṭana (al-Maḳrīzī, ii, 157; cf. al-sarīr, royal throne: Ibn Ḥawḳal 1, 282, 285; kursī similarly: cf. Ibn ʿArabs̲h̲āh, Vita Timuri, ed. Manger, ii, 186) or martabat al-mulk (Quatremère, op. cit., i/2, 61), the reference is no longer to the minbar. This does not mean that the ruler could no longer make public appearances in the mosques: thus in 648/1250, al-Muʿizz Aybak regularly gave audiences in al-madāris al-ṣāliḥiyya (Quatremère, Hist. Sult. Maml., 17) and memorial services for Baybars were held a year after his death in several mosques, madāris and k̲h̲awāniḳ in Cairo (677/1278; ibid., i/2, 164-5).

The caliph spoke chiefly from the minbar of the capital, but when he made the pilgrimage he also spoke from the manābir in Mecca and Medina (cf. e.g. al-Ṭabarī, ii, 1234; al-Yaʿḳūbī, ii, 341, 501; Chron. Mekka, i, 160). Otherwise, in the provinces, the governor stood in the same relation to the mosques as the caliph in the capital. He was appointed “over ṣalāt and sword” or he administered “justice among the people” and the ṣalāt (al-Ṭabarī, iii, 860), he had “province and minbar” under him (ibid., ii, 611), al-wilāyāt wa ’l-k̲h̲uṭba (al-Muḳaddasī, 337). Speaking from the minbar was a right which the caliph had delegated to him and it was done in the name of the caliph. ʿAmr b. al-ʿAṣ therefore refused to allow people in the country to hold d̲j̲umaʿ except under the direction of the commander (al-Maḳrīzī, iv, 7). This point of view was never quite abandoned. The k̲h̲uṭba was delivered “in the name of the caliph (ibid., 94) or “for” him (lahu, ibid., 66, 74, 198; Ibn Tag̲h̲rībirdī, ii/1, 85 below; al-Muḳaddasī, 485 above), and in the same way an amīr delivered a k̲h̲uṭba “for” a sultan (al-Maḳrīzī, iv, 213, 214). The sultan did not have the “secular” and the caliph the “spiritual” power, but the sultan exercised as a Muslim ruler the actual power which the caliph possessed as the legitimate sovereign and had formally entrusted to him. During the struggles between the different pretenders, there was thus a confession of one’s politics if one performed the ṣalāt with the one or the other governor (al-Ṭabarī, ii, 228, 234, 258; Chron. Mekka, ii, 168). The pretenders disputed as to whether the one or the other could put up his standard beside the minbar (al-Ṭabarī, iii, 2009).

Like the caliph, the governor also made his formal entry into office by ascending the minbar and delivering a k̲h̲uṭba; this was the symbol of his authority (e.g. al-Ṭabarī, ii, 91, 238, 242; Chron. Mekka, ii, 173; cf. Ḥamāsa, 660, vv. 2-3; al-D̲j̲āḥiẓ, Bayān, iii, 135). After glorifying God and the Prophet, he announced his appointment or read the letter from the caliph and the remainder of his address, if there was a war going on, was exclusively political and often consisted of crude threats. The k̲h̲uṭba was not inseparably connected with the Friday service. The commander-in-chief could at any time issue a summons to the ṣalāt and deliver his k̲h̲uṭba with admonitions and orders (see al-Ṭabarī, ii, as above and 260, 297-8, 300, 863, 1179) and it was the same when he left a province (ibid., 241); a governor, who could not preserve his authority with the k̲h̲uṭba was dismissed (ibid., 592). Since war was inseparably associated with early Islam, and since the mosque was the public meeting-place of ruler and people, it often became the scene of warlike incidents. While the governor in his k̲h̲uṭba was issuing orders and admonitions relating to the fighting, cheers and counter-cheers could be uttered (ibid., 238) and councils of war were held in the mosque (al-Ṭabarī, i, 3415; ii, 284; al-Balād̲h̲urī, 267). Soon after his election ʿAbd al-Malik asked from the minbar who would take the field against Ibn al-Zubayr, and al-Ḥad̲j̲d̲j̲ād̲j̲ shouted that he was ready to go (Chron. Mekka, ii, 20). After the Battle of the Camel, ʿAlī sent the booty to the mosque of Baṣra and ʿĀʾis̲h̲a looked for another mosque (al-Ṭabarī, i, 3178, 3223). Rowdy scenes occasionally took place in the mosques (al-Kindī, Wulāt, 18); Ziyād was stoned on the minbar (al-Ṭabarī, ii, 88); one could ride right into the mosque and shout to the governor sitting on the minbar (ibid., 682); fighting often took place in and beside the mosque (ibid., 960, 1701 ff; Wüstenfeld, Medina, 13-14). Sometimes for this reason, the governor was surrounded by his bodyguard during the ṣalāt or the minbar or even clothed in full armour (al-Walīd: al-Ṭabarī, ii, 1234; al-Yaʿḳūbī, ii, 341; al-Ḥad̲j̲d̲j̲ād̲j̲: al-Ṭabarī, ii, 254). ṣalāt and sword were thus closely associated in reality.

It thus came to be the custom for the enemies of the ruler and his party to be cursed in the mosques. This custom continued the old Arab custom of regular campaigns of objurgation between two tribes, but can also be paralleled by the Byzantine ecclesiastical anathematisation of heretics (cf. Becker, Islamstudien, i, 485, Zur Gesch. d. islamischen Kultus).

The first to introduce the official cursing of the ʿAlids from the minbar of the Kaʿba is said to have been K̲h̲ālid al-Ḳasrī (Chron. Mekka, ii, 36). The reciprocal cursing of Umayyads and ʿAlids became general (cf. al-Ṭabarī, ii, 12, 4; Ag̲h̲ānī 2, x, 102; Ibn Tag̲h̲rībirdī, i, 248; see also Lammens, Moʿâwia, 180-1). Like the blessing upon the ruler, it was uttered by the ḳuṣṣāṣ (al-Maḳrīzī, iv, 16); it was even recorded in inscriptions in the mosque (Ibn Tag̲h̲rībirdī, ed. Popper, ii/2, 63, 64; cf. also Mez, Renaissance, 61, Eng. tr., 64). As late as 284/897, al-Muʿtaḍid wanted to restore the anathematisation of Muʿāwiya from the minbar but abandoned the idea (al-Ṭabarī, iii, 2164). Anathemas were also pronounced on other occasions, for example, Sulaymān had al-Ḥad̲j̲d̲j̲ād̲j̲ cursed (Chron. Mekka, ii, 37), and al-Muʿtamid had Ibn Ṭūlūn solemnly cursed from the manābir (al-Ṭabarī, iii, 2048, 5 ff.); and other rulers had Muʿtazilī heretics cursed from the pulpits (see Mez, op. cit.,198, Eng. tr. 206; cf. against Ibn Taymiyya, Quatremère, Hist. Sult. Maml, ii/2, 256). Ibn Baṭṭūṭa describes the tumultuous scene with thousands of armed men uttering threats in a mosque in Bag̲h̲dād when a S̲h̲īʿī k̲h̲aṭīb was on the minbar (ii, 58).

In was very natural to mention with a blessing upon him the ruler in whose name the Friday k̲h̲uṭba was delivered. Ibn ʿAbbās, when governor of Baṣra, is said to have been the first to pronounce such a duʿāʾ over ʿAlī (Ibn K̲h̲aldūn, Muḳaddima, faṣl 37, end); it is not improbable that the custom arose out of the reciprocal objurgations of ʿAlids and Umayyads; the ḳuṣṣāṣ, who had to curse the ʿAlids in the mosques, used to pray for the Umayyads (al-Maḳrīzī, iv, 17). Under the ʿAbbāsids, the custom became the usual form of expressing loyalty to the ruler (Ibn Tag̲h̲rībirdī. ii/1, 151). After the caliph, the name of the local ruler or governor was mentioned (ibid., 156, 161): even in Bag̲h̲dād in 369/979-80 by order of the caliph al-Ṭā-ʾiʿ the actual ruler ʿAḍud al-Dawla was mentioned in the duʿāʾ (Ibn Miskawayh, vi, 499; ed. Cairo 1915, 396) and the Būyids, according to al-Muḳaddasī, were generally mentioned in the k̲h̲uṭba even in the remotest parts of the kingdom (this is evident from the above-mentioned expression k̲h̲uṭiba lahū, for which we also find ʿalayhi : see Ibn Ḥawḳal 1, 20; al-Muḳaddasī, 337, 338, 400, 472, 485; cf. Glossarium, s.v.). There is also evidence that prayers used to be uttered for the heir-apparent (al-Maḳrīzī, iv, 37; Kitāb al-Wuzarāʾ, ed. Amedroz, 420). Under the Mamlūks also, the sultan’s heir was mentioned (Quatremère, Hist. Sult. Maml, ii/1, 101; ii/2, 3). Under the Fāṭimids, it was even the custom to call salām upon the ruler from the minaret after the ad̲h̲ān al-fad̲j̲r (al-Maḳrīzī, iv, 45); this also took place under the Mamlūks, e.g. in 696/1297, when Lād̲j̲īn was elected (Quatremère, Hist. Sult. Maml, ii/2, 45). The prayer for the sovereign in the k̲h̲uṭba did not find unanimous approval among the learned (see Snouck Hurgronje, Verspreide Geschriften, ii, 214-15).

In general, the mosque, and particularly the minbar, was the place where official proclamations were made, of course as early as the time of the Prophet (al-Buk̲h̲ārī. Ṣalāt, bābs 70, 71), ʿUt̲h̲mān’s bloodstained shirt was hung upon the minbar (al-Ṭabarī, i, 3255); messages from the caliph were read from it (ibid., iii, 2084). Al-Walīd announced from the minbar the deaths of two distinguished governors (Ibn Tag̲h̲rībirdī, i, 242); the results of battles were announced in k̲h̲uṭbas (Yāḳūt, i, 647; al-ʿIḳd al-farīd, Cairo 1321, ii, 149-50). In the Fāṭimid and ʿAbbāsid periods also, proclamations, orders, edicts about taxation, etc., by the ruler were announced in the principal mosque (al-Ṭabarī, ii, 40; iii, 2165; Ibn Tag̲h̲rībirdī, ii/2, 68; al-Maḳrīzī, Ittiʿāẓ, ed. Bunz, 87 above; Quatremère, Hist. Sult. Maml, i/2, 89; ii/2, 44, 151); documents appointing the more important officers were also read upon the minbar (al-Kindī, Wulāt, 589, 599, 603, 604, etc., passim; al-Maḳrīzī, ii, 246; iv, 43, 88); frequently the people trooped into the mosque to hear an official announcement (al-Kindī, Wulāt, 14; cf. Dozy, Gesch. d. Mauren in Spanien, ii, 170).

After the position of the caliph had changed, tradition was so far retained that he still delivered the k̲h̲uṭba in the principal mosque on special occasions, particularly at festivals. Thus the Fāṭimid al-ʿAzīz preached in the Mosque of al-Ḥākim on its completion (al-Maḳrīzī, iv, 55) and in the month of Ramaḍān he preached in the three chief mosques of Cairo, one after the other (ibid., 53, cf. 61-2; Ibn Tag̲h̲rībirdī, ii/1, 482 ff.; exceptionally also in al-Rās̲h̲ida: al-Maḳrīzī, iv, 63). The ʿAbbāsid caliph also used to preach at festivals (e.g. al-Rāḍī: Yāḳūt, Udabāʾ, ii, 349-50); it was the exception when a zealot like al-Muhtadī in the year 255/869 followed the old custom and preached every Friday (al-Masʿūdī, Murūd̲j̲, viii, 2 = § 3110). Even the fainéant caliph in Egypt preached occasionally (al-Maḳrīzī, iv, 94; Quatremère, Hist. Sult. Maml., ii/1, 138-9). Although the mosque lost its old political importance in its later history, it has never quite lost its character as the place of assembly on occasions of public importance. This is evident from al-D̲j̲abartī’s history, and even quite recently large meetings have been held in the mosques of Egypt on questions of nationalist politics.

2. The mosque and public administration. The actual work of government was very early transferred from the mosque into a special dīwān or mad̲j̲lis (see al-Ṭabarī, Glossarium, s.v.) and negotiations were carried on and business frequently done in the ḳaṣr al-imāra (cf. al-Ṭabarī, ii, 230-1). But when financial business had to be transacted at public meetings, the mosque was used; of this there is particular evidence from Egypt. Here the director of finance used to sit in the Mosque of ʿAmr and auction the farming out of the domains, with a crier and several financial officers to assist him. Later, the Dīwān was transferred to the D̲j̲āmiʿ of Aḥmad b. Ṭūlūn, but even after 300/912-13, we find Abū Bakr al-Mād̲h̲arāʾī sitting on such occasions in the Mosque of ʿAmr. Under the Fātīmids, the vizier Yaʿḳūb b. Killis used first the dār al-imāra of the Mosque of Ibn Ṭūlūn (see above); later his own palace and afterwards the caliph’s ḳaṣr was used (al-Maḳrīzī, i, 131-2). In the same way, in the reign of Muʿāwiya, the Coptic churches were used and the taxation commission took up their offices in them (Papyrus Erzherzog Rainer, Führer durch die Ausstelling, no. 577); and Ibn Rusta (ca. 290/903) says that the officials in charge of the measurement of the Nile, when they noticed the rising of the river, went at once to the chief mosque and announced it at one ḥalḳa after another, at the same time scattering flowers on those seated there (116).

The connection with administration was also seen in the fact that the treasury-chest, the bayt al-māl (identical with the tābūt ; al-Kindī, Wulāt, 70, 117) was kept in the mosque. In Fusṭāṭ, Usāma b. Zayd, the director of finance, in 97/715-16 and 99/717-18 built in the Mosque of ʿAmr a ḳubba, on pillars in front of the minbar for the bayt al-māl of Egypt. A drawbridge was placed between it and the roof. In the time of Ibn Rusta, it was still possible to move about freely below the ḳubba,, but in 378-9/988-9 al-ʿAzīz put up a running fountain below it (Ibn Rusta, 116; al-Maḳrīzī, iv, 9, 11, 13; al-Suyūṭī, ii, 136; Yāḳūt, iii, 899). Al-Kindī records an attempt to steal the chest in 145/762 (Wulāt, 112-13). In the disturbed period around the year 300/912, the wālī al-Nūs̲h̲arī closed the mosque between the times of ṣalāt for the safety of the chest, which was also done in Ibn Rusta’s time (al-Kindī, Wulāt, 266; Ibn Rusta, 116). New approaches to the bayt al-māl were made in 422/1031 from the k̲h̲izāna of the mosque and from the Dīwān (al-Maḳrīzī, iv, 13).

In Kūfa, the buyūt al-amwāl, at least during the early period, were in the Dār al-Imāra (al-Ṭabarī, i, 2489, 2491-2); in the year 38/658-9, during the fighting, it was saved from Baṣra and taken with the minbar to the Mosque of al-Ḥuddān (ibid., 3414-15). In Palestine, in the chief mosque of each town, there was a similar arrangement to that in the Mosque of ʿAmr (al-Muḳaddasī, 182). In Damascus the bayt al-māl was in the most western of the three ḳubbas in the court of the Mosque of the Umayyads; it was of lead and rested on 8 columns (ibid., 157; Ibn D̲j̲ubayr, 264, 267; Ibn Baṭṭūṭa, i, 200-1); it is still called ḳubbat el-k̲h̲azne (“treasure-cupola”, earlier ḳubbat ʿĀʾis̲h̲a (cf. Baedeker, Palästina und Syrien). In the time of the two travellers mentioned, the ḳubba, only contained property of the mosque. Ibn D̲j̲ubayr saw a similar ḳubba, in the chief mosque of Ḥarrān and says that it came from the Byzantines (246). In Ād̲h̲arbāyd̲j̲ān, also by the time of al-Iṣṭak̲h̲rī, the Syrian custom had been everywhere introduced (184); in Īrāns̲h̲ahr, in the centre of the court, there was a building with marble colums and doors (al-Muḳaddasī, 316), which perhaps points to a similar state of affairs, and in Armenia, it is recorded that the bayt al-māl was kept in the D̲j̲āmiʿ in the time of the Umayyads as in Miṣr and elsewhere (Ibn Ḥawḳal 1, 241). The ḳubba was usually of lead and had an iron door. Ibn al-Ḥād̲j̲d̲j̲ considers it highly illegal to shut off a dīwān in a mosque, since this is the same as forbidding entrance to it. This shows that the custom still survived in his time.

Ibn D̲j̲ubayr’s remark about Ḥarrān suggests that here again we have an inheritance from Byzantium. It was probably the building belonging to the piscina (cf. above) that the Muslims put to a practical use in this way. For the Byzantines had the treasury (sakellē) in the palace, and it is doubtful if the treasure-chambers of the church (skenophylakion) were built in this way (cf. F. Dölger, in Byzantinisches Archiv, Heft 9 [1927], 26, 34).

3. The mosque as a court of justice. That the Prophet used to settle legal questions in his mosque was natural (see al-Buk̲h̲ārī, Aḥkām, bābs 19, 29, etc.; cf. Ṣalāt, bāb 71; K̲h̲uṣūmāt, bob 4), but he could also deliver judgments in other places (ibid., passim). In Ḥadīt̲h̲, it is recorded that some ḳāḍīs of the earlier period (S̲h̲urayḥ, al-S̲h̲aʿbī, Yaḥyā b. Yaʿmar, Marwān) sat in judgment beside the minbar, others (al-Ḥasan, Zurāʿa b. Awfā) on the open square beside the mosque (al-Buk̲h̲ārī, Aḥkām, bāb 18). The custom had all the better chance of survival, as churches were used in the same way (Joshua Stylites, ed. Wright, ch. 29; cf. Mez, Renaissance, 223, Eng. tr., 224). Sitting in judgment was primarily the business of the ruler, but he had to have assistants and Abū Bakr’s ḳāḍī is mentioned as assisting ʿUmar (al-Ṭabarī, i, 2135), and a number of judges appointed by ʿUmar are mentioned (Ibn Rusta, 227). In the reign of ʿUt̲h̲mān, ʿAbd Allāh b. Masʿūd is said to have been judge and financial administrator of Kūfa (Ibn Ḳutayba, Maʿārif, ed. Wüstenfeld, 128). On the other hand, we are told that ʿAbd Allāh b. Nawfal, appointed by Marwān in 42/662, was the first ḳāḍī in Islam (al-Ṭabarī, iii, 2477); it is recalled that in the year 132/749-50 the ḳāḍī of Medina administered justice in the mosque (ibid., 2505). In Baṣra, we are told that al-Aswad b. Sarīʿ al-Tamīmī immediately after the building of the mosque (i.e. in the year 14/635) worked in it as ḳāḍī (al-Balād̲h̲urī, 346). In the early period, ʿUmar wanted to choose a ḳāḍī, who had been already acting as a judge before Islam (al-Kindī, Wulāt, 301-2; al-Suyūṭī, ii, 86). Even the Christian poet al-Ak̲h̲ṭal was allowed to act as arbiter in the mosque of Kūfa (see Lammens, Moʿâwia, 435-6).

In Fusṭāṭ, as early as 23/643 or 24/644 by command of ʿUmar, ʿAmr b. al-ʿĀṣ appointed a ḳāḍī named Ḳays (al-Suyūṭī, ii, 86; al-Kindī, 300-1). The ḳāḍī held his sessions in the Mosque of ʿAmr but not exclusively there. The ḳāḍī K̲h̲ayr b. Nuʿaym (120-7/738-45) held his sessions sometimes before his house, sometimes in the mosque, and for Christians on the steps leading up to the mosque (al-Kindī, 351-2). A successor of his (177-84/793-800) invited Christians who had lawsuits into the mosque to be heard (ibid., 391); of another judge (205-11/820-6), it is recorded that he was not allowed to sit in the mosque (ibid., 428). It seems that the ḳāḍī could himself choose where he would sit. A judge, officiating in the year 217/832, sat in winter in the great pillared hall, turning his back towards the ḳibla wall, and in summer, in the ṣaḥn near the western wall (ibid., 443-4). During the Fāṭimid period, the subsidiary building on the north-east of the Mosque of ʿAmr was reserved for the judge. This judge, called from the year 376/986 onwards ḳāḍī ’ l-ḳuḍāt (cf. al-Suyūṭī, ii, 91; al-Kindī, 590), sat on Tuesday and Saturday in the mosque and laid down the law (al-Maḳrīzī, ii, 246; iv, 16, 22; cf. al-Kindī, 587, 589; cf. Nāṣir-i K̲h̲usraw, tr. Schefer, 149).

In al-Yaʿḳūbī’s time in Bag̲h̲dād, the judge of the east city used to sit in its chief mosque (Buldān, 245), in Damascus the vice-ḳāḍī in the 4th/10th century had a special riwāḳ in the Mosque of the Umayyads (al-Muḳaddasī, 158), and the notaries (al-s̲h̲urūṭiyyūn) also sat at the Mosque of the Umayyads at the Bāb al-Sāʿāt (ibid., 17). In Naysābūr, every Monday and Thursday, the mad̲j̲lis al-ḥukm was held in a special mosque (ibid., 328). In course of time, the judge was given a mad̲j̲lis al-ḥukm of his own (cf. al-Suyūṭī, ii, 96), and in 279/892 al-Muʿtaḍid wanted to forbid the ḳāḍīs to hold sessions in the mosques (Ibn Tag̲h̲rībirdī, ii/1, 87 above; perhaps, however, we should read ḳāṣṣ : see Goldziher, Muh. Stud., ii, 164, n. 4). Justice was also administered in the dār al-ʿadl (Quatremère, Hist. Sult. Maml, ii/2, 79). But the administration of justice did not at once lose all connection with the mosque. Under the Fāṭimids, the custom had been introduced that the ḳāḍī should hold sittings in his house, but Ibn al-ʿAwwām, appointed just after 400/1009-10, held them either in the D̲j̲āmiʿ at the Bayt al-Māl or in a side-room (al-Kindī, 612; cf. Ibn Tag̲h̲rībirdī. ed. Popper, ii/2, 69; al-Ḳalḳas̲h̲andī, Ṣubḥ al-aʿs̲h̲āʾ, iii, 487: for 439/1046, see Nāṣir-i K̲h̲usraw, ed. Schefer, text, 51, tr. 149). In Mecca, the dār al-ḳāḍī was in direct connection with the mosque (Ibn D̲j̲ubayr, 104). In the 8th/14th century, Ibn Baṭṭūṭa attended a court presided over by an eminent jurist in a mosque (madrasa) in S̲h̲īrāz (ii, 55, 63; cf. also Madk̲h̲al, ii, 54 below), and in Damascus the S̲h̲āfiʿī chief ḳāḍī held his sessions in the ʿĀdiliyya Madrasa (so Ibn K̲h̲allikān, in Quatremère, Hist. Sult. Maml, ii/1, 22; cf. also for Egypt: ibid., 87, ii/2, 253), the vice-ḳāḍīs sat in the Ẓāhiriyya Madrasa (Ibn Baṭṭūṭa, i, 218). The judgment might even be put into execution in the madrasa (ibid., 220). During the Mamlūk period in Egypt, we occasionally find a small mosque being used as a mad̲j̲lis for judges (al-Maḳrīzī, iv, 270; Ibn Duḳmāḳ, 98 above); Ibn K̲h̲aldūn held legal sittings in the Madrasa al-Ṣāliḥiyya (ʿIbar, vii, 453).

A mufṭī, especially in the large mosques, was also frequently appointed; he sat at definite times in a ḥalḳa li ’l-fatwā, e.g. in Cairo (al-Ḳazwīnī, in al-Suyūṭī, i, 182; D̲j̲alāl al-Dīn, ibid., 187), in Tunis (al-Zarkas̲h̲ī, Chronicle, tr. Fagnan, in Rec. Mém. Soc. Arch. Constantine, xxi [1895], 197, 202, 218, 248). In Bag̲h̲dād, Abū Bakr al-Dīnawarī (d. 405/1014-15) was the last to give fatwās in the Mosque of al-Manṣūr according to the mad̲h̲hab of Sufyān al-T̲h̲awrī (Ibn Tag̲h̲rībirdī, ed. Popper, ii/2, 120).

F. The administration of the mosque

1. Finances. The earliest mosques were built by the rulers of the various communities, and the members of the community did all the work necessary in connection with the primitive mosques. The later mosques as a rule were erected by rulers, amīrs, high officials or other rich men in their private capacity and maintained by them. The erection of the mosque of Ibn Ṭūlūn cost its builder 120,000 dīnārs, the Mosque of Muʾayyad 110,000 (al-Maḳrīzī, iv, 32, 137, 138). The upkeep of the mosque was provided for by estates made over as endowments (waḳf, ḥabs) (cf. thereon besides the fiḳh books, I. Krcsmárik, Das Waḳfrecht, in ZDMG, xlv [1891], 511-76; E. Mercier, Le code du hobous ou ouakf selon la législation musulmane, 1899). In the 3rd/9th century we thus hear of houses which belonged to the mosque and were let by them (Papyrus Erzherzog Rainer, Führer, nos. 773, 837), and Ibn Ṭūlūn handed over a large number of houses as an endowment for his mosque and hospital (al-Maḳrīzī, iv, 83). This custom was taken over from the Christians by the Muslims (see Becker, in Isl., ii [1911], 404). According to al-Maḳrīzī, estates were not given as waḳf endowments until Muḥammad Abū Bakr al-Mād̲h̲arāʾī (read thus) bequeathed Birkat al-Ḥabas̲h̲ and Suyūṭ as endowments (about 300/912-13; this was however cancelled by the Fāṭimids again (ibid.). Al-Ḥākim made large endowments not only for his own, but also for mosques previously in existence, such as the Azhar, al-Ḥākimī, Dār al-ʿIlm and D̲j̲āmiʿ al-Maḳs and D̲j̲āmiʿ Rās̲h̲ida; the endowments consisted of dwelling-houses, shops, mills, a ḳaysāriyya and ḥawānīt, and the document (ibid., 50-1) specifies how and for what purposes the revenues are to be distributed. Baths were also given as endowments for mosques (ibid., 76, for 529/1135; cf. 81 for the year 543/1148-9). Ṣalāḥ al-Dīn granted lands to his madāris: in 566/1170-1, for example, a ḳaysāriyya to the Ḳamḥiyya and a ḍayʿa in al-Fayyūm, and the teachers received wheat from al-Fayyūm; in the same year he endowed the Nāṣiriyya with goldsmiths’ shops and a village (ibid., 193-4; cf. another document, 196-7). During the Mamlūk period also, estates were given as endowments (for documents of this period, see van Berchem, CIA, i, nos. 247, 252, 528; Moberg, in MO, xii [1918], 1 ff.; JA, ser. 9, iii, 264-6; ser. 11, x, 158 ff., 222 f.; xii, 195 ff., 256 ff., 363 ff.). They were often a considerable distance apart: the mosques in Egypt often had estates in Syria (van Berchem, CIA, i, no. 247; al-Maḳrīzī, vi, 107, 137). Not only were mosques built and endowed, but already existing ones were given new rooms for teachers, minbars, stipends for Ḳurʾān reciters, teachers, etc. There were often special endowments for the salaries of the imām and the muʿad̲h̲d̲h̲ins, for the support of visitors, for blankets, food, etc. (see Ibn D̲j̲ubayr, 277 with reference to the Mosque of the Umayyads). The endowments, and the purpose for which they might be used, were precisely laid down in the grant and the document attested in the court of justice by the ḳāḍī and the witnesses (cf. al-Maḳrīzī, iv, 50, 196 below). The text was also often inscribed on the wall of the mosque (cf. ibid., 76; the above-mentioned inscriptions amongst others. For documents from Tas̲h̲kent, see RMM, xiii [1911], 278 ff.). Certain conditions might be laid down, e.g. in a madrasa that no Persian should be appointed there (al-Maḳrīzī, iv, 202 below), or that the teacher could not be dismissed or some such condition (van Berchem, CIA, i, no. 201); that no women could enter (JA, ser. 9, iii, 389); that no Christian, Jew or Ḥanbalī could enter the building (ibid., 405); etc. Endowments were often made with stipulations for the family of the founder or other purposes. That mosques could also be burdened with expenses is evident from an inscription in Edfū of the year 797/1395 (van Berchem, CIA, i, no. 539). If a mosque was founded without sufficient endowment, it decayed (e.g. al-Maḳrīzī, iv, 115, 201, 203) or else the stipends were reduced (ibid., 251), but in the larger mosques as a rule the rulers provided new endowments. According to al-Māwardī, there were also special “Sultan mosques” which were directly under the patronage of the caliph and their officials paid from the bayt al-māl (al-Aḥkām al-sulṭāniyya, ed. Enger, 172 above, 176 above).

Just as the bayt al-māl of the state was kept in the mosque, so was the mosque’s own property kept in it, e.g. the kanz or k̲h̲izānat al-Kaʿba, which is mentioned in ʿUmar’s time and may be presumed to have existed under his predecessors (al-Balād̲h̲urī, 43 above; Chron. Mekka, i, 307, ii, 14). The Bayt Māl al-Ḏj̲amiʿ in Damascus was in a ḳubba, in the ṣaḥn (al-Muḳaddasī, 157; Ibn D̲j̲ubayr, 267; Ibn Baṭṭūṭa, i, 201. cf. for Medina, Wüstenfeld, Medina, 86). Rich men also had their private treasure-chambers in the mosque (see above, I. E. 2), as used to be the case with the Temple at Jerusalem (see E. Schürer, Gesch. d. jüd. Volkes 4, ii, 1907, 322-8; F. Cumont, Fouilles de Doura-Europos, 1926, 405-6).

2. Administration. As Imām of the Muslim community, the caliph had the mosques under his charge. This was also the case with the sultan, governor or other ruler who represented the caliph in every respect. The administration of the mosques could not however be directly controlled by the usual government offices. By its endowment, the mosque became an object sui generis and was withdrawn from the usual state or private purposes. Their particular association with religion gave the ḳāḍīs special influence, and, on the other hand, the will of the testator continued to prevail. These three factors decided the administration of the mosque, but the relation between them was not always clear.

a. Administration of the separate mosques. The mosque was usually in charge of a nāẓir or walī who looked after its affairs. The founder was often himself the nāẓir or he chose another and after his death, his descendants took charge or whoever was appointed by him in the foundation charter. In the older period, the former was the rule and it is said to have applied especially in the case of chief mosques, if we may believe Nāṣir-i K̲h̲usraw, according to whom al-Ḥākim paid the descendants of Ibn Ṭūlūn 30,000 dīnār s for the mosque and 5,000 for the minaret, and similarly to the descendants of ʿAmr b. al-ʿĀṣ 100,000 dīnārs for the Mosque of ʿAmr (Safar-nāma, ed. Schefer, text 39-40, tr. 146, 148). In 378/988 we read of an administrator (mutawallī) of the mosque in Jerusalem (al-Maḳrīzī, iv, 11). In the case of mosques and madāris founded during the Mamlūk period, it is often expressly mentioned that the administration is to remain in the hands of the descendants of the founder, e.g. in the case of a mosque founded by Baybars (al-Maḳrīzī, iv, 89), in the D̲j̲āmiʿ Maḳs when the vizier al-Maḳsī renovated it (ibid., 66), the Ṣāḥibiyya (ibid., 205), and the Ḳarāsunḳuriyya (ibid., 232), etc.; so also in the Badriyya in Jerusalem (“to the best of the descendants”, cf. van Berchem, CIA, ii/1, 129). Other cases are also found. Sometimes an amīr or official was administrator, e.g. in the Muʾayyad (al-Maḳrīzī, iv, 140), the Ṭaybarsiyya (ibid., 224), the Azhar (ibid., 54-5) or the Mosque of Ibn Ṭūlūn (al-Ḳalḳas̲h̲andī, Ṣubḥ, xi, 159-62). In D̲j̲amāl al-Dīn’s madrasa, it was always the kātib al-sirr (al-Maḳrīzī, iv, 256), in the k̲h̲ānaḳāh of Baybars the k̲h̲āzindār and his successors (van Berchem, CIA, i, no. 252); but it was more frequently a ḳāḍī; for example, in the mosque of Baybars just mentioned, the Ḥanafī ḳāḍī was to take charge after the descendants (al-Maḳrīzī, iv, 89); in the Āḳbug̲h̲awiyya, the S̲h̲āfiʿī ḳāḍī was appointed but his descendants were expressly excluded (ibid., 225). In the Mosque of the Umayyads, during the Mamlūk period the S̲h̲āfiʿī chief ḳāḍī was as a rule the nāẓir (al-Ḳalḳas̲h̲andī, iv, 191), and thus also in the Nāṣir mosque in Cairo (ibid., xi, 262-4). In this city, we find during the Mamlūk period that amīrs and ḳāḍīs alternately acted as nāẓirs in the large mosques (e.g. the Mosque of Ibn Ṭūlūn, al-Maḳrīzī, iv, 42). Cases are also found, however, in which descendants of the founder unsuccessfully claimed the office of nāẓir (al-Maḳrīzī, iv, 218, 255). This was the result of the increasing power of the ḳāḍīs (see below). In the madāris, the nāẓir was often also the leading professor; the two offices were hereditary (ibid., 204, the Ṣāḥibiyya al-Bahāʾiyya; and 238 above, the D̲j̲amāliyya). In Tustar, a descendant of Sahl as nāẓir and teacher conducted a madrasa with the help of four slaves (Ibn Baṭṭūṭa, ii, 25-6).

The nāẓir managed the finances and other business of the mosque. Sometimes he had a fixed salary (in Baybars’ k̲h̲ānaḳāh, 500 dirhams a month, van Berchem, CIA, i, 252; in the Dulāmiyya in Damascus in 847/1443-4, only 60 dirhams a month, JA, ser. 9, iii, 261), but the revenues of the mosque were often applied to his personal use. His control of the funds of the mosque was however often limited by the central commission for endowments (see below). The nāẓir might also see to any necessary increase of the endowments. He appointed the staff and he fixed their pay (cf. e.g. al-Maḳrīzī, iv, 41). He could also interfere in questions not arising out of the business side of administration; for example, the amīr Sawdūb, the nāẓir of the Azhar in 818/1415-16, ejected about 750 poor people from the mosque. He was however thrown into prison for this by the sultan (ibid., 54). Generally speaking, the nāẓir’s powers were considerable. In 784/1382 a nāẓir in the Azhar decided that the property of a mud̲j̲āwir, who had died without heirs, should be distributed among the other students (ibid., 54). In Mecca, according to Ḳuṭb al-Dīn, the Nāẓir al-Ḥarām was in charge of the great festival of the mawlid of the Prophet (12 Rabīʿ I) and distributed robes of honour in the mosque on this occasion (Chron. Mekka, iii, 349). In the Azhar, no nāẓir was appointed after about 493/1100 but a learned man was appointed S̲h̲ayk̲h̲ al-Azhar, principal and administrator of the mosque (Sulaymān Raṣad al-Zayyātī, Kanz al-d̲j̲awhar fī taʾrīk̲h̲ al-Azhar, 123 ff.). Conditions were similar in Mecca in the late 19th century (Snouck Hurgronje, Mekka, ii, 235-6, 252-3).

As we have seen, ḳāḍīs were often nāẓirs of mosques. This was especially the case in the madāris, where the ḳāḍīs were often teachers (cf. al-Maḳrīzī, iv, 209, 219, 222, 238, etc.); the ḳāḍīs were particularly anxious to get the principal offices in the large schools (cf. al-Ḳalḳas̲h̲andī, xi, 235). Their influence was however further increased by the fact that, if a nāẓir qualified by the terms of the founder’s will no longer existed, the ḳāḍī of the mad̲h̲ab in question stepped into his place (cf. ZDMG, xlv [1897], 552). By this rule, which often gave rise to quarrels between the different ḳāḍīs (e.g. al-Maḳrīzī, iv, 218, the Ẓāhiriyya), a. ḳāḍī could accumulate a larger number of offices and “milk the endowments” (ibid., iii, 364). Sometimes their management was so ruthless that the schools soon declined (e.g. the Ṣāḥibiyya and the D̲j̲amāliyya, al-Maḳrīzī, iv, 204-5, 238). They also exercised influence through the committee of management of the mosque.

b. Centralisation in the management of the mosques. The large mosques occupied a special position in the Muslim lands, because the caliph had to interest himself particularly in them, especially those of Mecca and Medina, where the rulers and their governors built extensions and executed renovations (cf. Chron. Mekka, i, 145; iii, 83 ff.). During the ʿAbbāsid period, the ḳāḍī occasionally plays a certain part in this connection; for example al-Mahdī (158-69/775-85), presented the ḳāḍī with the necessary money to extend and repair the Meccan mosque (ibid., i, 312; ii, 43). In 263/877, al-Muwaffaḳ ordered the governor of Mecca to undertake repairs at the Kaʿba (ibid., ii, 200-1). In 271/1884-5, the governor and the ḳāḍī of Mecca co-operated to get money from al-Muwaffaḳ for repairs, and they saw the work through (ibid., iii, 136-7). In 281/894, the ḳāḍī of Mecca wrote to the vizier of al-Muʿtaḍid about the Dār al-Nadwa and backed up his request by sending a deputation of the staff there (sadana). The caliph then ordered the vizier to arrange the matter through the ḳāḍī of Bag̲h̲dād and a man was sent to Mecca to take charge of the work (ibid., iii, 144 ff.).

The importance of the ḳāḍī was based primarily on his special knowledge in the field of religion. A zealous ḳāḍī like al-Ḥārit̲h̲ b. Miskīn in Cairo (237-45/851-9) forbade the ḳurrāʾ of a mosque to recite the Ḳurʾān melodiously; he also had the maṣāhif ’in the mosque of ʿAmr inspected and appointed an amīn to take charge of them (al-Kindī, Wulāt, 469). After the building of the Ṭūlūnid mosque, a commission was appointed under the ḳāḍī ’ l-ḳuḍāt to settle the ḳibla of the mosque (al-Maḳrīzī, iv, 21-2). But at a quite early date they also obtained a say in the management of the funds. The first ḳāḍī to lay his hands on the aḥbās was Tawba b. Namir al-Ḥaḍramī; while hitherto every endowment had been administered by itself by the children of the testator or someone appointed by him, in 118/736 Tawba brought about the centralisation of all endowments and a large dīwān was created for the purpose (al-Kindī, 346). How this system of centralisation worked is not clear at first, but it was carried through under the Fāṭimids.

Al-Muʿizz created a special dīwān al-aḥbās and made the chief ḳāḍī head of it as well as of the d̲j̲awāmiʿ wa ’l-mas̲h̲āhid (al-Maḳrīzī, iv, 83 and 75; cf. al-Kindī, 585, 587, 589, according to whom al-ʿAzīz specially appointed the chief ḳāḍī over the two d̲j̲āmiʿ s), and a special bayt al-māl was instituted for it in 363/974; a yearly revenue of 150,00 dirhams was guaranteed; anything left over went to form a capital fund. All payments were made through his office after being certified by the administration of the mosque (al-Maḳrīzī, iv, 83-4). The mosques were thus administered by the ḳāḍīs, directly under the caliph. The dīwān al-birr wa ’l-ṣadaḳa in Bag̲h̲dād (Mez, Renaissance, 72, Eng. tr., 80) perhaps served similar purposes.

Al-Ḥākim reformed the administration of the mosques. In 403/1012-13 he had an investigation made, and when it proved that 800 (or 830) had no income (g̲h̲alla), he made provision for them by a payment of 9,220 dirhams monthly from the Bayt al-Māl ; he also made 405 new endowments (of estates) for the officials of the mosque (al-Maḳrīzī, iv, 84, 264). Under the Fāṭimids, the ḳāḍīs used to inspect all the mosques and mas̲h̲āhid in and around Cairo at the end of Ramaḍān and compare them with their inventories (ibid., 84). The viziers of the Fāṭimids, who also had the title ḳāḍī, did much for the mosques (D̲j̲awhar, Yaʿḳūb b. Killis, Badr al-D̲j̲amālī, cf. van Berchem, CIA, i, nos. 11, 576, 631).

Under the Ayyūbids, conditions were the same as under the Fāṭimids. The dīwān al-aḥbās was under the ḳāḍīs (al-Maḳrīzī, iv, 84). Ṣalāḥ al-Dīn gave a great deal to the mosques, especially the madāris: 20,000 dirhams a day is a figure given (ibid., 117). When Ibn D̲j̲ubayr says that the sultan paid the salaries of the officials of the mosques and schools of Alexandria, Cairo and Damascus 43, 52, 275), he must really mean the Dīwān already mentioned.

The same conditions continued for a time under the Mamlūks. In the time of Bay bars, for example, the chief ḳāḍī Tād̲j̲ al-Dīn was nāẓir al-aḥbās. He caused the Mosque of ʿAmr to be renovated, and when the funds from the endowments were exhausted, the sultan helped him from the Bayt al-Māl (al-Maḳrīzī, iv, 14); after conferring with experts, the chief ḳāḍī forbade a water-supply brought by Ṣalāh al-Dīn into the mosque (ibid., 14; al-Suyūṭī, ii, 137). In 687 the chief ḳāḍī Taḳī al-Dīn complained to Ḳalāwūn that the ʿAmr and Azhar mosques were falling into ruins, while the aḥbās were much reduced. The sultan would not however permit their restoration but entrusted the repairs of the mosques to certain amīrs, one to each (al-Maḳrīzī, iv, 14, 15). This principle was several times applied in later times, and the amīrs frequently gained influence at the expense of the ḳāḍīs. Thus after the earthquake of 707/1303 (cf. thereon Quatremère, Hist. Sult. Maml, ii/2, 214 ff.), the mosques were allotted to amīrs, who had to see that they were rebuilt (al-Maḳrīzī, iv, 15, 53). From the middle of the 7th/13th century, we often find amīrs as administrators of the chief mosques. The ḳāḍī had however obtained so much authority that he was conceded “a general supervision of all matters affecting the endowments of his mad̲h̲hab ” (al-ʿUmarī, al-Taʿrīf bi ’l-muṣṭalaḥ al-s̲h̲arīf, 117; cf. ZDMG, xlv [1891], 559); according to this theory the ḳāḍī could intervene to stop abuses. In Syria in 660/1262 Ibn K̲h̲allikān became ḳāḍī over the whole area between al-ʿArīs̲h̲ and the Euphrates and superintendent of waḳfs, mosques, madrasas, etc. (Quatremère, Hist. Sult. Maml, i/1, 170).

Sultan Baybars reformed these endowments and restored the office of nāẓir al-awḳāf or nāẓir al-aḥbās al-mabrūra or n. d̲j̲ihāt al-birr (al-Ḳalḳas̲h̲andī, iv, 34, 38; v, 465; ix, 256; xi, 252, 257; cf. K̲h̲alīl al-Ẓāhirī, Zubdat kas̲h̲f al-mamālik, ed. Ravaisse, 109). According to al-Maḳrīzī, the endowments were distributed among the Mamlūks in three departments (d̲j̲ihāt): 1. d̲j̲ihāt al-aḥbās, managed by an amīr, the Dawādār : this looked after the lands of the mosques, in 740/1339-40, in all 130,000 faddāns 2. d̲j̲ihāt al-awḳāf al-ḥukmiyya bi-Miṣr wa ’l-Ḳāhira, which administered dwelling-houses; it was managed by the S̲h̲āflʿī ḳāḍī ’l-ḳuḍāt, with the title Nāẓir al-Awḳāf. This department came to an end in the time of al-Malik al-Nāṣir Farad̲j̲ because an amīr, supported by the opinion of the Ḥanafī chief ḳāḍī, spent a great deal and misused the funds; 3. d̲j̲ihāt al-awḳāf al-ahliyya, comprised all the endowments which still had particular nāẓirs, either descendants of the testator or officials of the sultan and the ḳāḍī. The amīrs seized their lands and Barḳūḳ, before he became sultan, sought in vain to remedy the evil by appointing a commission. The endowments in general disappeared somewhat later because the ruling amīrs seized them (al-Maḳrīzī, iv, 83-6). In modern times, as a rule, endowments in Muslim lands have been combined under a special ministry, a Wizārat al-Awḳāf.

To be distinguished from the administrators of the mosque is the nāẓir who is only concerned with the supervision of the erection of mosques. Anyone could be entrusted with the building of a mosque (e.g. al-Maḳrīzī, iv, 92). Under the Mamlūks, there was also a clerk of works, mutawallī s̲h̲add al-ʿamāʾir or nāẓir al-ʿimāra : he was the overseer of the builders (ibid., 102; see Zubdat kas̲h̲f al-mamālik, ed. Ravaisse, 115, cf. 109; van Berchem, CIA, i, 742, no. 751).

The caliph or the ruler of the country was in this, as in other matters, supreme. As we have seen, he intervened in the administration and directed it as he wished. He was also able to interfere in the internal affairs of the mosque, if necessary through his usual officers. In 253/867 after the rising in the Fayyūm, the chief of police issued strict orders by which it was forbidden to say the basmala aloud in the mosque; the number of prayers in the month of Ramaḍān was cut down, the ad̲h̲ān from the minaret forbidden, etc. (Papyrus Erzherzog Rainer, Führer, 788). In the year 294/908, the governor ʿĪsā al-Nūs̲h̲arī had the Mosque of ʿAmr closed except at the ṣalāts, because the bayt al-māl was kept in it, which however produced protests from the people (al-Maḳrīzī, iv, 11; al-Kindī, Wulāt, 266; Ibn Rusta, 116). Many similar examples could be mentioned, especially during periods of unrest. In 205/821 the nāʾib, in conjunction with the ḳāḍīs, revised the budget of the Mosque of the Umayyads and made financial reforms (JA, ser. 9, vii, 220). The ad̲h̲ān formulae were laid down in edicts by the ruler (al-Maḳrīzī, iv, 44, 45). In the year 323/935 the vizier in Bag̲h̲dād had a man whipped who had recited a variant text of the Ḳurʾān in the miḥrāb, after he had been heard in his defence in the presence of the ḳāḍīs and learned men (Yāḳūt, Udabāʾ, vi, 300). The importance of the sovereign in connection with the mosque depended on his personality. As a rule, he recognised the authority of the regular officials. When, for example, al-K̲h̲aṭīb al-Bag̲h̲dādī asked the caliph al-Ḳāʾim for authority to read ḥadīt̲h̲ in the mosque of al-Manṣūr, the latter referred the question to the naḳīb al-nuḳabāʾ (Yāḳūt, Udabāʾ, i, 246-7; cf. Wüstenfeld, Schâfiʿî, iii, 280).

The consecration of the mosque was attended by certain ceremonies. When, for example, the midday worship was conducted for the first time in the D̲j̲āmiʿ al-Ṣāliḥ in Cairo, a representative from Bag̲h̲dād was present (al-Maḳrīzī, iv, 81). At the consecration of the Mosque of Ibn Ṭūlūn, the builder gave al-Rabīʿ b. Sulaymān, a pupil of al-S̲h̲āfiʿī. who lectured on ḥadīt̲h̲ there, a purse of 1,000 dīnārs (al-Suyūṭī, ii, 139). Al-Maḳrīzī describes the consecration ceremony at several mosques. In the Mosque of al-Muʾayyad the sultan was present seated on a throne surrounded by his officers; the basin of the ṣaḥn was filled with sugar and ḥalwa, the people ate and drank, lectures were given, then the ṣalāt was read and k̲h̲uṭba delivered and the sultan distributed robes of honour among the officials of the mosques and Ṣūfīs (al-Maḳrīzī, iv, 139); similarly at the Ẓahiriyya in 662/1264 where poems were also recited: cf. Quatremère, Hist. Sult. Maml, i/1, 228), Madrasat D̲j̲amāl al-Dīn, in 811/1408-9; al-Ṣarg̲h̲itmis̲h̲iyya, 757 (al-Makrīzī, iv, 217-18, 253, 256).

G. The personnel of the mosque.

1. The Imām. From the earliest days of Islam, the ruler was the leader of the ṣalāt; he was imām as leader in war, head of the government and leader of the common ṣalāt. The governors of provinces thus became leaders of the ṣalāt and heads of the k̲h̲arād̲j̲, and when a special financial official took over the fiscal side, the governor was appointed ʿalā ’l-ṣalāt wa ’l-ḥarb. He had to conduct ritual prayer, especially the Friday ṣalāt, on which occasion he also delivered the k̲h̲uṭba. If he was prevented, the chief of police, ṣāḥib al-s̲h̲urṭa, was his k̲h̲alīfa (cf. al-Maḳrīzī, iv, 83). ʿAmr b. al-ʿĀṣ permitted the people of the villages to celebrate the two festivals, while the Friday divine service could only take place under those qualified to conduct it (who could punish and impose duties; ibid., 7). This was altered under the ʿAbbāsids. The caliph no longer regularly conducted the ṣalāts (after the conquest of the Persians; al-Maḳrīzī, iv, 45), and ʿAnbasa b. Isḥāḳ, the last Arab governor of Egypt (238-42/852-6), was also the last amīr to conduct the ṣalāt in the d̲j̲āmiʿ. An imām, paid out of the bayt al-māl, was now appointed (ibid., 83), but the governor still continued to be formally appointed ʿalā H-ṣalāt. Henceforth, the ruler only exceptionally conducted the service, for example, the Fāṭimids on ceremonial occasions, especially in the month of Ramaḍān (Ibn Tag̲h̲rībirdī, ed. Juyn-boll, ii, 482 ff.; al-Ḳalḳas̲h̲andī, iii, 509 ff.); in many individual mosques, probably the most prominent man conducted the service; according to the ḥadīt̲h̲, the one with the best knowledge of the Ḳurʾān and, failing him, the eldest, should officiate (al-Buk̲h̲ārī, Ad̲h̲ān, bābs 46, 49).

The imām appointed was chosen from among those learned in religious matters; he was often a Hās̲h̲imite (Mez, Renaissance, 147, Eng. tr., 150); he might at the same time be a ḳāḍī or his nāʾib (see al-Ḳindī, 575, 589; Ibn Baṭṭūṭa, i, 276-7). During the ṣalāt he stood beside the miḥrāb; al-Muḳaddasī mentions the anomaly that in Syria one performed one’s ṣalāt “in front of the imām” (202). He could also stand on an elevated position; on one occasion Abū Hurayra conducted the ṣalāt in the Meccan mosque from the roof (al-Buk̲h̲ārī, Ṣalāt, bāb 17). In Mecca, in Ibn D̲j̲ubayr’s time, each of the four recognised mad̲h̲āhib (with the Zaydīs in addition) had an imām; they conducted the ṣalāt, one after the other each in his place, in the following order: S̲h̲āfiʿīs. Mālikīs, Ḥanafīs and Ḥanbalīs; they only performed the ṣalāt al-mag̲h̲rīb together; in Ramaḍān, they held the tarāwīḥ in different places in the mosque, which was also often conducted by the ḳurrāʾ (Riḥla, 101, 102, 143-4). This is still the case; very frequently one performs the ṣalāt, not after the imām of one’s own mad̲h̲hab (Snouck Hurgronje, Mekka, ii, 79-80). In Jerusalem, according to Mud̲j̲īr al-Dīn, the order was: Mālikīs, S̲h̲āfiʿīs, Ḥanafīs and Ḥanbalīs, who prayed each in their own part of the Ḥaram; in Hebron the order was the same (Sauvaire, Hist. Jér. et Hébron, 136-7). In Ramaḍān, extraordinary imāms were appointed (ibid., 138).

When the imām no longer represented a political office, each mosque regularly had one. He had to maintain order and was in general in charge of the divine services in the mosque. In al-Muḳaddasī’s time the imām of the Mosque of ʿAmr read a d̲j̲uzʾ of the Ḳurʾān every morning after the ṣalāt (205). It was his duty to conduct every ṣalāt, which is only valid fī d̲j̲amāʿa. He must conform to the standards laid down in the law; but it is disputed whether the ṣalāt is invalid in the opposite case. According to some, the leader of the Friday ṣalāt should be a different man from the leader of the five daily ṣalāts (al-Māwardī, al-Aḥkām al-sulṭāniyya, ed. Enger, 171; Ibn al-Ḥād̲j̲d̲j̲, Madk̲h̲al, ii, 41’, 43 ff., 50, 73 ff.; al-Subkī, Muʿīd al-niʿam, ed. Myhrman, 163-4; for ḥadīt̲h̲s, see Wensinck, Handbook, 109-10). Many misgivings against payment being made for religious services were held by certain authorities, who quoted in support of their view a saying of Abū Ḥanīfa (al-Muḳaddasī, 127).

2. The K̲h̲aṭīb or preacher [see k̲h̲atīb ].

3. The Ḳāṣṣ and Ḳāriʾ. On these, see above, I. C. 3. Sometimes, in later usage, wāʿiẓ is used of the official speaker, very like the k̲h̲aṭīb (cf. Ibn Baṭṭūṭa, iii, 9), while al-ḳāṣṣ is only applied to the street storyteller (al-Subkī, Muʿīd al-niʿam, 161-2). The ḳurrāʾ were also frequently appointed to madrasas and particularly to mausoleums (al-Maḳrīzī, iv, 223; Yāḳūt, iv, 509; al-Subkī, 162; van Berchem, CIA, i, no. 252).

4. The Muʾad̲h̲d̲h̲in. According to most traditions, the office of muʾad̲h̲d̲h̲in was instituted in the year 1, according to others only after the isrāʾ, in the year 2, according to some weak traditions, while Muḥammad was still in Mecca. At first, the people came to the ṣalāt without being summoned. Trumpets (būḳ) were blown and rattles (nāḳūs) used, or fires lit after the custom of Jews, Christians and Mad̲j̲ūs. ʿAbd Allāh b. Zayd learned the ad̲h̲ān formula in a dream; it was approved by the Prophet and when Bilāl proclaimed it, it was found that ʿUmar had also learned the same procedure in a dream (Ibn His̲h̲ām, 357-8; al-Diyārbakrī, i, 404-5; al-Buk̲h̲ārī, Ad̲h̲ān, bāb 1; al-Zurḳānī, i, 121 ff.). There are also variants of the story, e.g. that the Prophet and ʿUmar had the vision, or Abū Bakr or seven or fourteen of the Anṣār. According to some, the Prophet learned it at the miʿrād̲j̲ from Gabriel, hence the introduction of the ad̲h̲ān is dated after the isrāʾ, among the suggestions made, the hoisting of a flag is mentioned (Sīra Ḥalabiyya, ii, 100 ff.). Noteworthy is a tradition which goes back to Ibn Saʿd, according to which at ʿUmar’s suggestion, at first a munādī, Bilāl, was sent out who called in the streets: al-ṣalāta d̲j̲āmiʿatan. Only later were other possibilities discussed, but the method already in use was confirmed by the dream, only with another formula, the one later used al-Diyārbakrī, i, 404; Sīra Ḥalabiyya, ii, 100-1). According to this account, the consideration of other methods would be a secondary episode, and probably the tradition in general represents a later attitude to the practices of other religions. But in Islam, other methods were certainly used. In Fās, a flag was hung out in the minarets and a lamp at night (JA, ser. 11, xii, 341). The flag is also found in the legend of the origin of the practice.

The public crier was a well-known institution among the Arabs. Among the tribes and in the towns, important proclamations and invitations to general assemblies were made by criers. This crier was called munādī or muʾad̲h̲d̲h̲in (Sīra Ḥalabiyya, ii, 170; Lammens, La Mecque, 62 ff., 146; idem, Berceau, i, 229 n.; idem, Moʿâwia, 150). Ad̲h̲ān therefore means proclamation, sūra IX, 3, and ad̲h̲d̲h̲ana, muʾad̲h̲d̲h̲in, sūra VII, 70, “to proclaim” and “crier”. Munādī (al-Buk̲h̲ārī, Farḍ al-k̲h̲ums, bāb 15) and muʾad̲h̲d̲h̲in (ibid., Ṣawm, bāb 69; Ṣalāt, bāb 10 = D̲j̲izya, bāb 16; Sīra Ḥalabiyya, ii, 270) are names given to a crier used by the Prophet or Abū Bakr for such purposes. Official proclamations were regularly made by criers (cf. al-Ṭabarī, iii, 2131, 3). Sad̲j̲āḥ and Musaylima used a muʾad̲h̲d̲h̲in to summon the people to their prayers (al-Ṭabarī, i, 1919, 1932; cf. Annali dell’ Islām, i, 410; 638-9). It was therefore a very natural thing for Muḥammad to assemble the believers to common prayer through a crier (nādā li’l or ilā ’l-ṣalāt, sūra V, 63; lxii, 9); the summons is called nidāʾ and ad̲h̲d̲h̲ān, the crier munādī (al-Buk̲h̲ārī, Wuḍūʾ, bāb 5; Ad̲h̲ān, bāb 7) and mwʾad̲h̲d̲h̲in ; the two names are used quite indiscriminately (e.g. ibid., Wuḍūʾ, bāb 5; al-Ṭabarī, ii, 297 sq.). Munādī ’l-ṣalāt, al-Muḳaddasī, 182, 12, also ṣāʾiḥ “crier” is used (al-Ṭabarī, iii, 861; Chron. Mekka, i, 340).

In these conditions, it was very natural for the crier in the earliest period to be regarded as the assistant and servant of the ruler; he is his muʾad̲h̲d̲h̲in (Ibn Saʿd, i, 7; Muslim, Ṣalāt, tr. 4; al-Maḳrīzī, iv, 43, etc.; cf. al-Ṭabarī, ii, 1120). ʿUmar sent to Kūfa ʿAmmār b. Yāsir as amīr and ʿAbd Allāh b. Masʿūd “as muʾad̲h̲d̲h̲in and wazīr ” (Ibn al-Faḳīh, 165); he is thus the right hand of the ruler. Al-Ḥusayn had his munādī with him, and the latter summoned to the ṣalāt on al-Ḥusayn’s instructions (al-Ṭabarī, ii, 297, 298; cf. Ibn Ziyād, ibid, 260 and in the year 196/811-12, the ʿāmil in Mecca, ibid., iii, 861, 13; also. Chron. Mekka, i, 340). During the earliest period, the muʾad̲h̲d̲h̲in probably issued his summons in the streets and the call was very short: al-ṣalāta d̲j̲āmiʿat an (Ibn Saʿd, 7, 7; Chron. Mekka, i, 340; al-Ṭabarī, iii, 861; cf. also in the year 196/811-12, Sīra Ḥalabiyya, ii, 101 al-Diyārbakrī, i, 404-5). This brief summons was, according to Ibn Saʿd, also used later on irregular occasions (i, 7 ff.; cf. the passage in al-Ṭabarī). Perhaps also the summons was issued from a particular place even at a quite early date (see I. D. 2a). After the public summons, the muʾad̲h̲d̲h̲in went to the Prophet, greeted him and called him to prayer; the same procedure was later used with his successor; when he had come, the muʾad̲h̲d̲h̲in announced the beginning of the ṣalāt (aḳāma ’l-ṣalāt: cf. al-Buk̲h̲ārī. Wuḍūʾ, bāb 5; Ad̲h̲ān, bāb 48; Sīra Ḥalabiyya, ii, 104-5; al-Maḳrīzī, iv, 45; and iḳāma). The activity of the muʾad̲h̲d̲h̲in thus fell into three sections: the assembling of the community, the summoning of the imām and the announcement of the beginning of the ṣalāt. In the course of time, changes were made in all three stages.

The assembling of the community by crying aloud was not yet at all regular in the older period. During the civil strife in ʿIrāḳ, ʿUbayd Allāh b. Ziyād in the year 60/680 had his munādī summon people with threats to the evening ṣalāt in the mosque, and when after an hour the mosque was full, he had the iḳāma announced (al-Ṭabarī, ii, 260). When a large number of mosques had come into existence, the public call to prayer had to be organised lest confusion result, and the custom of calling from a raised position became general after the introduction of the minaret. While previously the call to prayer had only been preparatory and the iḳāma was the final summons, the public call (ad̲h̲ān) and the iḳāma now formed two distinct phases of the call to prayer. Tradition has retained a memory of the summoning in the streets, now completely fallen into disuse, when it tells us that ʿUt̲h̲mān introduced a third ad̲h̲ān, a call in al-Zawrāʾ, which was made before the call from the minaret: this call, however, was transferred by His̲h̲ām b. ʿAbd al-Malik to the minaret (al-Buk̲h̲ārī. D̲j̲umʿa. bābs 22, 25; Sīra Ḥalabiyya, ii, 110; Ibn al-Ḥād̲j̲d̲j̲, Madk̲h̲al, ii, 45). This may be evidence of the gradual cessation of the custom of summoning the community by going through the streets. Ibn Baṭṭūṭa (but this is exceptional) tells us that the muʾad̲h̲d̲h̲ins in K̲h̲wārazm still fetched the people from their houses and those who did not come were whipped (iii, 4-5), which recalls Wahhābī measures. When exactly the Sunnī and, in distinction to it, the S̲h̲īʿī formula, finally developed can hardly be ascertained [see ad̲h̲ān ]. The call ḥayya ʿalā ’ l-falāḥ is known from the time of ʿAbd al-Malik (65-85/685-705) (al-Ak̲h̲ṭal, ed. Ṣālḥānī, 254; see Horovitz, in Isl., xvi [1927], 154; on takbīr, see ibid.; on ad̲h̲ān formulae, see further Sīra Ḥalabiyya, ii, 105-6). At first, the call was only made at the chief mosque, as was the case in Medina and Miṣr (al-Maḳrīzī, iv, 43 below), but very quickly other mosques were also given muʾad̲h̲d̲h̲ins: their calls were sufficiently audible in the whole town. The chief mosque retained this privilege, that its muʾad̲h̲d̲h̲in called first and the others followed together (al-Maḳrīzī, iv, 43 below, 44).

The summoning by the imām in Medina was therefore quite a natural thing. The custom, at first associated with the ruler’s mosque, was not observed in Medina only (see for ʿUt̲h̲mān and ʿAlī, al-Ṭabarī, i, 3059-60), but was also usual under the Umayyads. The formula was al-salām ʿalayka ayyuhā ’l-amīr waraḥmatu ’llāh wa-barakātuhu, ḥayya ʿalā ’ l-ṣalat, ḥayya ʿalā ’l-falāḥ al-ṣalāt, yarḥamuka ’llāh (al-Maḳrīzī, iv, 45; Sīra Ḥalabiyya, ii, 105). After the alteration in the ad̲h̲ān and the greater distance of the ruler from the mosque, to summon him was no longer the natural conclusion to the assembling of the community. In the ʿAbbāsid period and under the Fāṭimids, there was a survival of the old custom, in as much as the muʾad̲h̲d̲h̲ins ended the ad̲h̲ān call before the ṣalāt al-fad̲j̲r on the minarets with a salām upon the caliph. This part of the muʾad̲h̲d̲h̲in’s work was thus associated with the first ad̲h̲ān call. When Ṣalāḥ al-Dīn came to power, he did not wish to be mentioned in the call to prayer, but instead he ordered a blessing upon the Prophet to be uttered before the ad̲h̲ān to the ṣalāt al-fad̲j̲r, which after 761/1360 only took place before the Friday service. A muḥtasib ordered that after 791/1389 in Egypt and Syria at each ad̲h̲ān a salām was to be uttered over the Prophet (al-Maḳrīzī, iv, 46; Sīra Ḥalabiyya, ii, 110). Ibn D̲j̲ubayr relates that in Mecca after each ṣalāt al-mag̲h̲rib, the foremost muʾad̲h̲d̲h̲in pronounced a duʿāʾ upon the ʿAbbāsid Imām and on Ṣalāḥ al-Dīn from the Zamzam roof, in which those present joined with enthusiasm (103), and according to al-Maḳrīzī, after each ṣalāt prayers for the sultan were uttered by the muʾad̲h̲d̲h̲ins (iv, 53-4). Another relic of the old custom was that the trumpet was sounded at the door of the ruler at times of prayer; this honour was also shown to ʿAḍud al-Dawla in 368/978-9 by order of the caliph (Miskawayh, vi, 499; ed. Cairo 1315, 396).

The iḳāma always remained the real prelude to the service and is therefore regarded as the original ad̲h̲ān (al-Buk̲h̲ārī. D̲j̲umʿa, bāb 24). In the earliest period, it was fixed by the arrival of the ruler and it might happen that a considerable interval elapsed between the summoning of the people and the iḳāma (cf. al-Ṭabarī, ii, 260, 297-8). The times were later more accurately defined; one should be able to perform one to three ṣalāts between the two calls (al-Buk̲h̲ārī, Ad̲h̲ān, bāb 14, 16). Some are said to have introduced the practice of the muʾad̲h̲d̲h̲in calling ḥayya ʿalā ’ l-ṣalāt at the door of the mosque between the two calls (Sīra Ḥalabiyya, ii, 105). From the nature of the case, the iḳāma was always called in the mosque; at the Friday service, it was done when the imām mounted the minbar (al-Buk̲h̲ārī, D̲j̲umʿa, bāb 22, 25; Sīra Ḥalabiyya, ii, 110; al-Maḳrīzī, iv, 43) while the muʾad̲h̲d̲h̲in stood in front of him. This muʾad̲h̲d̲h̲in. according to some, ought to be the one who called the ad̲h̲ān upon the minaret (Sīra Ḥalabiyya, ii, 109), while Ibn al-Ḥād̲j̲d̲j̲ ignoring the historical facts only permits the call from the minaret (Madk̲h̲al, ii, 45). In Tunis, the iḳāma was announced by ringing a bell as in the churches (al-Zarkas̲h̲ī, tr. Fagnan, in Rec. Soc. Arch. Constantine [1894], 111-12). A similarity to the responses in the Christian service is found in the fact that the call of the muʾad̲h̲d̲h̲in. which contains a confession of faith, is to be repeated or at least answered by every one who hears it (al-Buk̲h̲ārī, D̲j̲umʿa, bāb 23); this is an action which confers religious merit (Ibn Ḳuṭlūbug̲h̲ā, Ṭabaḳāt al-Ḥanafiyya, ed. Flügel, 30). It is possible that we should recognise in this as well as in the development of the formulae the influence of Christians converted to Islam (cf. Becker, Zur Gesch. d. islam. Kultus, in Isl., iii [1912], 374 ff., and Islamstudien, i, 472 ff., who sees an imitation of the Christian custom in the iḳāma in general; on the possibility of Jewish influence, see Mittwoch, in Abh. Pr. A. W. [1913], Phil.-Hist. Cl. 2).

The muʾad̲h̲d̲h̲in thus obtained a new importance. His work was not only to summon the people to divine service, but was in itself a kind of religious service. His sphere of activity was further developed. In Egypt we are told that Maslama b. Muk̲h̲allad (47-62/667-82) introduced the tasbīḥ. This consisted in praises of God which were uttered by the muʾad̲h̲d̲h̲ins all through the night until fad̲j̲r. This is explained as a polemical imitation of the Christians, for the governor was troubled by the use of the nawāḳīs at night and forbade them during the ad̲h̲ān (al-Maḳrīzī, iv, 48). In the time of Aḥmad b. Ṭūlūn and K̲h̲umārawayh. the muʾad̲h̲d̲h̲ins recited religious texts throughout the night in a special room. Ṣalāḥ al-Dīn ordered them to recite an ʿaḳīda in the night ad̲h̲ān and after 700/1300-1, d̲h̲ikr was performed on Friday morning on the minarets (ibid., 48-9, Sīra Ḥalabiyya, ii, 111). In Mecca also, the muʾad̲h̲d̲h̲ins performed d̲h̲ikr throughout the night of 1 S̲h̲awwāl on the roof of the ḳubba, of the Zamzam well (Ibn D̲j̲ubayr. 155, 156; cf. for Damascus, al-Maḳrīzī, iv, 49). Similar litanies are kept up in modern times, as well as a special call about an hour before dawn (ebed, tarḥīm: see Lane, Manners and customs, Everyman’s Library, 75-6, cf. 86; Snouck Hurgronje, Mekka ii, 84 ff.).

The original call of the muʾad̲h̲d̲h̲in thus developed into a melodious chant like the recitation of the Ḳurʾān. Al-Muḳaddasī tells us that in the 4th/10th century in Egypt during the last third of the night, the ad̲h̲ān was recited like a dirge (205). The solemn effect was increased by the large number of voices. In large mosques, like that of Mecca, the chief muʾad̲h̲d̲h̲in called first from a minaret, then the others came in turn (Chron. Mekka, iii, 242-5); Ibn D̲j̲ubayr, 145 ff.; (cf. Ibn Rusta, 111, 1 ff. and above). But in the mosque itself, the iḳāma was pronounced by the muʾad̲h̲d̲h̲ins in chorus on the dakka (see above, I. D. 2e) erected for this purpose, which is also traced to Maslama. In the 3rd/9th and 4th/10th centuries we hear of these melodious recitations (taṭrīb) of the muʾad̲h̲d̲h̲ins on a raised podium in widely separated parts of the Muslim world (Ṣanʿāʾ, Egypt, K̲h̲urāsān, al-Muḳaddasī, 327; Ibn Rusta, 111; the expression al-mutalaʿʿibīn, “the musicians”, if correct, probably refers to the muʾad̲h̲d̲h̲ins, al-Muḳaddasī, 205; cf. also al-Kindī, Wulāt, 469; for Fārs we are expressly told that the muʾad̲h̲d̲h̲ins call without taṭrīb, al-Muḳaddasī, 439, 17). Sometimes in large mosques, they were stationed in different parts of the mosque to make the imām’s words clear to the community (tablīg̲h̲). The singing, especially in chorus, like the tablīg̲h̲, was regarded by many as bidʿa (al-Kindī, op. cit.; Madk̲h̲al, ii, 45-6, 61-2; Sīra Ḥalabiyya, ii, 111). In other ways also, the muʾad̲h̲d̲h̲ins could be compared to deacons at the service. The k̲h̲aṭīb on his progress to the minbar in Mecca was accompanied by muʾad̲h̲d̲h̲ins, and the chief muʾad̲h̲d̲h̲in girded him with a sword on the minbar (Ibn D̲j̲ubayr, 96-7).

The new demands made on the muʾad̲h̲d̲h̲ins necessitated an increase in their number, especially in the large mosques. The Prophet in Medina had two muʾad̲h̲d̲h̲ins, Bilāl b. Ribāḥ, Abū Bakr’s mawlā, and Ibn Umm Maktūm, who worked in rotation. ʿUt̲h̲mān also is said occasionally to have called the ad̲h̲ān in front of the minbar, i.e. the iḳāma (al-Maḳrīzī, iv, 43). It is therefore regarded as commendable to have two muʾad̲h̲d̲h̲ins at a mosque (Muslim, Ṣalāt, tr. 4; cf. al-Subkī, Muʿīd, 165). Abū Maḥd̲h̲ūra was also the Prophet’s muʾad̲h̲d̲h̲in in Mecca. Under ʿUmar, Bilāl’s successor as muʾad̲h̲d̲h̲in was Saʿd al-Ḳaraẓ, who is said to have called to prayer for the Prophet in Ḳubāʾ (al-Maḳrīzī, op. cit.; cf. Sīra Ḥalabiyya, ii, 107 ff.). In Egypt under ʿAmr, the first muʾad̲h̲d̲h̲in in al-Fusṭāṭ was Abū Muslim; he was soon joined by nine others. The muʾad̲h̲d̲h̲ins of the different mosques formed an organisation, the head (ʿarīf) of which, after Abū Muslim, was his brother S̲h̲uraḥbīl b. ʿĀmir (d. 65/684-5); during his time, Maslama b. Muk̲h̲allad built minarets (al-Maḳrīzī, iv, 44).

The office of muʾad̲h̲d̲h̲in was sometimes hereditary. The descendants of Bilāl were for example muʾad̲h̲d̲h̲ins of the Medina Mosque in al-Rawḍa (Ibn D̲j̲ubayr. 194). We also find in Medina the sons of Saʿd al-Ḳaraẓ officiating (Ibn Ḳutayba, Maʿārif, ed. Wüstenfeld, 132, 279), in Mecca, the sons of Abū Maḥd̲h̲ūra (ibid., 278; Sīra Ḥalabiyya, ii, 106), in Baṣra, the sons of al-Mund̲h̲ir b. Ḥassān al-ʿAbdī, muʾad̲h̲d̲h̲ins of ʿUbayd Allāh b. Ziyād (Ibn Ḳutayba, 279); it is, however, possible that this was really the result of a system of guilds of muʾad̲h̲d̲h̲ins. In the d̲j̲awāmiʿ of the Mag̲h̲rib in the 8th/14th century, each had regularly four muʾad̲h̲d̲h̲ins who were stationed in different parts of the mosque during the ṣalāt (Madk̲h̲al, ii, 47 above); but there were often quite a large number. In the Azhar mosque in the time of al-Ḥākim, there were fifteen, each of whom was paid two dīnārs a month (al-Maḳrīzī, iv, 51). Ibn Baṭṭūṭa found seventy muʾad̲h̲d̲h̲ins in the Mosque of the Umayyads (i, 204). About 1900, in Medina there were in the Mosque of the Prophet fifty muʾad̲h̲d̲h̲ins and twenty-six assistants (al-Batanūnī, Riḥla, 242). Blind men were often chosen for this office; Ibn Umm Maktūm, for example, was blind (al-Buk̲h̲ārī. Ad̲h̲ān, bāb 11; Sīra Ḥalabiyya, ii, 104; cf. Lane, op. cit., 75). The Prophet is said to have forbidden T̲h̲aḳīf to pay a muʾad̲h̲d̲h̲in (al-Wāḳidī-Wellhausen, 383). ʿUt̲h̲mān is said to have been the first to give payment to the muʾad̲h̲d̲h̲ins (al-Maḳrīzī, iv, 44) and Aḥmad b. Ṭūlūn gave them large sums (ibid., 48). They regularly received their share in the endowments, often by special provisions in the documents establishing the foundations.

The muʾd̲h̲d̲h̲ins were organised under chiefs (ruʾasāʾ : al-Maḳrīzī, iv, 14). In Mecca, the raʾīs al-muʾad̲h̲d̲h̲inīn was identical with the muʾad̲h̲d̲h̲in al-Zamzamī who had charge of the singing in the upper story of the Zamzam building (Chron. Mekka, iii, 424-5; Ibn D̲j̲ubayr. 145; cf. Snouck Hurgronje, Mekka, ii, 322). The raʾīs was next to the imām but subordinate to him; in certain districts, it was the custom for him to mount the pulpit during the sermon with the imām (when the latter acted as k̲h̲aṭīb) (Madk̲h̲al, ii, 74). The position which they originally occupied can still be seen from the part which they play in public processions of officials, e.g. of the Ḳāḍī ’ l-Ḳuḍāt, when they walk in front and laud the ruler and his vizier (al-Maḳrīzī, ii, 246).

Closely associated with the muʾad̲h̲d̲h̲in is the muwaḳḳit, the astronomer, whose task it was to ascertain the ḳibla and the times of prayer (al-Subkī, Muʿīd, 165-6 and see mīḳāt); sometimes the chief muʾad̲h̲d̲h̲in did this (Snouck Hurgronje, Mekka, ii, 322).

5. Servants. According to Abū Hurayra, the Mosque of the Prophet was swept by a negro (al-Buk̲h̲ārī, Ṣalāt, bāb 72, cf. 74). The larger mosques gradually acquired a large staff of servants (k̲h̲uddām), notably bawwāb, farrās̲h̲, and water-carriers (cf. e.g. van Berchem, CIA, i, 252). In Mecca there have always been special appointments, such as supervisor of Zamzam and guardian of the Kaʿba (sādin, pl. sadana, also used of the officials of the mosque: al-Maḳrīzī, iv, 76; cf. Ibn D̲j̲ubayr, 278). In Ibn Baṭṭūṭa’s time, the servants (k̲h̲uddām) of the Mosque of the Prophet were eunuchs, particularly Abyssinian; their chief (s̲h̲ayk̲h̲ al-k̲h̲uddām) was like a great amīr and was paid by the Egyptian-Syrian government (i, 278, 348); cf. the title of an amīr of the year 798/1395-6, s̲h̲ayk̲h̲ mas̲h̲ā ʾ ik̲h̲ al-sāda al-k̲h̲uddām bi ’ l-ḥaram al-s̲h̲arīf al-nabawī (van Berchem, CIA, i, no. 201). In the Mosque of Jerusalem in about 300/912-13, there were no less than 140 servants (k̲h̲ādim Ibn al-Faḳīh, 100); others give the figure 230 (Le Strange, Palestine, 163) and according to Mud̲j̲īr al-Dīn, ʿAbd al-Malik appointed a guard of 300 black slaves here, while the actual menial work was done by certain Jewish and Christian families (Sauvaire, Hist. Jér. et Hébr., 56-7).

In other mosques, superintendents (ḳayyim, pl. ḳawama) are mentioned, a vague title which covered a multitude of duties: thus the Madrasa al-Mad̲j̲diyya had a ḳayyim who looked after the cleaning, the staff, the lighting and water-supply (al-Maḳrīzī, iv, 251), the Azhar Mosque had one for the miʾḍaʾa, who was paid twelve dīnārs, (ibid., 51) and also 4 ḳawama, who were paid like muʾad̲h̲d̲h̲ins (two dīnārs a month) and are mentioned between them and the imāms, probably supervisors of the staff (ibid., 51). In other cases, a ḳayyim al-d̲j̲āmiʿ, sometimes a ḳāḍī, is mentioned, who is apparently the same as the imām, the k̲h̲aṭīb or some similar individual of standing (ibid., 75, 121, cf. 122; cf. Ibn D̲j̲ubayr, 51). A mus̲h̲rif, inspector, is also mentioned, e.g. in the Azhar (al-Maḳrīzī, iv, 51).

Bibliography

given in the article.

(J. Pedersen)

H. The architecture of the mosque.

1. Introduction. Attemps to generalise about regional variations in mosque architecture are fraught with difficulty and have often miscarried. One solution, admittedly a compromise, is to select a few of the most celebrated mosques, to imply in more or less arbitrary fashion that they are typical, and to base the requisite generalisations on them. This approach has at least the merit of clarity, and it could indeed be argued that it is in the finest mosques of a given period and region that local peculiarities are apt to find their fullest expression. Nevertheless, such a broad-brush approach, for all its superficial attractions, is simply not specific enough. Another approach, which might be termed typological, cuts across regional and temporal boundaries in order to isolate the significant variants of mosque design and trace their development. Yet, precisely because it ignores such boundaries, this approach tends to minimise the significance of regional schools and fashions. The categories and sub-species which it proposes tend to have a somewhat academic flavour; while technically defensible, they somehow miss the point. A third approach might be to rely on statistics and, by chronicling all known mosques of pre-modern date, to discover the types and distribution of the most popular varieties. The picture to emerge from such a study might indeed be literally accurate, but it would not distinguish between the d̲j̲amiʿ and the masd̲j̲id, that is, between the major religious building of a town or city and the neighbourhood mosque (on the d̲j̲āmiʿ and its functions, see above, I. C. 2.). Since virtually all the mosques under discussion here fall into the category of d̲j̲āmiʿ, such a study would be of limited value in this context, and would assuredly blur the sharp outlines of regional peculiarities of mosque design. After all, the simplest types of mosques not only vastly outnumber the more complex ones but are also to be found throughout the Muslim world. It is such mosques, therefore, which make up the standard distribution of this building type. They dominate by sheer weight of numbers, but—by the same token—they distort the overall picture, suggesting a uniformity that actually exists only at the level of the most primitive buildings. Only when a statistical survey of this kind is relieved of the effectively dead weight of such buildings can regional and temporal distinctions stand out in their full clarity.

Such are the difficulties attendant on venturing a tour d’horizon of formal developments in the pre-modern mosque. What, then, is the best way of tackling this problem? The most promising line of approach is probably to identify those mosque types which are most distinctive of a given area and period, describing their constituent features but avoiding a detailed analysis of individual buildings. It should be emphasised that the over-riding aim of highlighting significant regional developments entails the suppression of much corroborative detail and, more importantly, of those periods when a given region was simply continuing to build mosques in a style already well established. Admittedly the lulls in innovation have their own part to play in the history of mosque architecture; but that part is too modest to rate any extended discussion here.

For that same reason, areas in which the pace of change was sluggish are allotted less attention in the following account than those which were consistently in the forefront of experiment. The Mag̲h̲rib, for example, receives less space than Iran, while ʿIrāḳ and the Levant take second place to Egypt and Anatolia. These emphases, moreover, reflect the basic truth that the design of a mosque was often less liable to take on a distinctively local colouring than were its decoration, its structural techniques or even specific components of that design, such as the minaret [see manāra ]. The time-span covered by this article is also limited. The mosque architecture of the last two centuries, which have seen the gradual invasion of a long-established Islamic idiom by European ideas and motifs, and in which a general decline is unmistakable, is omitted from this account. One final caveat should be sounded: the ensuing generalisations deliberately exclude the peripheral areas of the Islamic world, notably Indonesia, Malaysia, China and sub-Saharan Africa, for which see sections III-VII below. Nearly all the mosques in these areas are of post-mediaeval date, and therefore lie in the shadow of developments in the Islamic heartlands. There is, moreover, a strong vernacular element in these regional traditions, for they draw very heavily on a reservoir of ideas, practices and forms which owe very little to Islam. Thus for reasons which are as much historical and cultural as geographical they do not belong in the mainstream of mosque architecture.

This survey, then, will cover the central Islamic lands from al-Andalus to Afg̲h̲ānistān. The very nature of the material, however, makes it undesirable to embark directly on a series of regional summaries: the sheer lack of surviving monuments would require each summary to start at a different date. In most areas of the Islamic world it is not until the 5th/11th century that mosques survive in sufficient quantities for the lineaments of a local style to emerge. To explain that style would in most cases entail reference to earlier mosques in other regions, with consequent repetition and overlap. The crucial decisions which dictated the subsequent formal development of the mosque were taken in the early centuries of Islam; and the buildings which embodied those decisions are themselves thinly scattered over the entire area bounded by al-Andalus and Afg̲h̲ānistān. Yet the interconnections between these buildings are such as to make light of their geographical remoteness from each other.

Accordingly, a pan-Islamic survey of the early architectural history of the mosque will preface the individual accounts of local developments. These accounts in turn will be of unequal length. Pride of place will go to the Arab mosque plan, which not only had the widest diffusion but also covers the longest chronological span. Next in length will be the survey of the Persian tradition, almost as ancient as that of the Arab plan but more restricted in geographical scope. Shortest of all will be the discussion of the Turkish mosque type, whose creative development is confined in time to the 8th-11th/14th-17th centuries and in space to Anatolia.

2. Early history of the mosque

622-1000 A.D.

(a) The house of the Prophet. Beyond doubt, the genesis of the mosque is to be sought in a single seminal building: the house of the Prophet, erected to Muḥammad’s own specifications in Medina in 1/622. It was a near-square enclosure of some 56 × 53 m. with a single entrance; a double range of palm-trunk columns thatched with palm leaves (a feature of many African mosques to this day) was added on the ḳibla side, with a lean-to for destitute Companions to the south-east and nine huts for Muḥammad and his wives along the western perimeter. By a curious paradox, it was not built even secondarily as a mosque. This fact cannot be over-emphasised, since to ignore it is to misinterpret the subsequent history of mosque architecture. The venerated model for all later mosques itself became a mosque only, as it were, by the way and in the course of time. How is this to be explained? The accumulated deposit of many centuries of reverence makes it difficult to disinter the full original context of the building. Yet this much is clear: it was first and foremost a house for Muḥammad and his family to live in. It was also conceived from the beginning as a gathering place for the growing band of Muslims: in fact a kind of community centre, complete with the attendant associations of welfare. At the same time it served political, military and legal functions, while its high walls and single entrance allowed it at need to act as a place of refuge for the community. To be sure, by degrees people began to pray in it; but they prayed in many other places too and there is no evidence that it was used as the regular place of worship in the earliest years of the community. The mere fact that dogs and camels were allowed free access to it effectively disposes of such a notion. In short, Muḥammad had, it seems, no intention of creating a new type of building here. It is in no sense radical. In its extreme simplicity and austerity it well reflects his own life-style at that time. Its substantial scale may seem to contradict this, but is in fact somewhat deceptive, for some 80% of the interior consists of a vast empty courtyard. Yet it was this very emptiness that gave the mosque its innate flexibility, and in subsequent centuries a large open space became a standard feature of most large mosques. It is surely à propos to note that the earliest Christian places of worship, the so-called tituli, were also ordinary houses. (For a detailed discussion of the Prophet’s masd̲j̲id and its various functions, see above, I. A. 1.).

(b) The so-called “Arab plan”. Although there was thus a large measure of accident in the adoption of Muḥammad’s house as the model par excellence of later mosques, that form could not have enjoyed the popularity it did unless it had answered to a nicety the needs of Muslim liturgy and prayer. Its components—an enclosed square or rectangular space with a courtyard and a covered area for prayer on the ḳibla side—could be varied at will so as to transform the aspect of the building. Thus there evolved the so-called “Arab” or “hypostyle” mosque plan. From the first it showed itself capable of quite radical modification according to circumstances. At Kūfa in 17/638 the location of the mosque within one of the garrison cities (amṣār) allowed the builders to dispense with the element of security, and the perimeter—its dimensions fixed, according to al-Balād̲h̲urī. by four bowshots—is marked by ditches; elsewhere, as at Baṣra in the year 14/635, a reed fence served the same purpose. At Fusṭāṭ in the rebuilt mosque of ʿAmr (53/673), corner turrets served simultaneously to articulate the exterior, to single out the mosque from afar and to provide a place from which the call to prayer could be made: the germ of the future minaret. Multiple entrances became a feature as early as the first mosque of ʿAmr at Fusṭāṭ (22/643), admitting light to the muṣallā [q.v., and also above, I. B. 6] and allowing maximum ease of circulation.

The sunny climate of the southern Mediterranean and the Near East allowed the courtyard to accommodate the huge numbers of extra worshippers attending the Friday service. This was when its large expanse justified itself. For the rest of the week it was largely empty, and the heat and light emitted by this expanse could cause discomfort. This was especially likely if there were no provision for shade on three of the four sides, as in the early versions of the Great Mosques of Cordova (170/787), Ḳayrawān (221/836) and Tunis (250/864). Hence there arose the practice of adding arcades along the three subsidiary sides, so that people could walk around the mosque in cool shade. In time these arcades could be doubled, tripled or even quadrupled. A change in the alignment of their vaulting from one side of the mosque to another brought welcome visual relief and excluded the danger of monotony; so too did variations in the depth or number of the arcades (the second ʿAmr mosque in Cairo). As the surface area of the covered sanctuary was increased so did new spatial refinements suggest themselves, such as the progressive unfolding of seemingly endless vistas in all direction. Rows of supports (often spolia) with fixed intercolumniations created hundreds of repetitive modular units, perhaps deliberately mirroring the long files of worshippers at prayer.

Externally, the accent was on simplicity, with regular buttresses giving the structure a warlike air. At the Great Mosque of Sāmarrā (completed 238/852) there are a dozen of these on each long side, not counting the corners, with doorways after every second buttress. At Susa the exterior dispenses with buttresses in favour of rounded corner bastions, while in the mosque of al-Ḥākim in Cairo (381/991 onwards) the minarets at the corner of the façade rise from two gigantic square salients. The emplacement of the miḥrāb [q.v.] was marked by a corresponding rectangular projection on the exterior wall. Entrances were commonly allotted a measure of extra decoration—as in the series of shallow porches along the flank of the Cordova mosque—but massive portals on the scale of those in Western cathedrals found no favour in the early mosques of Arab plan. The absolute scale of some mosques (the mosque of Sāmarrā, for instance, could have accommodated 100,000 people) encouraged the adoption of fixed proportional ratios such as 3:2, which contributed in large measure to the impression of satisfying harmony which these mosques produced. The Ḳarak̲h̲ānid mosque of Samarḳand (5th/11th century) illustrates the continuing use of such ratios. Sometimes the scale of the mosque was illusionistically increased by the addition of a broad open enclosure (ziyāda) on three of the four sides (Mosque of Ibn Ṭūlūn, Cairo, finished 264/878, presumably copying the mosques of Sāmarrā). In comparison with later mosques of similar scale, which catered for multiple subsidiary functions by adding appropriate purpose-built structures to the central core, these early mosques maintain simple and symmetrical lines, especially for their outer walls.

The architectural vocabulary of these early mosques brought further scope for diversity. In the first half-century of Islamic architecture, the system of roofing was still primitive, and even when columns and roof-beams had replaced palm-trunks and thatching, the basic scheme remained trabeate (Baṣra; Kūfa; and Wāsiṭ, 83/702) whether the roof was flat or pitched. Thus the post-and-lintel system long familiar from Graeco-Roman buildings was perpetuated, and the pervasive classical flavour was strengthened by the lavish use of spolia. Sometimes, however, as in the bull-headed capitals of the Iṣṭak̲h̲r mosque, these were of Achaemenid origin.

By degrees, wooden roofs resting on arcades gained popularity, and this was the prelude to full-scale vaulting in durable materials (especially in Iran: Tārīk̲h̲āna mosque, Dāmg̲h̲ān, and Fahrad̲j̲ d̲j̲āmiʿ, both perhaps 3rd/9th century; Nāʾīn d̲j̲āmiʿ, perhaps 4th/10th century). The earliest mosques all use columns, and were thereby restricted to relatively low roofs. By the 3rd/9th century the pier had ousted the column as the principal bearing member, though it occurs as early as the mosques of Damascus, Baʿlabakk and Ḥarrān, and though the column was still used for some mosques (Ḳayrawān; al-Azhar, Cairo, 362/973). This change made it possible to raise the height of the roof, an important development given the oppressive sensation produced by a low roof extending over a large surface area. At the Cordova mosque the column shafts bore piers braced by strainer arches; but this device, for all its ingenuity, could not rival the popularity of superposed arcades in the fashion of Roman aqueducts (Damascus mosque, finished 98/716).

The apparently minor detail of whether the arcades ran parallel to the ḳibla or at right angles to it was sufficient to transform the visual impact of the roof. In the latter case, it focused attention on the ḳibla, and this was the solution that recommended itself to Mag̲h̲ribī architects (mosques of Cordova, Tunis and Ḳayrawān). Syrian architects, on the other hand, with only one major exception (Aḳṣā mosque, Jerusalem), preferred arcades parallel to the ḳibla (Damascus; Ḳaṣr al-Ḥayr East, ca. 109/728; Baʿlabakk, ca. 6th/12th century; Ḥarrān, ca. 133/750; and Raḳḳa, ca. 3rd/9th century), possibly reflecting in this the influence of the Christian basilica ubiquitous in that region, and several Egyptian mosques followed suit, including those of Ibn Ṭūlūn, al-Azhar and al-Ḥākim. It was a natural development to build mosques with arcades running in both directions (Great Mosques of Sfax and Susa, both finished 236/850), but with these exceptions the early experiments with this idea are all on a relatively modest scale which betrays some uncertainty of purpose. They comprise a small group of 9-bayed mosques with a dome over each bay and no courtyard: a type represented in Toledo, Ḳayrawān, Cairo and Balk̲h̲ and dating mainly from the 4th/10th century. These buildings inaugurate the much more ambitious use of vaults in later mosques. No such solutions are to be found in the larger mosques built before the 5th/11th century. This early Islamic vaulting drew its ideas impartially from the Romano-Byzantine tradition and from Sasanian Iran, and quickly developed its own distinctive styles, in which the pointed vault soon dominated.

(c) The secular element in early mosque architecture. In some mosques, the desire to emphasise the covered sanctuary (muṣallā) was achieved simply by adding extra bays and thus increasing its depth. In other mosques, especially those with royal associations, the requisite emphasis was achieved by some striking visual accentuation of the muṣallā: a more elaborate façade, a higher and wider central aisle, a gable or a dome. Once this idea of glorifying the muṣallā had taken root it was enthusiastically exploited, for example by furnishing this area with several carefully placed domes (Cordova, al-Azhar). On occasion, indeed, the muṣallā—complete with such distinguishing features as wider central aisle, dome in front of the miḥrāb and transversely vaulted bays adjoining the ḳibla—could itself become the mosque, with no attached courtyard (al-Aḳṣā).

The effect of singling out the muṣallā by these various means is to emphasise that this area is more important than any other in the mosque. Since this latter notion runs counter to the widely-expressed belief that all parts of the mosque are equally sacred, and that gradations of sanctity within it run counter to the spirit of Islam, its origins are worth investigating. It should be stressed at the outset that these various articulating devices cannot all be explained as attempts to draw attention to the ḳibla. Some measure of emphasis for this purpose was certainly required. Hence, no doubt, the greater depth of arcades on that side and the provision of an elaborate façade for the muṣallā alone. Similarly, the use of a different alignment or type of vaulting for the bays immediately in front of the ḳibla would make sense as a means of signposting this crucial area. Yet the addition of a dome or gable, or both, along the central aisle of the muṣallā, and the greater width and height of that aisle, cannot be explained—as is so often the case—simply as a means of highlighting the miḥrāb. After all, the entire ḳibla wall served to mark the correct orientation for prayer, so that the miḥrāb was technically redundant. The relatively late appearance of the miḥrāb (no 1st/7th century mosque appears to have possessed one and it is described as an innovation introduced by al-Walīd I in his re-building of the Mosque of the Prophet in Medina in 84/703) further suggests that it was not devised to meet some liturgical imperative.

The evidence points rather to the desire to assert in as public a way as the dictates of religious architecture would permit, the importance of the ruler in religious ceremonies. It was the duty of the caliph or of his representative to lead his people in prayer and to pronounce the k̲h̲uṭba [q.v.]. The political overtones of the latter ritual, which proclaimed allegiance to the ruler in much the same spirit as the diptychs in the contemporary Byzantine liturgy, in large part explain the physical form of the minbar [q.v.] from which the k̲h̲uṭba was pronounced. Similarly, the miḥrāb, another latecomer to mosque architecture, can be interpreted in secular terms, most conveniently as a throne apse transposed into a religious setting. These royal connotations could only be intensified by the addition of a dome over the bay directly in front of the miḥrāb.

Underneath that same dome was the preferred location for the maḳṣūra (see for this, above, I. D. 2.b.), usually a square enclosure of wood or stone reserved for the ruler, and ensuring both his privacy and his physical safety. Each of these elements in the mosque—miḥrāb, minbar, maḳṣūra, dome—drew added power from the proximity of the others, and together they stamped a secular and princely significance on this particular area of the mosque. The earliest surviving mosque which illustrates this emphasis, the Great Mosque of Damascus, adds a further refinement: a high transverse gable with a pitched roof cuts across the lateral emphasis of the muṣallā and thus highlights not just the miḥrāb area but also the way to it. The extra height of the gable and the way it cleaves across the grain of the mosque underscore its pro-clamatory role. Sometimes, as in the d̲j̲āmiʿ s of Tunis and Ḳayrawān, another dome over the central archway of the muṣallā façade sufficed to create an axis focused on the miḥrāb. As at Damascus, this axis asserted itself both inside the muṣallā and—by virtue of its greater width and the consequent break in the even tenor of the roofing—externally, at roof level. In later mosques, such as al-Azhar and al-Ḥākim (which possibly derive in this from al-Aḳṣā) the notion of the external gable is toned down to a broad flat strip projecting only modestly above roof level; but internally, the emphasis on the broader central nave terminating in the dome over the miḥrāb remains unchanged. It seems likely that these articulating devices were intended to mark out a processional way, presumably the formal route by which the ruler approached the miḥrāb.

So much, then, for the various elements in mosque design for which princely associations have been proposed. Yet their mere enumeration does not tell the full story. For it is above all the occurrence of these features in mosques located next to the residence of the ruler that places their political associations beyond doubt. This close juxtaposition of the secular and the religious may well have had its roots in the Prophet’s house. Be that as it may, at Baṣra, Kūfa, Fusṭāṭ, Damascus, to name only a few very early examples, the principal mosque and the private residence of the ruler adjoined each other, and the viceroy Ziyād b. Abīhi [q.v.] said of this arrangement “it is not fitting that the Imām should pass through the people”—a sentiment, incidentally, not shared by many later Islamic rulers. The analogy with the palatine chapel in Byzantium and mediaeval Europe—at Constantinople and Ravenna, Aachen and Palermo—is striking. Perhaps the most public expression of the idea in the mediaeval Islamic world was in the Round City of Bag̲h̲dād, where the huge and largely empty space at the heart of the city held only two buildings: the palace and the mosque, next door to each other. It would be hard to find the concept of Caesaropapism expressed more explicitly, or on a more gargantuan scale, than this.

The local expression of the articulating features under discussion varied from one part of the Islamic world to another, but they had come to stay. Henceforth, the d̲j̲amiʿ of Arab plan only rarely returned to the simplicity of the 1st/7th century. Such, however, was the strength of the traditions formed at that time that the basic nature of the earliest mosques remained substantially unchanged. They were proof, for example, against immense increases in size and against a growing interest in embellishment by means of structural innovations and applied ornament. Even the conversion into mosques of pre-Islamic places of worship, as at Damascus and Ḥamā, was powerless to affect their essential nature. The component parts of the Arab mosque could be redistributed and rearranged almost at will without impairing their functional effectiveness.

In much the same way, their idiosyncrasies of structure and decoration were purely cosmetic. The range of options in these areas was gratifyingly wide. Windows and lunettes bore ajouré grilles in stone or plaster with geometric and vegetal designs (Damascus mosque); wooden ceilings were painted or carved and coffered (Ṣanʿāʾ mosque, 1st/7th century onwards); a wide range of capitals, at first loosely based on classical models but in time featuring designs of Central Asian origin (Sāmarrā) was developed; and piers with engaged corner colonnettes (Ibn Ṭūlūn mosque, Cairo) rang the changes on the traditional classical column. Finally, the aspect of these early mosques could be varied still further by the type of flooring employed—stamped earth, brick, stone or even marble flags—and by applied decoration in carved stone or stucco, fresco, painted glass, embossed metal work or mosaic.

3. Later history of the “Arab plan” mosque.

The essentially simple components of the Arab plan set a limit to the degree of diversity that could be achieved within these specifications. Most of the room for manoeuvre had been exhausted within the first four centuries of Islamic architecture. Thus the subsequent history of the Arab plan cannot match the early period for variety and boldness; the later mosques, moreover, lie very much in the shadow of their predecessors, to such an extent, indeed, that it is hard to single out significant new departures in these later buildings. It can scarcely be doubted that the presence of the great Umayyad and ʿAbbāsid mosques, built at the period when the Islamic world was at the peak of its material prosperity, acted as a signal deterrent to later architects with substantially less money, men and materials at their disposal. In these early centuries the caliphal permission, not readily granted, had been required for the construction of a d̲j̲āmiʿ making it therefore a major undertaking, and correspondingly hard to emulate. By the 5th/11th century, moreover, most of the major Muslim cities had their own d̲j̲āmiʿ, so that the need for huge mosques had much declined.

Although mosques of Arab plan have continued to be built throughout the Islamic world until the present day, in the mediaeval period there were only two areas where they achieved dominance: in the Western Islamic lands before they fell under Ottoman rule, and in pre-Ottoman Anatolia. These areas will therefore provide the material for most of the discussion which follows. Nevertheless, sporadic references will be made to mosques elsewhere, for instance in Egypt and the Yemen.

(a) The Mag̲h̲rib. The Mag̲h̲rib rightfully takes pride of place in this account because for almost a millennium virtually no mosque that was not of Arab type was built there. Here, then, is to be found the most homogeneous and consistent development of that type. Its sources lie, like so much of Mag̲h̲ribī art, in Syria, and specifically in the Great Mosque of Damascus. Its transverse gable becomes a leitmotif in Mag̲h̲ribī mosques, and in some cases (such as the Ḳarawiyyīn Mosque [q.v.], Fez, founded 226/841 but largely of the 6th/12th century) is associated with the same proportions as the Syrian building, including the relatively shallow oblong courtyard imposed on the Damascus mosque by the classical temenos but copied thereafter in other mosques as a deliberate feature. In the Mosque of the Andalusians at Fez (600-4/1203-7) the Damascus schema is retained despite a jaggedly irregular perimeter and trapezoidal courtyard; and, as at the Ḳarawiyyīn mosque, the main entrance to the mosque is aligned to it, a refinement not found at Damascus. The length of the gable has also increased considerably, though its height is modest.

In later Mag̲h̲ribī mosques especially, the emphasis shifted from the exterior elevation of the gable to its impact from within the building. It attracts unusually intricate vaulting, often of muḳarnas [q.v.] type, or may be marked by domes ranging in number from two (Tlemcen, 531/1136) to six (second Kutubiyya, Marrakesh, mid-6th/mid-12th century). The latter mosque has a further five cupolas placed three bays apart along the transverse ḳibla aisle. Thus by means of vaulting alone is created a T-shape which combines the secular and religious emphases of the d̲j̲āmiʿ. Fewer vaults or domes, more strategically placed—for example at the miḥrāb, the muṣallā entrance and the corners of the ḳibla wall—could suffice to carry the T-shape into the elevation, but the form could be created at ground level alone by means of a wider central nave and by ensuring that the vaults stopped one bay short of the ḳibla, thus opening up dramatically the space immediately in front of it. The T-shape can indeed claim to be the principal Mag̲h̲ribī contribution to the development of mosque form, though horseshoe arches and square minarets were equally characteristic of the style.

Three other features distinguish Mag̲h̲ribī mosques from those found elsewhere in the Islamic world, though all have their origins in al-Andalus: the use of pierced ribbed or fluted domes, especially over the miḥrāb; the manipulation of arch forms to create hierarchical distinctions by means of gradual enrichment; and a readiness to alter the size, shape and location of the courtyard in response to the imperatives of a specific design. The ribbed domes (e.g. d̲j̲āmiʿs of Taza, 537/1142 and 691/1292, and Algiers, ca. 490/1097) derive from those of the Cordova mosque, but elaborate on them by cramming them with vegetal designs in carved stucco or by increasing the number of ribs from the usual eight to twelve (Tlemcen d̲j̲āmiʿ) or even sixteen (Taza d̲j̲āmiʿ). This practice gives free rein to the characteristically Mag̲h̲ribī obsession with non-structural arched forms, here used as a lace-like infill between the ribs; the overall effect is one of feathery lightness and grace. The light filtered through these domes suffuses the area of the miḥrāb with radiance, perhaps as a deliberate metaphor of spiritual illumination, an idea rendered still more potent when, as is often the case, that miḥrāb bears the popular text of sūra XXIV, 36-7, “God is the Light of the heavens and the earth; the likeness of His Light is as a niche wherein is a lamp...”

Long files of arcaded columns stretching in multiple directions and generating apparently endless vistas are a particular feature of Mag̲h̲ribī mosques. The distinctive “forest space” thereby created finds its fullest expression in the fourth major rebuilding of the Cordova mosque, the supreme generative masterpiece of Western Islamic architecture, and the major Almoravid and Almohad mosques are best interpreted as reflections of this great original. Where the Cordova mosque, however, employed systems of intersecting arches and carefully differentiated types of capital to establish hierarchical distinctions, later Mag̲h̲ribī d̲j̲āmiʿ s typically use a wide range of arch profiles to the same end. These include, besides the ubiquitious horseshoe type already noted, lobed, multifoil, interlaced cusped, trefoil, lambrequin and other varieties. They spring from piers, not columns, and this, coupled with the low roof, dim lighting and the general absence of ornament unconnected with vaulting, lends these interiors a ponderous austerity. Against this general background of parsimonious simplicity, the sudden switch from plain arch profiles for most of the sanctuary to elaborate ones for the axial nave alone constitutes a dramatic enrichment of the interior. Sometimes the transverse aisle in front of the ḳibla wall attests a third type of arch profile, and thus a further gradation of importance is emphasised.

In most western Islamic mosques the courtyard is something of an appendage. It is almost always very much smaller than the covered space. Custom decreed that it was isolated at the opposite end of the mosque from the miḥrāb, and that it should either be contiguous to the outer wall or be separated from it by no more than a single aisle. By contrast, the sanctuary tended to be of disproportionate depth and extent. This meant that the courtyard was never able to function as the heart of the mosque. Only when the sanctuary was reduced, as in the Ḳaṣba mosque in Marrakesh (581-6/1185-90), with its pronounced cruciform emphasis, was the courtyard able, both literally and figuratively, to play a more central role. In narrow rectangular plans, it can be a diminutive square box hemmed in by deep lateral aisles (Mosque of al-Manṣūra, 704-45/1304-44) or an extended shallow oblong (Mosque of Seville, ca. 571/1175). In oblong plans, it faithfully mirrored that emphasis on a diminutive scale (Tinmal, 548/1153; first Kutubiyya, Marrakesh, ca. 555/1160). Exceptional on all counts is the gigantic but unfinished mosque of Ḥasan, Rabat (ca. 591/1195), whose scale of 180 × 139 m. makes it the second largest mosque in the world, after the Great Mosque of Sāmarrā. Here the typical shallow oblong courtyard is supplemented by two lesser and narrow courtyards perpendicular to the ḳibla and along the lateral walls. These were, it seems, intended for men and women respectively, but they would also have served for ventilation and lighting, besides offering visual relief to the endless march of columns.

(b) Anatolia. For all that pre-Ottoman Anatolia was a fertile field for innovation in later mediaeval experiment with the hypostyle mosque, its contribution cannot seriously match that of the Mag̲h̲rib and al-Andalus, not least because of the much shorter time span, a mere three centuries; discussion of it will accordingly be brief. The earliest surviving mosques well illustrate the dependence of local builders on more developed traditions of Arab and Persian origin. The Great Mosque of Diyārbakir (484/1091) follows the transept schema of Damascus, while those of Mayyāfariḳīn (550/1155), Dunaysir (601/1204) and Mārdīn (largely 6th/12th century) follow Iranian precedent in their emphasis on a monumental dome rearing up out of the low roofing of the sanctuary and set squarely in front of the miḥrāb bay. Their foreshortened courtyards, however, owe nothing to Iranian precedent and instead presage later developments. So too did the increasing tendency to use domical forms rather than modular trabeate units as the principal means of defining space.

The buildings of the 6th/12th and 7th/13th centuries sufficiently demonstrate the embryonic state of mosque design in Anatolia, for the variety of plans is bewildering and defies easy categorisation. The absence of direct copies of the classical Arab type of plan is striking, though modifications of it were legion. A common solution was to do without the courtyard altogether—perhaps a response to the severe Anatolian winter—and reduce the mosque to a wooden-roofed hall resting on a multitude of columns or pillars (ʿAlāʾ al-Dīn mosque, Konya, 530/1135 to 617/1220; Sivas, ca. 494/1101; Afyon, 672/1273; Beyşehir, 696/1296). Usually the minaret was outside the mosque and therefore not integrated into the layout. Sometimes a similar design was executed in multiple small vaults (Divriği, castle mosque, 576/1180; Niksar, 540/1145; Urfa, 6th/12th century), and indeed the preference for vaulted as distinct from trabeated construction is well marked even at this experimental stage. Whatever the roofing system adopted in these enclosed mosques, the scope for development in either direction was small, while poor lighting, a sense of cramped space and inadequate ventilation were virtually inevitable. Huge piers and low vaults gave many of these mosques a crypt-like appearance (ʿAlāʾ al-Dīn mosque, Niğde, 620/1223; Sivas, Ulu Cami).

The obvious way forward was to allot a more significant role to the dome, a decision made at an early stage (Great Mosque of Erzurum, 530/1135; Kayseri, 535/1140; and Divriği, 626/1229) but by no means universally accepted. In such mosques the domed bay is invariably the largest of all and is placed along the axis of the miḥrāb. This emphasis on the totally enclosed covered mosque was to remain the principal feature of Turkish mosque architecture, and as a natural corollary fostered a compact and integrated style. Sometimes a small courtyard is integrated into this design (Malatya, 635/1237; Kayseri, Mosque of K̲h̲wānd K̲h̲ātūn, 635/1237; Harput, 560/1165). By degrees, however, the courtyard was relegated to one of two functions: as a forecourt, akin to the atrium of Byzantine churches and thus heralding the mosque proper, instead of being co-equal to the sanctuary; and as a bay within the muṣallā, furnished with a skylight and a fountain as a symbolic reminder of the word outside. Sometimes these two uses coincided. The skylight bay (s̲h̲ādirwān) was normally placed along the axis of the miḥrāb and thus served as a secondary accent for it, in much the same manner as a central dome.

The 8th/14th century saw no major developments in hypostyle plans. Flat-roofed prayer halls, some with wooden-roofed porches (Merām mosque, Konya, 804-27/1402-24), others, especially in the Ḳaramān region, without them, continued to be built. So too did hypostyle mosques with vaulted domical bays (Yivli Minare mosque, Antalya, 775/1373; the type recurs both in eastern Anatolia and Ottoman territory in Bursa and Edirne). Variations in the Damascus schema, with the transept replaced by one or more domes, a raised and wider central aisle, a skylight bay, or any combination of these were frequent (ʿĪsā Bey mosque, Selcuk, 776/1374; Ulu Cami, Birgi, 712/1312; mosque of Ak̲h̲ī Elvān, Ankara, ca. 780/1378). Finally, mosques with an enlarged domed bay in front of the miḥrāb spread from their earlier base in south-eastern Anatolia, an area bounded to the east by the Ulu Cami in Van (791-803/1389-1400) and to the west by that of Manisa (778/1376). In the latter mosque the ḳibla side is dominated by the dome and takes up almost half the mosque; a large arcaded courtyard with a portico accounts for the rest. With such buildings the stage is set for Ottoman architecture and Arab prototypes are left far behind.

These Anatolian mosques depart still further from the norm of the hypostyle type in their predilection for elaborate integrated façades. While earlier mosques of Arab type frequently singled out the principal entrance by a monumental archway, often with a dome behind it, the tendency was to keep the façade relatively plain. Only in the highly built-up areas of the major cities of the Near East, such as Cairo, Jerusalem, Damascus and Aleppo, did the extreme shortage of space, and often the small scale of the mosques themselves, oblige architects to decorate mosque façades if they wished to draw attention to them, e.g. the Aḳmar mosque, Cairo, 519/1125. In Anatolia the tenacious Armenian tradition, which favoured extensive external sculpture and articulation, may well have predisposed Muslim architects in Anatolia to develop integrated decorative schemes for the main façades of their mosques. A monumental stone portal or pīs̲h̲tāḳ [q.v.], often an īwān [q.v.] was the standard centrepiece for such designs. It could be strongly salient and tower well above the roofline (Divriği Cami). Further articulation was provided by ranges of recessed arches with decorative surrounds (Dunaysir), open or blind arcades along the upper section of the façade (Mayyāfāriḳīn and ʿAlāʾ al-Din mosque, Konya), and windows with densely carved frames (ʿĪsā Bey mosque, Selcuk).

(c) Egypt and Syria. It seems possible that some of the more elaborate Mamlūk mosque façades in Cairo, such as those of Baybars (660/1262) and Suḷtān Ḥasan (757/1356) may derive, if at several removes, from Anatolian prototypes of the kind discussed above. It is noteworthy, however, that in general the mosques of the Ayyūbid and Mamlūk period offer little scope for large-scale reworking of the hypostyle plan, since they were too small. The mosque of Baybars and that of al-Nāṣir Muḥammad b. Ḳalawūn in the Cairo citadel (718/1318), which is a free copy of it, provide exceptions to this rule; in both cases a monumental dome over the miḥrāb bay is the principal accent of an extensive covered space. The relative scarcity of major mosques in this period not only reflects the primacy of the great early d̲j̲āmiʿs which were still in use, and which made further such buildings redundant; it also marks a shift in patronage away from mosques towards mausolea, madrasas, k̲h̲ānḳāhs and the like. In time, not surprisingly, joint foundations became the norm, in which the mosque was a mere oratory, a component in some larger complex. Eventually, too, the forms of mosques came to reflect those of contemporary madrasas more than the hypostyle plans of earlier periods. Hence the dominance of small domed mosques such as the 7th/14th century Mamlūk d̲j̲āmiʿ s of Tripoli. Such buildings have no bearing on the history of the Arab mosque plan.

(d) The Yemen. Apart from the Mag̲h̲rib, it was principally in the Yemen that the large hypostyle mosque maintained its popularity throughout the mediaeval period. Inadequate publication has meant that these buildings are less well known than they deserve, and without excavation the dating of many of them will remain problematic. This is particularly regrettable because several of them were built on the site of pre-Islamic temples, churches or synagogues (e.g. al-D̲j̲ilāʾ mosque, Ṣanʿāʾ), and spolia from these earlier buildings—such as columns, capitals, inscriptions and even sculptures of birds—are used very widely. Persistent local tradition attributes the d̲j̲āmiʿs of Ṣanʿāʾ and al-D̲j̲anad to the time of the Prophet; both were probably rebuilt by al-Walīd I. The former has preserved much more of its original appearance: perimeter walls of finely cut stone in stepped courses enclose a roughly square shape with a central courtyard with the muṣallā only slightly deeper than the other sides. Al-D̲j̲anad, on the other hand, has had its similar original layout transformed by a domed transept and numerous subsidiary buildings. This gradual transformation by the addition of prayer halls, mausolea, ablutions facilities and the like is a recurrent pattern in the Yemen (d̲j̲āmiʿs of Zabīd and Ibb).

Small hypostyle mosques of square form (al-ʿAbbās, 7th/13th century), or of rectangular shape, whether broad and shallow oblongs (Tit̲h̲id, 7th/13th century) or narrow and deep (Tamur, 5th/11th century or earlier), are common, and a few larger mosques of this kind, still without a courtyard, are known (D̲h̲ibin, after 648/1250). The commonest form, however, comprises a structure that is rectangular or trapezoidal (Masd̲j̲id al-Ṣawmaʿa, Hūt, 7th/13th century) with a central courtyard and extensive covered riwāḳs on all sides (Rawda d̲j̲āmiʿ, 7th/13th century). Often this formula is enriched by a lavishly carved or painted wooden ceiling over the sanctuary area alone (S̲h̲ibām d̲j̲āmiʿ, 4th/10th century) or by the incorporation of mausolea (Ẓafār D̲h̲ibin, 7th/13th century; funerary mosque of the Imām al-Hādī Yaḥyā, Ṣaʿda, 4th/10th century and later) or of minarets (D̲j̲ibla, 480/1087; D̲h̲ū As̲h̲raḳ, 410/1019). Influences from the central Islamic lands explain the use of wider central aisles in the muṣallā (Ẓafār D̲h̲ibin, Ibb, D̲j̲ibla, D̲h̲ū As̲h̲raḳ) and a concentration of domes along the ḳibla wall (enlargement of Ibb d̲j̲āmiʿ; D̲j̲āmiʿ al-Muẓaffar and As̲h̲rafiyya mosque, both 7th/13th century, Taʿizz). The glory of these Yemeni mosques as a group lies in their decoration: exceptionally long bands of stucco inscriptions (mosques of D̲h̲amār and Rada, 7th/13th century and later), frescoes with epigraphic, floral and geometric designs (Rasūlid mosques of Taʿizz) and a matchless series of carved and painted wooden ceilings (Ẓafār D̲h̲ibin. al-ʿAbbās, Sirha, D̲h̲ibin. S̲h̲ibām. Ṣanʿāʾ and others).

4. The Iranian tradition.

(a) The early period. Such was the prescriptive power of the “Arab plan” that its influence permeated mosque architecture in the non-Arab lands too. It would therefore be an artificial exercise to consider the development of the Iranian mosque in isolation, the more so as many early mosques in Iran (Bīs̲h̲āpūr, Sīrāf, Susa, Yazd) were of Arab plan. Some also had the square minarets which were an early feature of that plan (Dāmg̲h̲ān; Sīrāf). Rather did the Iranian mosque acquire its distinctive character by enriching the hypostyle form by two elements deeply rooted in pre-Islamic Iranian architecture: the domed chamber and the īwān, a vaulted open hall with a rectangular arched façade. The domed chamber derived from the mostly diminutive Sasanian fire temple with four axial arched openings, the so-called čahār tāḳ. Set in the midst of a large open space, it served to house the sacred fire. This layout obviously lent itself to Muslim prayer, and literary sources recount how such fire temples were taken over and converted into mosques (e.g. at Buk̲h̲ārā) by the simple expedient of blocking up the arch nearest the ḳibla and replacing it with a miḥrāb but conclusive archaeological evidence of this practice is still lacking, though the mosques of Yazd-i K̲h̲āst and Ḳurwa may be examples of it. Such domed chambers, whether converted fire temples or purpose-built Muslim structures, may have served as self-contained mosques, with or without an attached courtyard; certainly the earliest part of many mediaeval Iranian mosques is precisely the domed chamber.

The associations of the īwān, by contrast, were markedly more secular than religious; its honorific and ceremonial purpose in Sasanian palaces is epitomised by the great vault at Ctesiphon, where it announced the audience chamber of the Emperor. The īwān form was therefore well fitted to serve as a monumental entrance to the mosque, to mark the central entrance to the muṣallā (Tārīk̲h̲āna, Dāmg̲h̲ān; Nāʾīn) or, indeed, itself to serve as the sanctuary (as at Nīrīz perhaps 363/973 onwards?). Thus both the domed chamber and the īwān quickly found their way into the vocabulary of Iranian mosque architecture, and by their articulating power gave it a wider range of expression than the Arab mosque plan could command. It was in the interrelationships between the domed chamber, the īwān and the hypostyle hall that the future of the Iranian mosque was to lie.

(b) The Sald̲j̲ūḳ period. The tentative experiments of early Iranian mosque architecture crystallised in the Sald̲j̲ūḳ period, especially between ca. 473/1080 and ca. 555/1160. The major mosques built or enlarged at this time have as their major focus a monumental domed chamber enclosing the miḥrāb and preceded by a lofty īwān. This double unit is commonly flanked by arcaded and vaulted prayer halls. This arrangement represents the final transformation of the muṣallā in Iranian mosques, using the vocabulary of Sasanian religious and palatial architecture for new ends. The sanctuary īwān opens onto a courtyard with an īwān at the centre of each axis punctuating the regular sequence of riwāḳs. These arcades attain a new importance as façade architecture by their arrangement in double tiers. Yet the focus of attention is undoubtedly the great domed chamber. The simplicity of the prototypical čahār ṭāḳ is scarcely to be recognised in these massive Sald̲j̲ūḳ maḳṣūra domes with their multiple openings in the lower walls and their complex zones of transition. This concentration on the domed chamber was often achieved at the expense of the rest of the mosque (Gulpāyagān d̲j̲āmiʿ, ca. 510/1116). The new combination of old forms created the classical, definitive version of the already ancient 4-iwan courtyard plan that was to dominate Iranian architecture for centuries to come, infiltrating not only other building types such as madrasas and caravansarais, but also spreading as far west as Egypt and Anatolia and eastwards to Central Asia and India. The 4 -īwān mosque thus became in time the dominant mosque type of the eastern Islamic world.

Up to the end of the Sald̲j̲ūḳ period, however, the way was still open for numerous other combinations of hypostyle hall, domed chamber and īwān. Bas̲h̲an, for example (4th/10th century) has a square layout with courtyard, hypostyle hall, domed sanctuary and sanctuary īwān, but lacks any further articulation of the courtyard façade by īwāns. The mosques of Dandānḳān [q.v. in Suppl.] and Mas̲h̲had-i Miṣriyān [q.v.] (both 5th/11th century) are typologically related. At Urmiya/Riḍāʾiyya (7th/13th century) the mosque is an extensive shallow oblong with the domed chamber at one end of a hypostyle hall, and no īwān. Sometimes the mosque is entirely covered by five (Masd̲j̲id-i Diggaron, Hazāra, 5th/11th century) or nine domed bays (Čar Sutūn mosque, Tirmid̲h̲, 5th/11th century; Masd̲j̲id-i Kūča Mīr, Nātanz, 6th/12th century). In its Sald̲j̲ūḳ form the mosque at Ardabīl comprised a domed chamber with an īwān in front of it, while at Sīn (528/1136) the sanctuary, comprising a deep īwān with muḳarnas vaulting, engulfs one side of the diminutive courtyard. The huge courtyard of the Firdaws d̲j̲āmiʿ (597/1201) is dominated by its single īwān which heralds a low vaulted sanctuary. The d̲j̲āmiʿs of Faryumad (7th/13th century?) and Gūnābād (606/1210) have only two īwāns facing each other across a narrow courtyard, and no domed chamber. Other mosques in K̲h̲urāsān are simpler still, comprising only the domed chamber itself (Sangān-i Pāʾīn, 535/1140; Birrābād and ʿAbdallāhabād, both possibly Sald̲j̲ūḳ) or with insignificant bays adjoining it (Tak̲h̲latan Baba, 6th/12th century). Often too, the various elements were added in an unpredictable sequence, for instance at Simnān where a probably 5th/11th century columned hall had a complete mosque “unit” comprising a domed chamber, īwān and courtyard tacked on to its side. Even within the classical 4-īwān model, considerable diversity could be attained by varying the scale of the components: from long narrow courtyards (Simnān) or small square ones of domestic scale (Zawāra, 527/1133) to huge open expanses broken up by trees (S̲h̲īrāz d̲j̲āmiʿ, mainly 10th/16th century), pools or fountains.

The principal emphasis on the internal façade was, however, unchanging. The exterior, by contrast, was unadorned and unarticulated to the point of austerity. Variations in the height or breadth of īwāns reinforced axial or hierarchical distinctions. By common consent the sanctuary īwān was the largest and deepest; the opposite īwān was next in size, though often very shallow, while the two lateral īwāns were usually the smallest. Minarets at the corner of the sanctuary īwān underlined its importance, while the twin-minaret portal īwān first encountered in the Sald̲j̲ūḳ period (Nak̲h̲čivān. ca. 582/1186; Ardistān, Masd̲j̲id-i Imām Ḥasan, 553/1158) became increasingly monumental and elaborate in later centuries (d̲j̲āmiʿs of As̲h̲tard̲j̲ān, 715/1315, and Yazd, 846/1442). Īwān minarets of this kind gradually replaced the freestanding cylindrical minarets so popular in the Sald̲j̲ūḳ period.

(c) The Īlk̲h̲ānid period. As in Mamlūk Egypt, so too in Iran the later mediaeval history of the mosque is sometimes hard to disentangle from that of the madrasa -, tomb- or shrine-complex. Prayer and communal worship were, after all, integral to the operation of such “little cities of God” as the shrines of Ardabīl, Nātanz, Turbat-i D̲j̲ām, Basṭām and Lind̲j̲ān—all of them the scene of much building activity in the 8th/14th century—to say nothing of the great shrines of Ḳumm and Mas̲h̲had. Such new foundations as these were simply perpetuated Sald̲j̲ūḳ models (Hafs̲h̲ūya, early 8th/14th century), though these were subtly altered by having their proportions attenuated or otherwise modified. At As̲h̲tard̲j̲ān everything is subordinated to the principal axis announced by the double minaret façade, an emphasis which is taken up and intensified by the single great īwān which takes up the full width of the courtyard and leads into the domed sanctuary. At Warāmīn, too (722/1322 onwards), which is of standard 4-īwān type, the sense of axial progression is strong, and is made rather more effective than at As̲h̲tard̲j̲ān by the absolute length of the mosque and the extended vestibule. The d̲j̲āmiʿ of ʿAlī S̲h̲āh in Tabriz, by contrast (ca. 710-20/1310-20) deliberately returned, it seems, to much earlier models, for it comprised essentially a huge cliff-like īwān preceded by a courtyard with a central pool and clumps of trees in the corners—perhaps a deliberate reference to the Ṭāḳ-i Kisrā itself. For smaller mosques, Sald̲j̲ūḳ models were again at hand; hence, for example, the trio of domed chamber mosques with īwāns at Azirān, Kad̲j̲ and Das̲h̲tī, all datable ca. 725/1325. Yet another compliment to earlier masters was the Īlk̲h̲ānid tendency to add new structures to existing mosques: a madrasa to the Iṣfahān d̲j̲āmiʿ: (776-8/1374-7), an īwān to the mosque at Gaz (ca. 715/1315), and so on.

(d) The Tīmūrid period. The Tīmūrid period took up still further ideas which had been no more than latent in earlier centuries. While some mosques of traditional form were built such as the Mosque of Gawhar S̲h̲ād, in Mas̲h̲had, of standard 4 -īwān type (821/1418), attention focused particularly on the portal and ḳibla īwāns, which soared to new heights. Turrets at the corners magnified these proportions still further. This trend towards gigantism is exposed at its emptiest in the 4 -īwān d̲j̲āmiʿ of Ziyāratgāh, near Harāt (887/1482), where the absence of decoration accentuates the sheer mass of the sanctuary īwān looming over the courtyard. At its best, however, as in the mosque of Bībī K̲h̲ānum, Samarḳand (801/1399) where these exceptional proportions are consistently carried through to virtually every part of the mosque, the effect is overwhelming. Here the 4-īwān plan is transformed by the use of a domed chamber behind each lateral īwān; by the profusion of minarets—at the exterior corners and flanking both portal and sanctuary īwāns—and by the four hundred-odd domes which cover the individual bays.

As in the Mongol period, however, the fashion for building k̲h̲ānḳāhs, madrasas and funerary monuments, all of them capable of serving as places of worship (shrine of Aḥmad Yasawī, Turkestan, begun 797/1394; the Rīgistān complex, Samarḳand, begun in its Tīmūrid form in 820/1417; Gawhar S̲h̲ad complex, Harāt, 821/1418) excluded an equal emphasis on architecture. This may explain the continued popularity of so many standard mosque types—the domed hypostyle (Ziyāratgāh, Masd̲j̲id-i Čihil Sutūn ca. 890/1485) and the two-īwān type so long familiar in K̲h̲urāsān (Bad̲j̲istān and Nīs̲h̲āpūr d̲j̲āmiʿs, both later 9th/15th century)—to say nothing of the emphasis on refurbis̲h̲ing earlier mosques (d̲j̲āmiʿs of Iṣfahān, 880/1475 and Harāt, 903-5/1497-9), which, in accordance with the Tīmūrid predilection for innovative vaulting, often took the form of transversely vaulted halls (d̲j̲amiʿ s of Abarḳūh, 808/1415; Yazd, 819/1416; S̲h̲īrāz, ca. 820/1417; Maribud, 867/1462; and Kās̲h̲ān, 867-8/1462-3; and the mosques of Sar-i Rīg, 828/1424 and Mīr Čaḳinaḳ, 840-1/1436-7, at Yazd). There was also still ample room for surprises. The winter prayer hall added to the Iṣfahān d̲j̲āmiʿ in 851/1447 has multiple aisles of huge pointed arches springing directly from the ground and lit by ochre alabaster slabs let into the vaults and diffusing a golden radiance. The hoary 4-īwān formula was given a new twist by the addition of twin domed chambers flanking the sanctuary īwān (Harāt d̲j̲āmiʿ, 9th/15th century), an idea which infiltrated other plan types too (Rus̲h̲k̲h̲ar d̲j̲āmiʿ, 859/1454). At D̲j̲ād̲j̲arm (late 9th/late 15th century?) the central axis marked by the domed chamber and the courtyard is flanked on each side by a trio of vaulted bays.

Yet perhaps the most original mosque designs of the period were those which focused on the single dome and thus echoed, if only distantly, the preoccupations of contemporary Ottoman architects. This concept manifested itself in several different ways. In the Masd̲j̲id-i Gunbād, Ziyāratgāh (ca. 887-912/1483-1506), a square exterior encloses small corner chambers and a cruciform domed central area, a layout more reminiscent of a palace pavilion than a mosque. The core of the Masd̲j̲id-i S̲h̲āh, Mas̲h̲had (855/1451), is again a large domed chamber, but this is enclosed by a vaulted ambulatory and preceded by a long façade with corner minarets and a portal īwān. Most ambitious of all, however, is the Blue Mosque in Tabrīz (870/1465) in which a similar idea is given much more integrated expression by virtue of the open-plan arrangement of the central space. The dome springs from eight massive piers, but this octagon has further piers in the corners, making it a square with twelve openings, and thus offering easy access to the multidomed ambulatory. A similar openness characterises the gallery area and ensures that this mosque, though entirely covered, was airy, spacious and flooded with light. The range and subtlety of its polychrome tilework makes this mosque an apt coda for a period which exploited to an unprecedented degree the role of colour in architecture.

(e) The Ṣafavid period. The restoration and enlargement of existing mosques, a trend already noted in Tīmūrid times, continued apace in the Ṣafavid period, and involved over a score of mosques in the 10th/16th century alone. Yet not one new mosque of the first importance survives from this century, though the Masd̲j̲id-i ʿAlī in Iṣfahān (929/1522), a classic 4-īwān structure, has a sanctuary whose openplan dome on pendentives provides a bridge between the Blue Mosque in Tibrīz and the Luṭfallāh mosque in Iṣfahān (1011-28/1602-10). The latter, a private oratory for S̲h̲āh ʿAbbās I, makes a very public break with tradition, for it is simply a huge square chamber. Its lofty dome rests on eight arches via an intermediary zone of 32 niches. The whole interior is sheathed in glittering tilework whose smooth surfaces simplify all structural subtleties. Though the mosque is correctly oriented towards Mecca, it is set at an angle to the great square (maydān) from which it is entered, an angle dissimulated by the portal īwān which instead obeys the orientation of the maydān towards the cardinal points of the compass. A low vaulted passage linking īwān and dome chamber, but invisible from either, resolves these conflicting axes. It also draws attention to a discrepancy which could easily have been avoided and is therefore deliberate.

In the nearby Masd̲j̲id-i S̲h̲āh (1021-40/1612-30), which also fronts the maydān, the problem of discordant axes is solved with sovereign ease, for the portal leads into a diagonal vestibule which in turn opens into a 4-īwān courtyard now correctly orientated. Both portal and ḳibla īwāns have paired minarets to assert their importance. The scale is vast, but the entire mosque is conceived in due proportion to it. As at the comparably large mosque of Bībī K̲h̲ānum, dome chambers behind the lateral īwāns give extra space for prayer, while two madrasas with courtyards flank the main courtyard to the south. Thus even at the height of its popularity, the 4- īwān mosque could accommodate quite major innovations without impairing its essential character. Later Ṣafavid mosques, such as the d̲j̲āmiʿ s of Sarm and Čas̲h̲um, the Masd̲j̲id-i Wazīr in Kās̲h̲ān and that of ʿAlī Ḳulī Ag̲h̲a in Iṣfahān, serve by their very modesty, however, to highlight the altogether exceptional status of the two mosques on the Iṣfahān maydān. Even such a spacious and handsome version of the traditional 4-īwān schema as the Masd̲j̲id-i Ḥakim, Iṣfahān (1067/1656) could not fail to be an anticlimax in their wake.

5. The Turkish tradition.

(a) Early domed mosques. The earliest Anatolian mosques follow Arab prototypes, and by degrees some of them take on an Iranian colouring, especially in their free use of īwāns for portals and for sanctuary entrances. Already by the 7th/13th century, however, an emphasis on the isolated domed chamber as a mosque type began to make itself felt. This idea too might have had Iranian origins, but it soon developed in ways that owed nothing to Iran, since the contemporary preference for entirely covered mosques with no courtyard was itself enough to encourage experiments in the articulation of interior space. The dome quickly became the most favoured device to this end. In Iran, by and large, the domed chamber behind the ḳibla īwān remained spatially isolated from the rest of the mosque. In Anatolia, by contrast, architects were always seeking new ways of integrating the main domed space with the area around it. A consistent emphasis on domical forms created the necessary visual unity to achieve this. Already in the Sald̲j̲ūḳ period tentative experiments in this direction may be noted, for example the ʿAlāʾ al-Dīn mosque, Niğde (620/1223), whose ḳibla is marked by three domed and cross-vaulted bays with further parallel aisles behind. In the Ulu Cami of Bitlis (555/1160), a single great dome replaces these smaller bays, while in the Gök mosque and madrasa, Amasya (665/1266), the masd̲j̲id comprises a series of tripledomed aisles. Experiment with domical forms was therefore deeply rooted in Anatolian architecture from the beginning. It is above all, however, the hallmark of mosques erected by the Ottomans, and can be traced to the very earliest years of that dynasty.

(b) Ottoman architecture before 857/1453. The sequence begins very modestly with a series of mosques comprising a simple domed cube with a lateral vestibule (ʿAlāʾ al-Dīn mosque, Bursa, 736/1335, a structure typical of well over a score of such Ottoman mosques built in the course of the 8th/14th century) and minor variants of this schema, such as the mosque of Orhan Gazi, Bilecik, and the Yeşil Cami, Iznik, 780/1378. Such structures have a natural affinity with larger mausolea throughout the Islamic world, and with the simplest forms of Iranian mosques. It is only with hindsight that their significance for later developments, in which the theme of the single, and (above all) central, dominant dome of ever-increasing size becomes steadily more important, can be appreciated. This, then, is the main line of evolution in Ottoman mosque architecture, and the discussion will return to it shortly.

Meanwhile, two other types of mosque, in which the dome also loomed large, deserve brief investigation, especially as they bade fair in the formative early years to oust the domed, centrally planned mosque as the favoured Ottoman type, and also because they had their own part to play in the final synthesis of the 10th/16th century. The presence of three major types of domed mosque in the same century is a reminder that the pace of change was uneven. Several mosques conceived on an altogether larger scale rejuvenated the hypostyle form by investigating the impact of multiple adjoining domes. In some cases, like the Ulu Cami, Bursa, of 797/1394, a simple square subdivided into 20 domed bays of equal width though of varying height—the choice of the dome as the agent of vaulting is a diagnostic Ottoman feature—the effect was distinctly old-fashioned. At ground level this is an Arab mosque, even if its elevation is Anatolian. Contemporary with this, but marking a very different attitude to interior space, are two mosques in Bursa, that of Yi̊ldi̊ri̊m Bāyazīd, 794/1390, and the Yeşil Cami of 816/1413, which use the dome motif on various scales and thus far more imaginatively. They represent a second preparatory stage on the way to the mature Ottoman mosque, and their large layout is by turn cruciform, stepped or of inverted T-type. Their distinguishing feature is the use of several domes of different sizes. In the two cases under discussion, the inverted T-plan highlights the miḥrāb aisle by two adjoining domes along the central axis flanked by a trio of domed or vaulted bays on each side, the whole knit together laterally by a 5-domed portico. Sandwiched between these two buildings in date is the Ulu Cami of Edirne, 806/1403, where the square is subdivided into nine equal bays, eight of them domed, with a domed and vaulted portico tacked on. At the mosque of Čelebi Sulṭān Meḥemmed, Dimetoka, this arrangement is refined by an increased concentration on the central dome, which is enveloped by vaults on the main axes and diagonals, the whole preceded by a 3-domed portico. Such a combination cannot fail to recall the standard quincunx plan, complete with narthex, of mid-Byzantine churches, and it was of course these buildings which dominated the Anatolian countryside in the early centuries of Turkish occupation. Steady Byzantine influence can be seen to have affected the evolution of Ottoman architecture even before the capture of Istanbul brought Turkish architects face to face with Hagia Sophia. Yet it would be grossly mistaken to regard mature Ottoman mosques as mere derivatives of Hagia Sophia. The Uç Şerefeli mosque, Edirne, of 851/1447, with its huge central dome on a hexagonal base flanked on either side by a pair of much smaller domes and preceded by a lateral courtyard enclosed by 22 domed bays, makes excellent sense within a purely Ottoman perspective as a key stage in the evolution which terminated in the great masterpieces of Sinān. The divergence between the great dome and the lesser ones flanking it has already become acute and was to end in their total suppression.

Yet one significant element, crucial to Hagia Sophia and a cliché of Ottoman architecture after 857/1453, had not yet entered the architectural vocabulary of the Turkish mosque before that date. This was the use of two full semi-domes along the miḥrāb axis to buttress the main dome. The long-rooted Islamic custom of marking the miḥrāb bay by a great dome rendered such a feature otiose. Once the decision had been taken to make the largest dome the central feature of a much larger square, the way was open for the adoption of this Byzantine feature, and with it the transformation and enrichment of interior space was a foregone conclusion. Otherwise, most of the architectural vocabulary used in mature Ottoman mosques was already to hand by 857/1453: flying buttresses, the undulating exterior profile created by multiple domes, tall pencil-shaped minarets and a certain parsimony of exterior ornament allied to exquisite stereotomy. It has to be admitted, however, that these features had yet to find their full potential, notably in the failure to develop a suitably imposing exterior to match the spatial splendours within. That potential could be realised only when these features were used in tandem with each other by masters seeking to express a newly-won confidence and bent on creating an integrated style for that purpose. The mosque was, moreover, their chosen instrument; indeed, Ottoman architecture is, first and foremost, an architecture of mosques.

(c) Ottoman architecture after 857/1453. The capture of Constantinople in 857/1453 provided both a terminus and an impetus to a radical rethinking of mosque design. Appropriately enough, the first building to express the new mood was a victory monument, as its name indicates: the Fātiḥ Mosque (867-75/1463-70). This has a single huge semi-dome buttressing the main one but also displacing it off the main axis; clearly, the spatial, aesthetic and structural implications of such a semidome had not yet been fully grasped. Within a generation, this anomaly at least had been rectified; the mosque of Bāyazīd II (completed 913/1506) has two such semi-domes on the miḥrāb axis, with four lesser domes flanking this central corridor on each side. On the other hand, the projecting portico sandwiched between dome chamber and courtyard is a clumsy and lopsided expedient with little functional justification. Yet the resultant emphasis on the portico is wholly typical of a period in which this feature re-appeared under numerous guises, especially in doubled form (Mihrimah mosque, completed ca. 973/1565). The Şehzade mosque (955/1548) presents a much more streamlined appearance, with dome chamber and courtyard of approximately equal proportions. Within the sanctuary, the great central dome opens into semi-domes on all four sides, with small diagonal semi-domes opening off the main ones and corner domes. It is instructive thus to see Ottoman architects developing the possibilities of the centralised plan like the builders of Christian churches and martyria a millennium before, and coming to very similar conclusions. Smaller mosques with domes on hexagonal (Ahmed Paşa, completed ca. 970/1562) or octagonal bases (Mihrimah mosque) were scarcely less popular than domed squares. A small number of wooden-roofed mosques perpetuating earlier modes, and with their roots in the Arab tradition, survive (e.g. Ramazan Efendi in Koçamustafapaşa, 994/1585, and Tekkeci Ibrahim Ağa, 999/1590) as reminders of a very widespread type of Ottoman mosque now almost entirely eclipsed by more durable structures.

In the ferment of experiment which marks 10th/16th century Ottoman architecture, the key figure was undoubtedly Sinān, an Islamic equivalent to Sir Christopher Wren, who transformed the face of the capital city as of the provinces with some 334 buildings (mostly mosques) erected in his own lifetime, and whose pivotal role as chief court architect (effectively Master of Works) allowed him to stamp his ideas on public architecture from Algeria to ʿIrāḳ and from Thrace to Arabia in the course of a phenomenally long career which spanned virtually the entire century. The Süleymaniye mosque in Istanbul (963/1556) is by common consent the masterpiece of his middle age. It takes up and refines the model of the Bāyazīd II mosque by adding ideas taken from the Şehzade mosque, like the succession of semi-domed spaces billowing out from the main dome, though only along the principal axis. Huge arches serve to compartmentalise the spatial volumes.

All these mosques are preceded by an open courtyard whose cloister is roofed by long files of adjoining domes. This standard feature typifies the new emphasis on subsidiary structures, mausolea, ʿimārets, madrasas and the like, and the consistent attempt to integrate them visually with the sanctuary itself, for example by subordinating them to the principal axes of the design. All this implies a marked increase in scale and a new sensitivity to the landscaping of the ensemble. Hence the recurrent choice of dramatic sites for these mosques, especially in Istanbul with its built-in vistas along the Bosphorus. This awareness of topography as a feature of mosque design is evident as early as the Fātiḥ mosque; its three parallel axes are grouped around and within an enclosed open piazza measuring some 210 m. per side. The climax of mature Ottoman architecture is reached with Sinān’s final masterpiece, the Selimiye at Edirne (982/1574), in which the largest of Ottoman central domes (31.28 m. in diameter, hedged externally by the loftiest quartet of Ottoman minarets (70.89 m. high) rests on eight piers pushed as close to the walls as safety will allow so as to create the largest possible open space.

While the increase in the absolute height and breadth of these great domed chambers is striking, the amount of articulation and detail crammed into these spaces is scarcely less impressive. All is subordinated to a formidable concentration of purpose—for example, the carefully considered fenestration, surely a legacy from Hagia Sophia, with its superposed groupings of eights and sixes or sevens, fives and threes. In the interests of creating the maximum untrammelled space, thrusts are concentrated onto a few huge piers with spherical pendentives between them, and thus the layout is a model of clarity and logic. Flooded with light, their volumetric subdivisions apparent at a glance, these interiors are at the opposite pole from the dim mysteries of Hagia Sophia. Frescoes reminiscent of manuscript illumination and of carpet designs vie with Iznik tiles to decorate the interior surfaces, and often (as in the case of fluted piers) to deny their sheer mass.

Externally, these mosques attest a well-nigh fugal complexity by virtue of their obsessive concentration on a very few articulating devices like windows, arches and domes. The repetition of the same forms on varying scales intensifies the sense of unity. Even the minarets which mark the outer limits of the mosque’s surface area are brought into play; for example, those of the Sultan Ahmed Mosque (completed 1025/1616) have the bases of their balconies so calibrated as to coincide with the top of the main dome, its collar and the collar of the main subsidiary half-domes, while their location at the corners of the building binds it together and defines the sacred space from afar. Detailing is sparse and crisp, with a strong linear emphasis, a flawless sense of interval and a pronounced attenuation of features like wall niches and engaged columns (Süleymaniye mosque). Nothing is allowed to impair the primary aesthetic impact of cliff-like expanses of smooth grey stone. Most notable of all is a dramatic but ordered stacking of units culminating in the great dome which crowns and developes the entire ensemble. These individual units are each locked into place within a gently sloping pyramidal structure whose inevitable climax is the central dome. From this peak the subsidiary domes, semi-domes and domed buttresses cascade downwards to form a rippling but tightly interlocked silhouette. These highly articulated exteriors are a triumphant reversal of the standard Islamic preference in mosque architecture for stressing the interior at the expense of the exterior. As the viewpoint changes, so too does the profile of these mosques, from a continuous smoothly undulating line to a series of sharp angular projections formed by stepped buttresses and roof-turrets. The preference for saucer domes rather than pointed domes with a high stilt fosters the sense of immovable, rock-like stability, with the topmost dome clamped like a lid onto the mobile, agitated roof-lines beneath it.

This, then, can justly claim to be architects’ architecture. It merits that term by virtue of its unbroken concentration on the single germinal idea of the domed centralised mosque. It is against that consistent unity of vision that the role of the Hagia Sophia must be assessed. Of course, Turkish architects were not blind to its many subtleties, and they freely quarried it for ideas. But it was as much a challenge that inspired them to emulation as it was a source for technical expertise. Finally, it was the Ottomans who succeeded where the Byzantines had failed: in devising for these great domed places of worship an exterior profile worthy of the splendours within. The triumphant issue of their labours to that end can be read along the Istanbul skyline to this day.

Bibliography

G. Gurlitt, Die Baukunst Konstantinopels, 3 vols., Berlin 1907-12

F. Sarre and E. Herzfeld, Archäologische Reise im Euphrat- und Tigrisgebiet, 4 vols., Berlin 1911-20

M. S. Briggs, Muhammadan architecture in Egypt and Palestine, Oxford 1924

A. Gabriel, Les mosquées de Constantinople, in Syria, vii (1926), 353-419

E. Diez, Masd̲j̲id. III. Architecture, in EI 1, III, 378-89

E. Pauty, L’évolution du dispositif en T dans les mosquées à portiques, in Ber.Or., ii, (1932), 91-124

L. Y. Hautecoeur and G. Wiet, Les Mosquées du Caire, 2 vols., Paris 1932-4

Gabriel, Monuments turcs d’Anatolie, 2 vols., Paris 1934

M. B. Smith, Material for a corpus of early Iranian Islamic architecture. I-III, in Ars Islamica, ii (1935), 153-71, iv (1937), 1-40, vi (1939), 1-10

A. Godard, Les anciennes mosquées de l’Irān, in At̲h̲ār-é Īrān, i/2 (1936), 187-210, continued in Arts Asiatiques, iii (1956), 48-63, 83-8

idem, Historique du Masd̲j̲id-é D̲j̲umʿa d’Iṣfahān, in At̲h̲ār-é Īrān, i/2 (1936), 213-82

B. Maslow, Les mosquées de Fès, Paris 1937

J. Sauvaget, Observations sur quelques mosquées seldjoukides, in AIEO Alger, iv (1938), 81-120

A. U. Pope and P. Ackerman (eds.). A survey of Persian art from prehistoric times to the present, 6 vols., Oxford 1938-9

K. A. C. Creswell, Early Muslim architecture. II. Early ʿAbbāsids, Umayyads of Cordova, Aghlabids, Ṭūlūnids and Samānids, A.D. 751-905, Oxford 1940

Gabriel and Sauvaget, Voyages archéologiques dans la Turquie Orientale, 2 vols., Paris 1940

M. Akkush, Contribution à une étude des origines de l’architecture musulmane. La Grande Mosquée de Médine (al-Ḥaram al-Madanī), in Mélanges Maspero, iii (1940), 377-410

Sauvaget, La mosquée omeyyade de Médina, Paris 1947

G. Marçais, L’église et la mosquée, in L’Islam et l’Occident. Cahiers du Sud, Marseilles 1947, 174-84

E. Kühnel, Die Moschee. Bedeutung, Einrichtung und kunsthistorische Entwicklung der islamischen Kultstätte, Berlin 1949

E. Lambert, La synagogue de Dura-Europos et les origines de la mosquée, in Semitica, iii (1950), 67-72

H. Stern, Les origines de l’architecture de la mosquée omeyyade, in Syria, xxviii (1951), 269-79

L. Torres Balbás, Origen de las disposiciones arquitectónicas de las mezquitas, in Al-Andalus, xvii (1952), 388-99

Creswell, The Muslim architecture of Egypt. I. Ikhshīds and Fāṭimids, A.D. 937-1171. II. Ayyūbids and Early Baḥrite Mamlūks, A.D. 1171-1326. Oxford 1952-60

U. Vogt-Göknil, Türkische Moscheen, Zürich 1953

E. Egli, Sinan, der Baumeister osmanischer Glanzzeit, Zürich 1954

Marçais, L’architecture musulmane d’Occident: Tunisie, Algérie, Maroc, Espagne et Sicile, Paris 1955

D. N. Wilber, The architecture of Islamic Iran. The Il-Khānid Period, Princeton 1955

Lambert, Les origines de la mosquée et l’architecture religieuse des Omeyyades, in SI, vi (1956), 5-18

Gabriel, Une Capitale turque Brousse, Bursa, 2 vols., Paris 1958

G. A. Pugačenkova, Puti razvitiy̲a̲ ark̲h̲itekturi Yuz̲h̲nogo Turkmenistana pori rabovladeniy̲a̲ feodalizma, in Trudi̊ Yuz̲h̲no- Turkmenistanskoi Ark̲h̲eologičeskoi Ekspeditsii, vi, Moscow 1958

L. Golvin, La Mosquée. Ses origines. Sa morphologie. Ses diverses fonctions. Son rôle dans la vie musulmane, plus spécialement en Afrique du Nord, Algiers 1960

M. Useinov, L. S. Bretanitski and A. Salamzade, Istoriy̲a̲ ark̲h̲itekturi Azerbaidz̲h̲ana, Moscow 1963

A. Dietrich, Die Moscheen von Gurgan zur Omaijadenzeit, in Isl., xl (1964), 1-17

Pugačenkova and L. I. Rempel’, Istoriy̲a̲ Iskusstva Uzbekistana, Moscow 1965

A. Lézine, Architecture de l’Ifriqiya, Paris 1966

Bretanitski, Zodčestvo Azerbaidz̲h̲ana XII-XV v. v., i ego mesto v ark̲h̲itekture perednego vostoka, Moscow 1966

A. Kuran, The mosque in early Ottoman architecture, Chicago and London 1968

O. Grabar, La Grande Mosquée de Damas et les origines architecturales de la mosquée, in Synthronon. Art et Archéologie de la fin de l’Antiquité et du Moyen Âge. Recueil d’Études, Paris 1968, 107-14

idem, The architecture of the Middle Eastern city from past to present: the case of the mosque, in Middle Eastern cities, ed. I. M. Lapidus, Berkeley and Los Angeles 1969, 26-46

Creswell, Early Muslim architecture. Umayyāds. A.D. 622-750, 2 vols., Oxford 1969

J. Sourdel-Thomine, La mosquée et la madrasa. Types monumentaux caractéristiques de l’art islamique médiéval, in Cahiers de civilisation médiévale X e -XII e siècles, Université de Poitiers, Centre d’Études Supérieures de Civilisation Médiévale, xiii/2 (1970), 97-115

Golvin, Essai sur l’architecture religieuse musulmane, i-iv, Paris 1970-6

G. Goodwin, A history of Ottoman architecture, London 1971

O. Aslanapa, Turkish art and architecture, tr. A. Mill, London 1971

Kuran, Thirteenth and fourteenth century mosques in Turkey, in Archaeology, xxiv/3 (1971), 234-54

S. Ögel, Der Kuppelraum in der türkischen Architektur, Istanbul 1972

R. A. Jairazbhoy, An outline of Islamic architecture, Bombay 1972

R. Bourouiba, L’art religieux musulman en Algérie, Algiers 1973

Grabar, The formation of Islamic art, New Haven 1973

D. Kuban, Muslim religious architecture, 2 vols., Leiden 1974-85

R. Hillenbrand, Saljūq dome chambers in North-west Iran, in Iran, xiv (1976), 93-102

D. Hill, Golvin and Hillenbrand, Islamic architecture in North Africa, London 1976

J. D. Hoag, Islamic architecture, New York 1977

Vogt-Göknil, Die Moschee. Grundformen sakraler Baukunst, Zürich 1978

C. Ewert and J.-P. Wisshak, Forschungen zur almohadischen Moschee. Lieferung 1: Vorstufen. Hierarchische Gliederungen westislamischer Betsäle des 8. bis 11. Jahrhunderts: Die Hauptmoscheen von Qairawan und Cordoba und ihr Bannkreis, Mainz 1981

B. Finster, Islamische Bau- und Kunstdenkmäler im Yemen, in Archäologische Berichte aus dem Yemen, i (1982), 223-75

R. B. Serjeant and R. Lewcock (eds.), Ṣanʿāʾ. An Arabian Islamic city, London 1983

Hillenbrand, The mosque in the medieval Islamic world, in Architecture in continuity. Building in the Islamic world today, ed. S. Cantacuzino, New York 1985, 30-51.

(R. Hillenbrand)

II. In Muslim India

A. Typology.

The nature of the regional building styles and their characteristic decoration have been treated s.v. hind. vii. Architecture, in Vol. III above. This section deals with the essential typology of mosques in India, and excludes the simplest structures used only for occasional. prayer such as the ḳibla -indications at some tombs and graveyards [see maḳbara. 5. India], and the special structures (ʿīdgāh) provided for the ʿīds ; for these see muṣallā. 2.

The continuous history of the mosque begins with the M. Ḳuwwat al-Islām in Dihlī, founded immediately after the Muslim conquest in 587/1191. There are however records of mosques founded earlier, e.g. under the ʿAbbāsid caliphate in Sind [q.v.], by small communities of Muslim traders, especially in Gud̲j̲arāt and the Malabār coast, and by individual Ṣūfī pīrs who gathered a community around them. The remains of these are mostly too exiguous to be of value in a general statement. Recent explorations by M. Shokoohy, not yet published, have revealed a few structures, of a century or two before the conquest, at Bhadres̲h̲war in Gud̲j̲arāt. These, in common with the first structures of any fresh conquest of expansion, are constructed from the remains of Hindū buildings; in the case of mosques built after a conquest there has been a deliberate pillaging of Hindū or D̲j̲ayn temples, as an assertion of superiority as well as for the expediency of making use of material already quarried and of local impressed labour before the arrival of Muslim artisans. Examples of this are cited for different regions of India s.v. hind, vii. Architecture, in Vol. III, p. 441 above. (It should be pointed out that the practice of pillaging the buildings of the conquered is known in India in the case of rival Hindū kings also.)

Where a mosque is actually constructed on the plinth of a destroyed Hindū building (e.g. M. Ḳuwwat al-Islām at Dihli; At́alā M. at D̲j̲awnpur) the ḳibla [q.v.] will probably not be accurately located and the original cardinal west made to serve the purpose; but in general an effort is made to observe the correct ḳibla, which varies between 20° north of west in the south of India to 25° south of west in the extreme. north, with a conventional west used only rarely in original buildings.

Mosques which might be described as “public”—i.e. not only the Masd̲j̲id-i d̲j̲āmiʿ of a particular locality (and of course in a conurbation there may be a separate d̲j̲āmiʿ for each original maḥalla) but also the individually-founded or endowed mosques within a town—are enclosed on all sides. This has not been required of mosques within a sarāʾī or a dargāh, or when the mosque is an adjunct of a tomb, and there are countless instances of small private mosques where there seems never to have been any enclosure. The enclosure for the public mosque is particularly necessary for Islam in partibus infidelium, and those courtyards which are not enclosed are protected from the infidel gaze in some other way, e.g. by the ṣaḥn standing on a high plinth (examples: the D̲j̲āmiʿ M. at S̲h̲āhd̲j̲ahānābād. Dihlī, At́alā M. at D̲j̲awnpur. where in both the courtyard is limited only by an open arcade or colonnade). The principal entrance is usually on the east, although any gate may be on occasion specified as a royal entrance; it is rare, though not unknown, for any entrance to be made in the western wall, and where this has happened it is not designed for access by the general public. The internal position of the principal miḥrāb [q.v.], sometimes of subsidiary miḥrābs also, is indicated on the outside of the west wall by one or more buttresses; a feature of mosques in India is the way the exterior elevation of the west wall is brought to life by decorative expedients.

The interior of the mosque admits of little variation outside two well-defined types. In one the western end (known in India as līwān) is a simple arrangement of columns supporting a roof, usually of at least three bays in depth but possibly of many more; the roof may be supported by beam-and-bracket or by the arch; the former arrangement being by no means confined to compilations of pillaged Hindū/D̲j̲ayn material. The līwān openings may be connected directly with the arcades or colonnades of other sides of the ṣaḥn. Where Hindū material has been used it is usually necessary to superimpose one column upon another in order to gain sufficient height, for not infrequently a mezzanine gallery may be incorporated in the structure, in the līwān or in the side riwāḳ s. These are frequently referred to as “women’s galleries”, but this is surely impossible unless they are placed to the rear of the structure so that women may not make their prayers in front of men; gallery structures in the līwān are more likely to be either reserved for royal (male) use or to be čillas for the use of a local pīr. In the other type, the līwān is physically separated from the ṣaḥn by a screen of arches (maḳṣūra), which may conceal a columnar structure to the west, as in the M. Ḳuwwat al-Islām where the maḳṣūra is a later addition to the original structure, or in the mosques of Gud̲j̲arāt where the arch is not used with as much freedom as in other styles. More commonly, however, the arches of the maḳṣūra are part of a vaulting system whereby the līwān is composed into one or more halls; there is always an odd number of maḳṣūra arches, and it is common for the bay which stands in front of the principal miḥrāb to be singled out for special treatment, either by being made taller than the rest, or by being specially decorated (the latter treatment common in the mosques of Bīd̲j̲āpur [q.v.]). (This is not invariably the central bay, as mosques are not necessarily symmetrical about the principal miḥrāb axis; cf. the “Stonecutters’ M.” in Fatḥpur Sikrī, where a čilia occupies two additional bays at the north end of the līwān, or the Aŕhāʾī Kangūra M. at Kās̲h̲ī Banāras, where the side riwāḳs of the līwān are of unequal length.) In one mosque at Bīd̲j̲āpur (Makkā M.), the līwān stands within and unattached to the surrounding courtyard. A staircase is commonly provided to give access to the līwān roof, either separately or incorporated within the walls or the base of a minaret, as this is a favourite place from which to call the ād̲h̲ān ; a staircase may be provided within a gateway for the same purpose. The līwān roof may be surmounted by one or more domes. Inside the līwān, the principal miḥrāb stands within the west wall opposite the main opening; if there are other miḥrābs, the central one is always the most sumptuously decorated and may be set deeper within the west wall than the other. The minbar is usually a permanent stone structure, with an odd number of steps, only occasionally made an object of decoration (splendid examples in the older Bengal mosques and in the Mālwā sultanate). A simple minbar is often provided when not liturgically necessary, as in the mosque attached to a tomb. There is an exceptional case at Bīd̲j̲āpur, at the mosque building for the cenotaph of Afḍal K̲h̲ān: the mosque is two-storeyed, the two halls being exactly similar except that a minbar is provided only in the lower one. (In another first-floor mosque at Bīd̲j̲āpur, the And́ā M., there is no minbar; the ground floor is apparently a well-guarded sarāʾī, and the suggestion has been made that the whole structure was intended for zanāna use.) The floor of the līwān is often marked out into muṣallās of miḥrābī shape for each individual worshipper. Lamps may be suspended from the līwān ceiling. The līwān façade is open to the ṣaḥn ; i.e. there is never any portion closed off like the zimistān of Persian mosques.

The ṣaḥn is usually an open courtyard, containing a ḥaẉḍ [q.v.] for the wuḍūʾ ; this is usually placed centrally, except that in some S̲h̲īʿī mosques the ḥawḍ may be placed to one side of the central axis. There are rare cases where the ṣaḥn is completely or partially covered (e.g. the D̲j̲āmiʿ M. at Gulbargā [q.v.] is completely covered; in two mosques of the Tug̲h̲luḳid period at Dihlī, K̲h̲iŕkī M. and Sand̲j̲ar (Kālī) M., additional riwāḳ s leave only four small open courtyards in the middle of the ṣaḥn). In such cases provision must be made for the wuḍūʾ outside the ṣaḥn; some major mosques may also make provision, outside the ṣaḥn, for the g̲h̲uṣl. In some Gud̲j̲arāt mosques there is a water reservoir under the floor of the ṣaḥn, sometimes with chambers wherein to take refuge from the heat of the sun, with some sort of kiosk standing in the ṣaḥn from which water may be drawn; the idea is imitated on a small scale in the floor of the D̲j̲āmiʿ M. in Fatḥpur Sikrī. In one complex (Rad̲j̲ōn kī bāʾīn) south of the M. Ḳuwwat al-Islām the mosque and an associated tomb seem subordinate to an enormous step-well (bāʾōlī [q.v.]).

One or more bays of the side or end riwāḳs may be closed off for a special purpose, e.g. to make a room for relics, or to serve as a room for the ḳāḍī or mutawallī ; in S̲h̲īʿī mosques, sometimes to house the ʿalam s, etc., but these are usually accommodated in the Imāmbāŕā or ʿĀs̲h̲ūrā-k̲h̲āna where there is one. The use of part of the mosque as a madrasa [q.v.] is commonplace, and many instances could be cited at the present day where there is no special provision for such a purpose; but there are instances of a special building forming an integral appendage of the mosque designated as a madrasa; e.g. M. K̲h̲ayr al-Manāzil, near the Purānā Ḳilʿa in Dihlī, where the northern riwāḳ, of two storeys, forms the madrasa of the foundation.

The ṣaḥn may be used also for graves, from the simplest tombstone to elaborate mausoleums (see maḳbara. 5); e.g. the D̲j̲āmīʿ M. of Fatḥpur Sikrī, where most of the northern side of the ṣaḥn is occupied by the tomb of Salīm Čis̲h̲tī, the Zanāna Rawḍa, and the tomb of Nawwāb Islām K̲h̲ān (not so designed originally, and possibly a d̲j̲amāʿat-k̲h̲āna for the saint’s disciples).

A mīnār is by no means an invariable appendage to the Indian mosque; apart from a few occasional early instances, only in the Gud̲j̲arāt sultanate, and in Burhānpur in K̲h̲āndēs̲h̲. was a functional mīnār provided for the ad̲h̲ān before the Mug̲h̲al period; after the 10th/16th century, the mīnār becomes common, but not invariable. See further manāra. 2. India.

The administration of the mosque may be under the ḳāḍī [q.v.] or, in the case of larger foundations, a committee headed by a mutawallī [q.v.]. Where a mosque stands on a high plinth there may be openings in it sufficiently large to be rented off as storerooms or to traders, in which case the revenues accrue to the mosque; see also waḳf.

Bibliography There are no studies dealing with mosque typology alone

for works on all architectural aspects, see the Bibliographies to hind. vii. Architecture, and Section B. below.

(J. Burton-Page)

B. The monuments.

The development of the mosque in the subcontinent can be recognised as an adaptation of the Arab prototype, largely as already modified by Iranian builders, to local materials, climate, and the proclivities of a long-established tradition of architecture and ornament. The Arab elements in this fusion were those basic to the expression of the d̲j̲amāʿat, the collective act of prayer and the simple, egalitarian liturgy: the courtyard and its protective enclosure, the ḳibla wall, here on the western side, the ẓulla or prayer hall, here known as līwān, along the western wall, and colonnades, riwāḳ or dālān, along the other sides, with an essential severity of outline and a spare orthogonal framework. The Iranian elements were rhythmic arcading, the prominent use of pīs̲h̲tāḳ [q.v.] or frontispiece alcoves, the voussoired dome, ultimately double, and a particular sense of proportion; minarets did not become general until relatively late, and then often as decorative rather than functional features. A gamut of Iranian decorative devices including ceramic tiles [see kās̲h̲ī ]. cut plaster-work, gač-barī, plaster relief work, munabbat-kārī, and pietra dura inlay, parčīn-kārī, besides the pseudostructural pendentive-work, ḳalūb-kārī, or squinch-netting. The Indian elements, within the context of an elaborated stonecutting technique, were initially a certain heaviness due to the stone itself (especially in corbelled domes), complexity in individual forms, a vibration set up by the reiteration of forms at different scales, an interest in diagonal axes, and an overwhelming fertility of imagination in carved ornament. Indian traditions of massing only influenced mosque design in a limited way, and then largely through changes in dome form and grouping. The traditions of temple building were in strong contrast, creating massive, highly ornate enclosures within which progressively more intimate cells led to individual confrontation with a deity; the vertical extension was frequently emphasised as much as the horizontal. Despite this difference, a reconciliation of these traditions led to an enlivening of the mosque outline, especially on the skyline, with a frequent play of pinnacles and pavilions, much use of receding planes, and in some cases a culminating centrality comparable with the Ottoman achievement. The underlying Arab archetype retained its simplicity of arrangement in most regions, though periodically transformed in others. Evidence for the direct transfer of skills from temple-building to mosque building, which can be deduced from the earlier forms, is provided by a Māru-Gurd̲j̲ara architectural manual of the 15th century A.D., the Vṛḳṣārṇava, in a chapter on the Rehmāṇa-prāsāda, or temple of Rehmāna, i.e. of Allāh, giving instructions for layout, orientation, superstructure and exclusively floral decoration, all within prescribed norms. The principal modifications attributable to the climate are a tendency to raise the courtyard level to catch wind currents and escape dust and noise, a tendency to pierce the courtyard walls to allow the currents through, and a preference for riverside sites. Specific architectural features are incorporated, notably the finaly pierced d̲j̲ālī screen to reduce glare, and the čhad̲j̲d̲j̲ā or eaves pent to throw off monsoon water and increase shade. A general trend in the chronological development is the movement from trabeated construction towards arcuate or vaulted forms, though this is achieved with some hesitation. This is in parallel with a progression from a somewhat provincial emulation of Iranian or Central Asian types through local technique to a much more accomplished creation of local types in which influence from the Vilāyat can still be traced. Although the relative neglect of the madrasa [q.v.] as a building form may have been due in part to a practice of teaching within the mosque, this seems not to have produced any overall adaptation of layout, unless in the development of the undercroft.

The Arab conquest of Sind. It is recorded that the first mosque in Sind was built by Muḥammad b. Ḳāsim at Daybul [q.v.] after his capture of the city in 92/711, followed by another at Multān [q.v.], next year; he was urged to build mosques in every town, the resources seized having proved unexpectedly large. A third great mosque was built at Manṣūra [q.v.] either by his son ca. 120/738, or in the early years of Abū D̲j̲aʿfar al-Manṣūr, i.e. after 136/754, with teak columns. Little remains of these. If Daybul is correctly identified with Bhambōr, and the uncertain date of 109/727 is right, then the mosque there may be among the oldest in Islam. Its plan is certainly close to that of Kūfa [q.v.], as rebuilt in 50/670, with the same double rows of columns for the riwāḳ, but only three aisles (of twelve bays) parallel to the ḳibla wall in lieu of five for the prayer hall; no trace has been found of a miḥrāb recess, but neither has one been found at Wāsiṭ [see miḥrāb ], as built under the same governor, al-Ḥad̲j̲d̲j̲ād̲j̲ b. Yūsuf. Outer bays of the riwāḳ were walled off to form cells, ḥud̲j̲ra, and stone bases contain traces of timber pillars. Another inscription gives 239/853-4; one in flowered Kufic for 294/906-7 probably refers to rebuilding after the earthquake of 280/893. The building thus conforms to the early ʿIrāḳī type, even to its strip foundations; though in yellow freestone, it lacks the stone columns. Pivots for gates in front of the līwān suggest some kind of maḳsūra. At Manṣūra, the D̲j̲āmiʿ Masd̲j̲id appears to have had a six-aisled prayer hall, built on an earlier Hindu site; three smaller mosques show careful alignment and external buttressing for a miḥrāb. In the absence of detail, the influence of these buildings is imponderable, but Daybul and Manṣūra survived until the 7th/13th century, and Manṣūra like Multān was taken by Maḥmūd of G̲h̲azna; they can hardly have been ignored. A further early mosque in Kačh, at Bhadreśvar, has been identified by Shokoohy as a rebuilding with purposely-carved stone ca. 560/1165. This has a prayer hall of two aisles, a double riwāḳ colonnade at the sides, and a single one to the east. The prominent miḥrāb is echoed outside the east wall, which faces an open hypostyle hall, no doubt for an overflow congregation. The roof is trabeated throughout, mostly on the east-west axis.

In the period preceding the Dihlī Sultanate, the principal mosques must have been at Lāhawr [q.v.], the G̲h̲aznawid centre (as Maḥmūdpūr) from 412/1021, including the K̲h̲is̲h̲tī Masd̲j̲id., of which nothing remains, though brickwork is still typical of the area.

Sultanate. At Dihlī [q.v.] the victory of Ḳuṭb al-Dīn [q.v.] was proclaimed by the creation (587/1191) of the Masd̲j̲id Ḳuwwat al-Islām, “The Might of Islam”, on a temple plinth, with stonework taken from 27 other temples by elephant-power. The plan, of the same ʿIrāḳī type, is here elongated on the east-west axis, and includes formally symmetrical entrances to the east, north and south. The colonnades in the prayer hall are four aisles deep, those to the east three, and those down the long sides two. The hall is now modified to include a row of five corbelled domes, above five miḥrābs, by adjustment of the bay spacing to carry octagonal systems of lintels; this roof was set higher than the riwāḳ roofs, and mezzanines were built at the four angles of the court, possibly for women. Ingenious use of the strongly articulated temple pillars, with cruciform capitals and internally tiered domes, achieved a relatively light, harmonious building, whose Hindu character was scarcely disguised. In 595/1199, however, a great frontal screen of five pointed arches was added to the hall. Its clearly-framed format, with the central arch much taller, is Iranian, and related to the G̲h̲ūrid S̲h̲āh-i Mas̲h̲had in G̲h̲ard̲j̲istān (571/1175-6), or the Ribāṭ-i S̲h̲araf [q.v.] (508/1114-15), but its construction is limited to Indian techniques, with corbelled arches. The marvellously vigorous combination of sinuous Hindu carving with ṭug̲h̲rā inscriptions makes fresh use of Indian skills for a Muslim purpose. The exaggerated height of this screen, with no direct relation to the hall behind, set a pattern for later buildings. In the same year Ḳuṭb al-Dīn began the immense Ḳuṭb Mīnār [q.v.] outside the southeast corner of the mosque, much like that at K̲h̲wād̲j̲a Siyāh Pus̲h̲ in Sīstān, as a symbol of the centrality of faith; minarets, if used at all in Hindustān, are usually symbolic rather than functional until Mug̲h̲al times. The exception is at Ad̲j̲mer. There the equally symbolic re-use of temple components as “the annihilation of idolatry” achieved more orderly expression in the Aŕhāʾī-dinkā D̲j̲hōńpŕā (595/1199), under Abū Bakr al-Hirawī, with some evidence of specially-cut masonry in the lower column-shafts and tiered domes (see Meister, op. cit.), and a single, exquisite, cusped marble miḥrāb. The court is almost square, and probably had nine domes on all four sides, though there are five aisles in the prayer hall to three elsewhere; the effect is spacious, well-lit and calm. A reeded shaft graced each external angle, and the site on a mound allowed a grand approach stair to the east. Here too a great screen wall was added, with seven arches, under Iltutmis̲h̲ (607-33/1211-36), two lateral arches on each side reflecting the cusped form of the miḥrāb; the central arch is less dominant than at Dihlī, but is surmounted by two minaret shafts (now stumps), reeded and creased like the Ḳuṭb, so emulating a Sald̲j̲ūḳ [q.v.] pīs̲h̲tāḳ. Iltutmis̲h̲ was to extend the work at Dihlī. Accepting Aybek’s plan, he enlarged the prayer hall by a further three domes to north and south, with corresponding miḥrābs and screen wall. Corbels on the latter suggest a double storey in each central bay, as in later work in Gud̲j̲arāt. The riwāḳ, built as before, now enclosed the first mosque, including the Mīnār, to which he added three storeys [see dihlī for plan and details] (completed 1229). The S̲h̲āhī D̲j̲āmiʿ Masd̲j̲id at Bari Khatu is of the same period and type, set on a high plinth; it introduces an ornate domed gallery over the east entrance. At Badāʾūn [q.v.] the great D̲j̲āmiʿ Masd̲j̲id built by Iltutmis̲h̲ in 620/1223-4 adheres to the same basic layout, but has been heavily rebuilt.

ʿAlāʾ al-Dīn K̲h̲ald̲j̲ī’s scheme to double the Ḳuwwat al-Islām again fell victim to its own ambition, for it was abandoned at his death. Remnants show that it respected the existing alignments in prayer hall, screen wall, and north gateway, and even in the immense ʿAlāʾī Mīnār which was to rise from the centre of the new prayer court. The inherent symmetry cannot have mitigated the disruption of worship by three courts set within each other. The only complete element to survive is the southern gateway, or ʿAlāʾī Darwāza (710/1311), set as a čārṭāḳ on the palace approach: an elegant, accomplished building of a new order. Its vocabulary is recognisable in the D̲j̲amāʿat K̲h̲āna at the dargāh of Niẓām al-Dīn (dated for his death 725/1325), fully Muslim in style, and built with new stone. This has no courtyard, but only a prayer hall of three domed chambers, to each of which there is a broad archway in the eastern façade. The square central space, almost the same size as the Darwāza, has a similar system of concentric keel arches for its squinches, as in earlier K̲h̲urāsānian work (cf. Ḳi̊rḳ Ḳi̊z near Termez), here carved, framed, and supporting an octagonal cornice; above, round the base of the smooth dome, are 32 arched niches, four of them pierced to admit light. The grace of the interior is achieved by a balance between the four main arches, the squinches, and at a reduced scale the miḥrāb and pairs of small arches at each corner, sustaining interest at each level. Each arch, inside or out, is contained by bands of inscriptions on the extrados (derived from Čis̲h̲t?), set off by lotus buds lining the intrados, in recessed planes above the angle shafts first introduced in Iltutmis̲h̲’s screen. The now-voussoired arch construction is masked by the carving. The lateral bays have two domes each on triangular pendentives, and may have been added rather later. Externally, the lateral bays are sunk, and the central one advanced and raised as a modest frontispiece; all are joined by a string course at mid-height and a lotus-bud parapet. Each archway is latticed. A provincial variant of the same style can be seen in the Ūk̲h̲a Masd̲j̲id at Bayāna, erected by Ḳuṭb al-Dīn Mubārak (716-20/1316-20). The mosques of the same period at D̲j̲ālōr, Dawlatābād, Pāt́an and Bharoč are built from temple spoil, but that at Dawlatābād continues the use of tapering, fluted corner buttresses, and Bharoč, with its more conscious blending of Hindu with Muslim elements, provides a starting point for the Gud̲j̲arātī style, with latticed windows, coffered ceilings over carefully-grouped columns, and domes of two sizes over the līwān. The D̲j̲āmiʿ Masd̲j̲id at K̲h̲ambāyat (ca. 1325) owes a more direct debt to Dihlī in its arches and massing, but local features are evident in the merlon parapet, pinnacles on the frontispiece, latticework set in a grid-like frame, and pillars carrying a cusped arch just inside the main archway. These examples attest to the diffusion of the style in western Hindūstān.

An altogether different treatment of the mosque was to characterise Tug̲h̲luḳ building. Most of the examples at Dihlī are undated, and have been ascribed to Fīrūz S̲h̲āh, but it has been suggested (Burton-Page, op. cit. in Bibl., 1974, 15) that the large Begampur Masd̲j̲id is better explained as built by Muḥammad b. Tug̲h̲luḳ for his new city of D̲j̲ahānpanāh (ca. 725/1325). Raised on a high plinth, it is important in introducing the Iranian four- īwān plan to India. North and south, the īwāns are advanced well into the court between heavy walls, boxing entrances at the centre of each side; to the east, the projection is outwards to a flight of steps, and to the west the tall arch rises to twice the roof height between tapering octagonal stair turrets, framing a triple entrance to the prayer hall. Here the main chamber is square, under a large pointed dome completely masked by this pīs̲h̲tāḳ. The hall on either side is three-aisled, with lesser domes, and 44 more domes cover the single riwāḳ all round the court, above arcades, and matching arched windows (for plan see ASIAR, iv [1871-2], pl. x). Muḥammad’s transfer of Dihlī’s population to Dawlatābād in 729/1329 appears to have depleted the skilled labour force and led to its dispersion elsewhere, notably in the Bahmanī Sultanate; southward expansion emptied the treasury. Nevertheless, the change of attitude introduced by Fīrūz S̲h̲āh (752-90/1351-88) was primarily an ethical one, in which his religious integrity required a return to prescribed simplicity and lack of ostentation. His building programme encompassed many mosques and 120 k̲h̲ānaḳāhs in Dihlī and Fīrūzābād alone, under the architect Malik G̲h̲āzī S̲h̲aḥna; given his stringent financial control, a modest but durable type of construction was inevitable. The fortified appearance of these mosques probably owes more to K̲h̲urāsānian prototypes, whose tapering round towers and massive walls had met the needs of mud construction, than to the need for defence (G̲h̲iyāt̲h̲ al-Dīn, a Ḳaraʾuna Turk, may have mediated this influence). The D̲j̲āmiʿ Masd̲j̲id at Fīrūzs̲h̲āh Kōt́lā (755/1354), now ruined, was built to incorporate a tahk̲h̲āna or undercroft, with arcaded vaults accessible from three sides, the east fronting the river. It once had three-aisled riwāḳs, with multiple domes, and 216 stone pillars about 16 ft. (4.87 m.) high, around a central octagonal pool with its own dome. To the north, one of the As̲h̲oka’s stone pillars was re-erected on a three-storey, arcaded pyramid as a marker. The materials for this and Fīrūz S̲h̲āh’s other mosques are rough rubble stonework faced with čūnā plaster, once whitewashed or painted, with a minimum of mouldings. The common répertoire included tall plain walls with merlons, plain lintels on plain, squared quartzite piers set in twos or fours, with elementary scrolled cross-brackets and capitals, still Hindu in type, and two-centred arches of variable width sunk in panels, sometimes concentric. Domes were of a similar, helmet-like profile, set on framed, recessed squinch arches. Externally, the mass is emphasised by long flights of steps, projecting porches, and battered towers at the angles. The device of the tahk̲h̲āna, which allowed the lease of shop spaces to sustain the mosque, is repeated at the Kalān Masd̲j̲id (798/1387?) which exhibits these features, and an unusual corridor around the prayer hall, besides cannon-like guldasta pinnacles crowning the angle-towers of the porch. The K̲h̲iŕkī Masd̲j̲id, also on a tahk̲h̲āna, repeats the three-aisled riwāḳ, but in combination with three-aisled passages which traverse the court on both axes, dividing it into four smaller square courts. This four-court plan is to be seen in a perhaps earlier form at the Sand̲j̲ar Masd̲j̲id (772/1370-1) at Niẓām al-Dīn, though there the riwāḳ and the passages are only one aisle deep, and the courts are rectangular. This scheme, possibly derived from D̲j̲ayn temple plans, was presumably intended to provide shade; the courts themselves were probably covered by awnings, as in palaces at the time. It intruded on the essential unity of the ṣaḥn and its congregation, and the experiment was not repeated. The mosque of S̲h̲āh ʿĀlam includes an early example of a mezzanine gallery in the northwest corner; the inaccessibility of such retreats leaves their purpose uncertain.

The D̲j̲āmiʿ Masd̲j̲id at Irič (815/1412), some 40 miles north of Jhansi, demonstrates the transition from the Tug̲h̲luḳ to the Sayyid manner. The plan, with single-aisled riwāḳs, is centred on a prayer chamber whose dome spans the full depth of the hall, with two aisles and six smaller domes on each side. The structure is wholly arcuate, on low piers carefully detailed to articulate both axes, with frequent use of recessed planes; the arches are now stilted, with marked corbelling at the impost giving a shouldered effect, and set in deep panels. The riwāḳ has groined vaulting. The dome is single, a little pointed inside, with ribs, and still set on concentric squinch arches. The generally ponderous effect is offset by the assured but simple proportions, and the skyline is relieved with merlons (see Mem. ASI, xix, Calcutta 1926, for drawings).

The Lōdī mosque (Tug̲h̲luḳid) at K̲h̲ayrpur (900/1494) incorporates similar features, while its massing shows the continuity of Tug̲h̲luḳid tradition despite Tīmūr’s incursion. Attached by a walled court to the Baŕā Gumbad, it is balanced by an arched structure opposite: a significant precedent for later tombs. An arcaded basement makes up the change in level at the rear, with tapering round buttresses at each rear corner, and at each angle of the projecting bay of the central miḥrāb, whose tops are alternately reeded below guldasta pinnacles; a Hindu window is corbelled out from the middle, and from either end wall. The hall has five bays; the three in the middle are domed, but the ends have low, flat vaults. The elevation reiterates the pattern, with three broad shouldered arches, and narrow ones at the extremities. As at Irič, the central pīs̲h̲tāḳ is raised a little, but here it is set between narrow, niched piers, and the outer two bays are united by the line of a čhad̲j̲d̲j̲ā. Like its dome, the central arch, thrice recessed, is a little higher than the others, and a muscular tension results from the contrast of line. The surfaces, worked outside and in with deeply cut plaster, vibrate with countless arabesques; each extrados is inscribed, and inscribed rosettes fill the spandrels. Inside, they enhance pendentive systems of oversailing lintels carved with muḳarnas [q.v.] niches. The vocabulary is further enlarged by blind merlon parapets, counterset trefoils around the octagonal dome bases (precursors of later foliation), and spreading lotus finials, mahāpadma. The development of this type is apparent in the Mōth kī Masd̲j̲id (ca. 911/1505), where the lateral domes are shifted to the end bays, in a much freer spacing. There they are supported on similar corbelled pendentives, as long used in Iran, while the central dome rests on squinch arches. The five façade arches are narrower, and a lancet window is added at each end. The pīs̲h̲tāḳ now encloses a lofty blind arch reacing the parapet, which frames the entrance arch below, and a window above, as anticipated in the miḥrāb at K̲h̲ayrpur. The two corner buttresses give way to polygonal towers, arcaded in two storeys. White marble is used to set off the red sandstone, with coloured tilework, notably on čhairī s at the courtyard corners, and painted carved plaster.

Despite his dissatisfaction with this style, Bābur appears to have secured little improvement at his mosques (932/1526) at Kābulī Bāg̲h̲, Pānipat and Sambhal, beyond introducing Tīmūrid squinch netting. Humāyūn, however, developed it further in the D̲j̲amālī Masd̲j̲id (943/1536) at Dihlī, in the same five-bay format. This only has one dome. The pīs̲h̲tāḳ is contained between engaged reeded shafts that anticipate the Mug̲h̲al use of minarets. The four-centred arches on either side are separated by large superimposed niches, which help to maintain the rhythm, and their haunches are slight. K̲h̲ald̲j̲ī lotus buds are re-introduced on the central intrados. The Masd̲j̲id-i Kuhna at the Purānā Ḳilʿa (ca. 1535-60?) shows further refinement. Each of its five arches is contained within a taller blind one, and that in a panel. The end bays, broken forward, resemble the D̲j̲amālī pīs̲h̲tāḳ, but the three middle ones are set deeper, with delicate angle shafts, and are proportionately taller. The fine ashlar incorporates the first geometric marble mosaic, after Tīmūrid models, and elaborate moulding profiles. Inside, the rippling recessed arches carry squinch arches below the prominent central dome, niched pendentives on either side, and arched cross ribs with vaulting at either end, again of a Tīmūrid type.

Regional developments. Bengal.

Remains from the early Muslim annexation are very limited. At Tribeni, an inscription framing the miḥrāb is dated 698/1298, but the mosque has been rebuilt, as has the Sālik mosque at Basīrhāt (705/1305). At Čhōt́a Pānd́uā [see pānd́uā, Čhōt́a] ruins of a large brick mosque include basalt Hindu columns supporting well-rounded, two-centred arches of a type that remained typical of Bengal, and miḥrābs with carved trefoil heads above ringed shafts, plainly derived from Hindu niches, though within diapered Muslim frames, and a kiosk-like minbar [q.v.]. It may have been the model for the huge Ādīna Masd̲j̲id at Ḥaḍrat Pānd́uā (776/1374-5)(154.70 × 87 m.), which has similar features. There the broad courtyard resembles that of the Great Mosque at Damascus in its proportions and the dominance of a maḳṣūra- like bay at the centre of the prayer hall, once vaulted over. This runs through the hall, with five arches leading to five aisles of 18 bays on either side, but the presence of a royal mezzanine in the north wing leaves its purpose in doubt. Triple-aisled riwāḳs surround the court behind plain, stone-faced arcades, each arch recessed once within a panel. The simple pillars support brick cross arches between which spherical pendentives of corbelled brick carry 378 identical low domes, punctuated only by the maḳṣūra. Outside, the ashlar wall is advanced and recessed in alternate vertical strips traversed by cornice and string course, each set off by an aedicule containing a cusped arch and lamp. Although never repeated at this scale (32 miḥrābs!) such treatment of detail was to inform most subsequent work. From the 9th/15th century onwards, mosques took a closed form in response to the wet climate, with the characteristically curved Bangālī eaves line, but still with the massive polygonal corner buttresses of the period. Thus the Čamkatta Masd̲j̲id at Gawŕ (ca. 880/1475?) has a single square chamber of brickwork surmounted by a single dome; it has single openings centred north and south, and three to the east giving on to a vaulted verandah running the full width, again with single doors to north and south, and three to the east. The piers between the arched openings carried aedicules set high, and glazed tilework. The Lattan Masd̲j̲id (880/1475-6) is similar, but with a more complete symmetry, having three openings to north and south, and three miḥrābs opposite the doors, three domes over the verandah, and intermediate “corner” buttresses; the central verandah dome has a roof with four curved eaves—a čawčāla. It was once tiled outside and in. The Gunmant Masd̲j̲id at Gawŕ (889/1484?) encloses four bays of three aisles, all domed, on either side of a central maḳṣūra, the stonework of whose vault is carved in relief. A further variant is illustrated by three mosques at Gawŕ. The Thāntipāra Masd̲j̲id (885/1480) is rectangular, enclosing five bays of two aisles, with a single line of four stone pillars to carry its ten domes. Fine terracotta reliefs fill the spandrels and the two registers of aedicules on the piers outside. At the Čhōt́a Sōnā Masd̲j̲id, built between 899/1493 and 925/1519, the plan is comparable, but of three aisles; its central bay is wider, and has three čawčāla roofs in lieu of domes. Its ashlar front is finely carved, and the dome was once gilded. The Baŕā Sōnā Masd̲j̲id (932/1526) combines eleven bays of three aisles with a verandah forming a further aisle down the front, facing an open quadrangle with arched gateways; the stone is remarkably plain. Such forms continued well into the Mug̲h̲al period, as seen in the Ḳuṭb S̲h̲āhī mosque at Ḥaḍrat Pānd́uā (990/1582).

D̲j̲awnpur. A mosque begun in 778/1376 by Fīrūz S̲h̲āh Tug̲h̲luḳ was completed under the independent S̲h̲arḳī sultans (811/1408); its name, At́āla Masd̲j̲id, apparently refers to the pylon-like pīs̲h̲tāḳ which was to become the dominant trait of subsequent buildings here (Sk. at́t́āla = “watch tower”, see Lehmann, op. cit., 23), exaggerating the great screen-arch at Dihlī. The four- īwan̄ plan is apparently derived from the Begampur Masd̲j̲id at Dihlī, though the īwān walls are reduced to massive spurs outside the enclosure, and those to north and south have domes carried on clustered columns, leaving the three-aisled riwāḳ unimpeded. Only the western īwān still boxes in space in the prayer hall, accessible through triple doors as before, but with biaxial symmetry, three arches on either side maintaining the continuity of the three prayer hall aisles; the frontal turrets are now resolved as square towers tapering five stories to accommodate the pīs̲h̲tāḳ arch, whose recessed tympanum is pierced in three registers to reveal the open air beyond. This pylon, used for giving the ad̲h̲ān, is echoed at 1/3 scale on either side in the līwān wings, and in the remaining īwāns outside the remarkable two-storey colonnade; it may have been suggested by the pierced archway of the S̲h̲ayk̲h̲ Bārha mosque at Ẓafarābād (711/1311), though its scale perhaps owes something to Pānd́uā [for further description, see d̲j̲awnpur ]. Tapering cylindrical turrets at the angles of the rear wall attest to Tug̲h̲luḳid influence. At the Lāl Darwāza Masd̲j̲id (ca. 852/1447), built on the same pattern, the structure behind the main īwān is still lighter, minimising obstruction of the prayer hall below the central dome, though mezzanines are set on either side; the absence of lateral domes, due to the smaller scale, leaves that at the centre uncluttered. The dome piers, with massive Hindu brackets, contrast oddly with the Iranic slenderness of the colonnades. In structural terms, the D̲j̲āmiʿ Masd̲j̲id (842/1438, but finished under Ḥusayn S̲h̲arḳī) is a reversion to the Begampur type, with boxed-in, domed īwāns on all four sides, and the same high undercroft. In the prayer chamber the colonnades are eliminated except under the mezzanines either side of the central chamber, where the pillars are paired to match its piers, for the wings are again boxed in by heavy masonry supporting the roof of a single pointed barrel vault spanning east and west, on either side. The prayer hall is thus divided into three spaces free of supports, but separated by their cross walls and the two-storey mezzanines. The same triality is seen in the façade. The simply niched towers and arcaded tympana of the earlier pīs̲h̲tāḳs are transmuted into a rhythmic display of framed and fretted openings. The dichotomy between high frontal screen and the dome hidden behind is nowhere more pronounced than here. Related mosques are to be found at It́āwā (D̲j̲āmiʿ Masd̲j̲id) and Banāras (Aŕhāʾī Kanguar).

Gud̲j̲arāt. In a sandstone architecture, drawing more than that of any other region on the Hindu and Jain traditions, two tendencies in mosque design had already emerged in the K̲h̲ald̲j̲ī phase already referred to: the screening of the prayer-hall front between a series of archways, as at K̲h̲ambāyat (after Niẓām al-Dīn at Dihlī), or the treatment of the hall as an open colonnade, given additional rhythm by the surge of domes above the čhad̲j̲d̲j̲a line, as at Bharoč. In either case the domes were carried by the Hindu device of beams spanning between two columns grouped to convert each square bay to an octagon. Remaining square bays were panelled in intricately recessed layers of coffering, whose cellular carving matched that of the domes. Pillars with markedly stratified round shafts above squared, faceted pedestals, carry vigorously curved brackets never far from living movement. The proportions of the three-arched screen are carefully repeated at D̲h̲ōlkā in the mosque of Hilāl K̲h̲ān Ḳāḍī (733/1333), but with bracketed, tiered pinnacles marking the pīs̲h̲tāḳ so prominently as to suggest the minarets which followed; the central dome, raised a storey above the roof, is surrounded by pierced screens. The same scheme, with its lower wings on either side, recurs at Aḥmadābād in Sayyid ʿĀlam’s mosque (815/1412), with half-rounded, tiered and bracketed buttresses framing the central arch as bases for fully functional mīnār s in a comparable style. The larger domes are now true, hemispherical ones. The development reaches fruition in the D̲j̲āmiʿ Masd̲j̲id at Aḥmadābād (826/1423) where the roof at the front of the three central bays is raised for a clerestory, with mezzanine galleries between, and the central dome is raised a further storey, so that light can enter indirectly at two levels, filtered by a pierced screen set in the usual Gud̲j̲arātī gridframe of stone: the remaining domes, three deep and five in the length of the hall, surround these three at the lower level. The mīnārs, once four times this height, fell in 1819 (see J. Forbes’ drawing of 1781 in ASWI, vii [1906], 30). The Masd̲j̲id of Malik ʿĀlam (1422?) combines a single arch with such minarets and an open front. Continuing interest in the open type of hall is seen, as at the mosque at Sarkhēd̲j̲ (855/1451), where 140 pillars, grouped as usual to support two rows of five equal domes, are set throughout in pairs to achieve an elegantly simple unity below a continuous roof line; there is little carving but for the miḥrāb s. The D̲j̲āmiʿ Masd̲j̲id at Čāmpānēr (Maḥmūdābād) (924/1518-9) works variations on that at Aḥmadābād. The eleven main domes are staggered, the central one being set over a single central bay rising through three roof levels, behind a pīs̲h̲tāḳ which now overlaps the mīnār on either side, and incorporates three corbelled bay windows. The hall wings (bāzūha) thus maintain a single roof line, with a plain walled front pierced by two arches each side, but there are now corner turrets to match the octagonal mīnārs. The main dome is ribbed inside, the side ones still corbelled, and the carved panels have filigree tendril-work. As at Aḥmadābād, the riwāḳ is one aisle deep; three entrance pavilions outside the wall carry prominent čhatris, and the wall itself is strongly modelled. The mosque of Rānī Rūpawātī (ca. 916/1510) shows a hall of only three domes treated similarly, with bay windows playing a more conspicuous role in modulating the front and ends. The culmination of the open hall design at the mosque of Rānī Siprī (Sabarī), also at Aḥmadābād (920/1514), fronting her tomb, has two rows of three corbelled domes, with only one row of pillars down the centre, and another, paired, in front, enlivened by alternate spacing. The extreme delicacy of this small-scale scheme is most evident in the slender, but solid and purely decorative mīnārs now set at each end of the façade—a device already introduced at the mosque of Muḥāfiẓ K̲h̲ān (897/1492) with full minarets. These two traditions were reconciled in the mosque of S̲h̲ayk̲h̲ Ḥasan Muḥammad Čis̲h̲tī (973/1565-6), a pillared hall of three miḥrābs in which the front is arcaded between terminal mīnārs, and the central five bays are raised in an upper storey of verandahs around a single dome. Sīdī Saʿīd al-Ḥabs̲h̲ī’s mosque (980/1572-3), still at Aḥmadābād, has five bays of three aisles with intersecting arches, supporting shallow domes over squinches, lintels and corbels, but is remarkable for its ten large tracery lunettes, of which two are unrivalled in the sinuous naturalism they bring to the interior.

Mālwā. An initial phase of redeployed temple material is distinguished by a simple grace which remained typical of the kingdom. At the D̲j̲āmiʿ Masd̲j̲id (or Lāt́ Masd̲j̲id) at D̲h̲ār (807/1404-5) the proportions of a single smooth hemispherical dome impart a spaciousness to the centre of the prayer hall colonnades, complemented by a pattern of flagstones, and a peaked, cusped miḥrāb arch; outside, its coronet of merlons enhances the traces of a tiled merlon parapet over the open front. One domed porch is surrounded by coved vaults, and in another false arch profiles are inserted between the pillars as in Gud̲j̲arātī temples. The first mosque at Mānd́ū, that of Dilāwar K̲h̲ān (808/1405-6) is spartan, however, with its hall of elemental columns relieved only by seven miḥrābs. Its successor, that of Malik Mug̲h̲īt̲h̲ (835/1432) presents a more Tug̲h̲luḳid exterior, with an arcaded undercroft in front between domed turrets, and the prominent stair often used here. The open, pillared prayer hall has three low, helmet-like domes. These, though still supported by an octagon of lintels, are partly enclosed by similar false arches below, with web spandrels, well integrated with the miḥrābs behind. The D̲j̲āmiʿ Masd̲j̲id (858/1454) has the same undercroft and steps, and the three main domes again span three rear aisles of the hall, but there are now two aisles in front of them, which with the triple aisles of the side riwāḳ s are covered with ranks of small domes, one to each bay, 158 in all. The building is mature, wholly Muslim, and of a sturdy dignity. The heaviness of strongly stilted domes is balanced by the grace of matching arcades round the court; the lofty hall is intersected by arches over plain, squared pillars, and articulated with blind wall arches and a characteristic flaring squinch. Each end dome covers a mezzanine set on nine bays of cross-vaulting. The pink stone is almost plain. The D̲j̲āmiʿ Masd̲j̲id at Čandērī is comparable, though remarkable for serpentine brackets developed from those of the minbar at Mānd́ū.

K̲h̲āndēs̲h̲. A similar restraint in the D̲j̲āmiʿ Masd̲j̲id at Burhānpur is conspicuous in its open hall front of 15 uniform arches, relieved only by a dancing alternation of large and small trefoil merlons, and the reiteration of čhad̲j̲d̲j̲ā brackets, the arcaded court appearing larger thereby (997/1589). The interior of the hall is equally regular, with five aisles of cross vaulting sustained by plain squared pillars decorated only on their bases, and a crested miḥrāb to each bay, rising above the string course, with three recessed arches finely chiselled in the dark stone. A substantial octagonal mīnār rises from a faceted square base at each end of the hall front, topped by a square lantern and a dome. Similar tall but plain mīnārs appear elsewhere in the city, and most notably as a pair flanking the pīs̲h̲tāḳ arch of the Bībī kī Masd̲j̲id, with four d̲j̲harōk̲h̲ā windows below their domes. Their tiered form otherwise resembles that at Čāmpānēr, there is even a d̲j̲harōk̲h̲ā on either side fronting the threedomed hall, whose organisation is apparently based on Rānī Rūpawatī’s mosque at Aḥmadābād (see ASI, NIS, ix, 1873-5).

×Bahmanī Sultanate. The interpretation of the līwān as a simple repetition of arched bays is already present in the S̲h̲āh Bazar Masd̲j̲id at Gulbargā (ca. 761/1360?), in an open-fronted hall of 15 bays of crossed arches in six aisles, all of them domed. The arches, set on tall piers, are recessed once, shouldered at the impost and stilted; the domes are low. At the D̲j̲āmiʿ Masd̲j̲id (769/1367, thus contemporary with the K̲h̲iŕkī and Sand̲j̲ar mosques at Dihlī) similar arches and squared piers are deployed quite differently to cover what would normally be the court with 63 domes on pendentives of corbelled work on angle. The riwāḳs are replaced by broader aisles roofed by rows of transverse pointed barrel vaults countering the thrust of these, with a large dome at each corner; these vaults rest on arches set on very low imposts, the contrasts in height adding interest to the interior, while light floods in from arcades in the outer wall. A still larger dome is set in front of the miḥrāb, heavily stilted, over trilobed squinches echoing the miḥrāb itself, and set in a square clerestory (cf. that in the mosque of Karīm al-Dīn at Bīd̲j̲āpur, 720/1320). The ensemble recalls bāzār architecture in Iran; it was without sequel, like the experiments at Dihlī. A variant of the arcaded open līwān at the Dargāh of Mud̲j̲arrad Kamāl (ca. 802/1400) has carved stucco archi volts and rosettes, with an extraordinary “entablature” of depressed cusped arches on sinuous brackets. The D̲j̲āmiʿ Masd̲j̲id (Solah K̲h̲amba) in the Fort at Bīdar [q.v.] is another version (827/1423-4), whose long front of 19 arched bays has square piers, and the five-aisle interior round pillars, carrying small domes on squinches. Heavy piers form a maḳṣūra enclosing the central three bays, from which squinches on sinuous brackets carry a tall 16-sided drum lit by fine d̲j̲ālīs, and a single large dome whose outer form is close to the domes at Mūltān [q.v.] while its supports recall the Tug̲h̲luḳid īwān at Begampur. The small three-bayed Langar kī Masd̲j̲id at Gulbargā (ca. 838/1435?) introduces a single pointed brick vault over two arched ribs.

Barīd S̲h̲āhī. At Bīdar, the use of tall arches on low imposts is resumed at the D̲j̲āmiʿ Masd̲j̲id (ca. 926/1520?), recessed once, with angular matching squinches articulated with great clarity below plain domes (cf. those in southern Iran). A transition to the Bīd̲j̲āpur vocabulary can be seen in the Kālī Masd̲j̲id (1106/1694-5), where the three front arches are framed by a pair of slender, formalised mīnārs, and the decagonal miḥrab recess is housed in a square rear tower carrying a čārṭāḳ lantern, and a slightly bulbous dome as introduced at the Madrasa of Maḥmūd Gawan (877/1472); a domical vault roofs the central bay. A small mosque at the tomb of ʿAlī Barīd (984/1576), handled similarly, has three domes on squinch-net pendentives, and a fretted cresting.

ʿImād S̲h̲āhī. The D̲j̲āmiʿ Masd̲j̲id at Gāwilgaŕh [q.v.], rebuilt in 893/1488, already combined a sevenarched hall façade on square piers with a square pylon at either end topped by a čhatrī with d̲j̲ālī -work in the sides, and čhad̲j̲d̲j̲a s on serpentine brackets, but otherwise follows the Bahmanī pattern of a dome over every bay, and a larger one raised on a tall drum at the centre; an arcaded screen wall surrounds its court. This is repeated at a smaller scale in the D̲j̲āmiʿ Masd̲j̲id at Rohankhed (990/1582), where four pylons with čhatrīs now form the hall ends, with a single central dome: the imposing south gateway has extensive carving.

Niẓām S̲h̲āhī. The Damrī Masd̲j̲id at Aḥmadnagar, small and precise, has a three-arched façade flanked by ornate pylons, which carry four graceful mīnārs capped by bud-like domelets. Octagonal pillars form two arched aisles supporting a flat roof. At the centre of a decorative parapet two slim minarets frame an arch profile, as in the Bādal Maḥall Darwāza at Čandērī. No superstructure remains on the corner piers of the Dilāwar K̲h̲ān mosque at Khed, but the exterior is enhanced by cusped arches, with two panelled bands running all round, and lotus medallions in relief. The central dome set on a square base imitates a tomb, complete with čhad̲j̲d̲j̲ās and corner čhatrīs. Inside, columns with volutes carry a coved ceiling.

ʿĀdil S̲h̲āhī. At Rāyčūr [q.v.] in the disputed Dōāb, a series of līwāns were built with flat ceilings over black basalt Čālukyan pillars whose short, heavy profiles are compensated by a deep parapet; the Ek Mīnār kī Masd̲j̲id has a tapering, free-standing mīnār 20 m. high (919/1513). In Bīd̲j̲āpūr [q.v.] the Bīdar vocabulary was elaborated in dark stone. Thus in the D̲j̲āmiʿ Masd̲j̲id of Yūsuf (918/1512-3) the slightly bulbous dome, set on a tall cylindrical drum, is familiar but for the foliation around its base, as is the dominance of the central arch, its form, and the articulation of line and squinch within; what is new, and characteristic, is the prominence given the dome, and the domed čārṭāḳ lanterns at each corner, well above the roof line. The same three-bay format is used in the D̲j̲āmiʿ Masd̲j̲id of Ibrāhīm (ca. 957/1550?), where a flat, domeless roof with sturdy domed guldasta pinnacles at each corner is relieved by a panelled mīnār set over each front pier. Cusped arches surround its miḥrāb. The mosque of Ik̲h̲lāṣ K̲h̲ān (ca. 968/1560?) is similar, with the addition of a lantern in two storeys above the miḥrāb, and a cusped central arch. All three arches are cusped, and repeatedly recessed, in the mosque of ʿAlī S̲h̲ahīd Pīr where a pointed vault (as at Gulbargā) runs parallel to the front, and a tall domed shaft rises over the miḥrāb. In all of these carved stucco decoration, notably rosettes, is prominent. A mosque in the fort at Naldrug (968/1560) may have one of the first double domes in India. At the D̲j̲āmiʿ Masd̲j̲id of Bīd̲j̲āpūr, the largest in the Deccan (985/1577-8?), these elements achieve mature expression. Its prayer hall, nine bays long and five aisles deep, is articulated with a calm strength, only an alternation of squinch detail varying a uniform structure with shallow domes; four piers at the centre are omitted, and intersecting pendentive arches are inserted in a miraculous change of scale to carry the dome (as already found in the tomb of Sulṭān Kalīm Allāh at Bīdar and based on Tīmūrid antecedents. Clerestory arches with fine d̲j̲ālī s light it through a square base rising above the roof, but the dome, still of the Mūltān shape above its foliation, remains dim, as usual here. Two features are innovations. At the east end of each seven-bayed riwāḳ is an octagonal base for an unbuilt mīnār; the entire external wall is modelled with two registers of arcading, the upper a corridor, and the lower blind. Both may be derived from the Muṣallā at Harāt (841/1437-8) [q.v.]. A central courtyard tank anticipates Mug̲h̲al practice. Stucco is partly replaced by carved stone at Malika D̲j̲ahān Bēgam’s mosque (ca. 995/1586-7), in which the dome now suggests a sphere in its collar of leaves, repeated at each stage of four corner minarets; guldasta lanterns, fretted cresting, and pendant stone chains compound a new elegance. The same character informs the Andā Masd̲j̲id (1017/1608) in fine ashlar, set back above a sarāʾī, with a gadrooned dome, and the mosque at the Mihtar-i Maḥall, domeless, with rod-like mīnārs, and four prolonged čhad̲j̲d̲j̲ā brackets engaged to the piers. Its acme is the mosque at the Ibrāhīm Rawḍa (1036/1626), facing the tomb across a plinth within a walled garden; brilliant use is made of elements repeated at a miniature scale to complement the whole. Afḍal K̲h̲an’s mosque (1064/1653) is on two floors, the upper probably for women, as at the Andā mosque. The style was taken as far south as Sante Bennur. Much of the extravagant ornament is discarded in the Makka Masd̲j̲id, in the latter half of the century, free-standing within a riwāḳ continued to the west.

Ḳuṭb S̲h̲āhī. At Golkond́ā [q.v.], the first capital, the ruins include a D̲j̲āmiʿ Masd̲j̲id built by Sulṭān Ḳulī Ḳuṭb al-Mulk in 924/1518 near the Bālā Ḥiṣār Darwāza. The regional achievement is best represented by the mosques at Ḥaydarābād [q.v.], which were given a new emphasis on height, accentuated by the concentration of external detail in the fascia between the čhad̲j̲d̲j̲ā and the skyline, and complemented by arcaded galleries around powerfully contoured mīnārs. The multiple guldastas on fretted parapets, and foliated bulbous domes are, like the stucco, inherited from Bīd̲j̲apūr. The D̲j̲āmiʿ Masd̲j̲id (1006/1597-8) has a spacious arched hall behind a front of seven bays divided unusually into two registers, the upper one of cusped arches being carried on struts from the pier imposts; the central arch, broader and taller than the others, is surmounted by a plain profile in the upper section. The Makka Masd̲j̲id, begun ca. 1026/1617, and continued until finished by Awrangzīb in 1105/1693, is set behind a square courtyard reputed to hold 10,000 worshippers, with a hall two aisles deep and five tall bays wide. In the plain ashlar façade, the central arch is slightly larger, as the only variation below the strong horizontal of a čhad̲j̲d̲j̲ā on linked brackets, spanning between the broad galleries of the turrets at either end, each of which is crowned by a bulbous dome on a marked necking. The columns carry arched pendentives and domes, with a coved central bay. Verticality is particularly pronounced in the Tolī Masd̲j̲id (1043/1633-4), where the five narrow arches of the front are stilted above impost blocks on the tall piers, and a tall parapet of arched screens joins the mīnār galleries for their full height; each shaft has two further galleries above roof level. Extensive use is made of cut plaster, syncretic in style. For other developments in the south, see mahisur. 2. Monuments.

Kas̲h̲mīr [q.v.]. The combination of a mountain climate and plentiful timber have resulted in a tradition of mosque building in a blockhouse technique of laid dewdār logs and pitched roofs with birchbark sarking topped by turf. In parallel with Dakhani mosques, the basic constructional unit had much in common with the local tomb type, a near-cubical volume set on a stone base, the corners emphasised by timber jointing, and roofed by a pyramid, sometimes tiered, with a slim spire at the centre. Frequent renewal after fires renders dating unreliable, though the type seems to have been used since the 8th/14th century. At S̲h̲rinagar in the mosque of S̲h̲āh Hamadān, the volume is modulated by large roofed balconies on each outside face, and the roof by a square arcaded muʾad̲h̲d̲h̲īn ’s gallery below the peaked spire. Four tapering octagonal columns support a painted ceiling, with small rooms ranged to north and south. Cusped round arches contrast with the rhythms of varying timber lattices and panelling. At the D̲j̲āmiʿ Masd̲j̲id (last built 1085/1674), a variant of the four- īwān plan places four of these units symmetrically around a square court, joined by four-aisled riwāḳs full of timber columns. Three form arched gateways, while the larger one to the west rises between walls of arched panelling over paired columns at the riwaḳ ends in an expansion of light and space, focussed on the simple arches of a large miḥrāb in a fenestrated wall. In this case the outer walls are of brick with a simple repeated window, contrasting with the four spired roofs. In Baltistān and Kuhistān simple open līwān s of one or two aisles are supported on wooden columns, often fluted above a waisted base, and with brackets carved in repeated waves supporting beams on the long axis; here the connection with Turkestan building is evident.

Mug̲h̲al Empire. During Akbar’s minority, the Tīmūrid innovations introduced under Humāyūn remained in currency, associated with the harem faction, as in the mosque and madrasa of Māham Anaga (Angā), the K̲h̲ayr al-Manāzil (969/1561-2) whose three bays to the court are close in format to the central three at Purānā Ḳilʿa with a slightly raised pīs̲h̲tāḳ advanced between clustered shafts, and four-centred arches whose tympana are pierced with archways at a lower level; only the single dome has an awkward, old-fashioned stilt. The arch spandrels are inlaid. The screened upper storey of rooms enclosing the court on three sides appears to be unique for the period, while the portal is the first to use a semidomed īwān. At Fatḥpur Sīkrī [q.v.] these forms are less in evidence. Although the front of the Stonecutters’ Mosque (ca. 973/1565) is arched, originally in five bays, the arch profile is cut from thin slabs set between thicker posts, the čhad̲j̲d̲j̲ā is supported by long, sinuous brackets, and the internal row of pillars is Hindū. The organisation of the great D̲j̲āmiʿ Masd̲j̲id (979/1571-2) stems from D̲j̲awnpur via Bayānā, where the technique of assembling cut stone components was already welldeveloped a century earlier (fieldwork by Shokoohy 1981). Three domed spaces at the centre and amid either wing of the līwān are each contained within massive walls pierced by symmetric arches to communicate with the columned spaces between, where flat, beamed roofs are supported on Hindū brackets, all in red sandstone; the central dome set on squinch arches is painted with swirling floral patterns, and the lateral ones are ribbed, lit through the drum, and carried on corbelled pendentives. The front of the hall with its alternation of broad and narrows bays, thin spandrels, long čhad̲j̲d̲j̲ās, and the form of the pillars appears to be Gud̲j̲arātī in origin, as does the great tank under the courtyard. At the centre, however, is a great pīs̲h̲tāḳ of the Dihlī type, with a semi-dome, completely screening the stilted and lumpish dome behind. The wings are of half the height, and relieved by queues of little čhatrī s along the skyline, like the riwāḳ s with their central īwāns: these once served as lanterns. Although the awkward column-spacing under the lateral domes of the At́āla Masd̲j̲id has been resolved, and much is made of the three main spaces, their walls still interrupt the unity of the hall.

The Mosque of Maryam Zamānī (1023/1614) at Lāhawr [q.v.], known as the Bēgam S̲h̲āhī Masd̲j̲id, and built of brick following local practice, achieves an unencumbered prayer hall of five square, domed compartments in line, interconnected by single arches springing from heavy piers at front and rear. The central compartment is wider, with a larger dome than the others, still stilted, but housing an inner shell which, though only of plaster, was probably the first used in a mosque in the north. The new arch shape extends to the squinches, with muḳarnas semi-domes, and the domes are articulated with netting, the whole being elaborately painted with floral, geometric, and inscriptional designs. Outside, the līwān front follows the model of the D̲j̲amālī Masd̲j̲id, with blind superimposed niches on the pier faces, but the arches are now simple in profile, the front is in one plane but for the vaulted īwān, and there are square, domed turrets at either end. The Masd̲j̲id-i Wazīr K̲h̲ān (1044/1634-5) in the same city has a līwān of the same kind, both outside and in, as before punctuated by a miḥrāb below a semidome in each bay, with pendentives rising to carry the inner dome shells in the wings, and squinches at the centre. The dome profile is lower, with minimal stilting, but still unlike the profile of the five arches. The turrets are here full-sized octagonal minars with čhatrīs above the galleries, and are echoed by a second pair at the east of a long court. The brickwork forms shallow panels between orthogonal fillets, containing a sumptuous variety of tile mosaic; the interior is painted.

A series of court mosques faced entirely in white marble—seen as “pure like the heart of the austere” (Bāds̲h̲āh-nāma, ii/1, 155)—was probably initiated at Agra [q.v.] with the tiny, perfectly simple Mīnā Masd̲j̲id and the larger, three-bayed Nagīna Masd̲j̲id within the Fort. The latter, in which the lower dome profile has been transformed by necking above a torus moulding into a smooth bulbous shape with a large pointed mahāpadma (Bīd̲j̲āpūrī influence is suggested by the crescent above), represents an attempt to eliminate the conflict between emphasis on the central bay, and that on the dome behind, by replacement of the pīs̲h̲tāḳ with an upward curve of the čhad̲j̲d̲j̲ā and parapet, in the new Bangālī fashion, at the middle. This accommodates the larger central arch; the arches are engrailed, probably to reduce glare when viewed from inside. In the mosque at the Tād̲j̲ Maḥall [q.v.], the same conflict is resolved by raising the level of the façade over the two lateral arches almost to pīs̲h̲tāḳ level, and including a blind arched panel above each. This scheme is repeated at Lāhawr in the mosque of Dāʾī Angā (1045/1635-6), the corner turrets containing the taller front as before; the side arches are surmounted by great cusped arch heads, and the Lāhawri panelling is of tile mosaic inside and outside the three interpenetrating square compartments. The treatment of the Madrasa Masd̲j̲id at Pat́nā (ca. 1040/1630) is comparable. The Fatḥpurī Masd̲j̲id at Āgra, flanked by the same flaring turrets, has a fully bulbous dome, but a tall marble pīs̲h̲tāḳ in front over a deep īwān, and low wings; its red stone is finely worked in relief, notably in the pendentives and inner dome. Like it, the Mōtī Masd̲j̲id at Lāhawr (ca. 1055/1645) is fronted by cusped arches flanking a plain central one, but it offers a further solution to the problem with a barely raised pīs̲h̲tāḳ linked to the wings by a continuous parapet in parčīn-kārī. The three marble domes still have the cavetto and profile of Dāʾī Angā’s mosque, now clearly visible. These smaller mosques owe much to the consonant detail of arcuate screens which separate their courts from the outside world, and a finesse that extends to sad̲j̲d̲j̲āda inlaid in the floor. On a larger scale, the S̲h̲āh D̲j̲ahānī Masd̲j̲id at Ad̲j̲mēr (1048/1638-9), with a prayer hall two aisles deep with arched piers, presents a long, unbroken façade of eleven bays, accented only by a needle-like guldasta over each octagonal column, to a balustrated court adjoining the dargāh of Muʿīn al-Dīn Čis̲h̲tī; the whole is in marble.

Some of these tendencies are resolved at the D̲j̲āmiʿ Masd̲j̲id at Āgra, completed in red stone in 1058/1648. Its plan is essentially that of the fivecompartment prayer hall from Lāhawr, complete with its corner turrets and another pair at the east corners of the court. Its capacity is increased by the addition of a second row of compartments in front of the first, the central one forming a deep īwān, whose pīs̲h̲tāḳ is thus spaced well forward from the domes over the main row behind; the two lateral domes are placed over the ends, for better balance, and all three are double and distinctly bulbous, with a pointed profile accentuated by inlaid chevrons of white marble (structural inner domes were from henceforward the norm). The front is of the tall type, with panels above and between the well-spaced plain arches, and two prominent shafts frame the marble pīs̲h̲tāḳ. Čhatrī s enliven the whole skyline. The interior is a smooth progression of netted pendentives and plain arches with a broad extrados, at a noble scale. Its equivalent at Dihlī (1066/1656), also raised on a high podium, and approached by three great pyramids of steps on the axes, is the largest enclosed mosque in northern India. Gateway īwāns on these axes regain their prominence, and the riwāḳ s are open to the external air on all three sides. A collision between these and the līwān, a weakness at Āgra, is avoided by returning them along the west, and then advancing the hall forward between full-size minarets at the corners. The līwān plan fuses those of Āgra and Fatḥpur Sīkrī, with alternating main compartments, and slimmer piers at the front; cusped arches are used throughout. The domes, now on tall drums are, like the mīnār s and the īwān, striped with marble inlay, and the entire front is panelled in marble, with plain merlons above. Such detail, and especially the marble calyces topping the angle shafts, introduce a mannered deviation from the former simplicity. The scale is such that the īwān itself forms a miḥrāb to the courtyard.

The Mōtī Masd̲j̲id at Āgra Fort (1063/1653), the largest of the marble series, complete with riwāḳ and axial gateways, combines a restraint of outline and of plan with an extravagance in the intersecting, cusped arch profiles. Eighteen identical piers in three aisles carry plain coved ceilings alternating with three domes on smooth pendentives, that rise bulbous among the čhatrī s outside. That in the Dihlī Fort (1074/1663-4) shows the full extent of the stylistic change at a small scale, with a Bangālī curve in the čhad̲j̲d̲j̲ā over the central bay, set off by Bangālī vaults within, reticulated coving, clustered guldasta s with calyces, and floral relief playing on many surfaces; the domes, rebuilt after the Mutiny, were originally lower, and gilded.

The last of the great congregational mosques, the Bāds̲h̲āhī Masd̲j̲id at Lāhawr (1084/1673-4) derives its plan almost entirely from the great mosque at Dihlī, the principal differences being that the three-storey octagonal mīnārs are now set at the four corners of the court, and the līwān itself reverts to the local scheme with a domed octagonal turret at each corner. The riwāḳs, too, are subdivided into an alternating series of hud̲j̲ra s for teaching, accessible only through doorways, and though raised as before, the court is thus closed in. The līwān, of brick faced with red stone, is rather taller than at Dihlī, and panelled in the local manner, but the surfaces swarm with relief carving; the marble domes formerly had dark drums to relate them to the wings. Internally the squinched dome chambers alternate with Bangālī vaults, and the walls, arch soffits and domes are panelled or worked in netpatterns, islīm-i k̲h̲aṭāʾī, of plaster relief, or else painted. The mosque is claimed as the largest in the world. The gateways of such structures served to house the imām and other staff. The Sonahrī Masd̲j̲id at Dihlī (1164/1751) repeats the Motī Masd̲j̲id at the Fort in fawn sandstone. In subsequent work in Awadh the curvilinear and vegetal elements were to become dominant [see lak̲h̲naw ], and were still vigorous in the D̲j̲āmiʿ Masd̲j̲id of ca. 1840 in the capital.

Provincial developments within the Mug̲h̲al empire predictably show an adaptation of the court style to local practice. In Bengal, the mosque of the Lālbāg̲h̲ Fort at D́hākā (1089/1678) has the closed appearance and panelled front typical of the area, but the height of the prayer hall, its three cusped and netted front īwāns, its three low domes and the four octagonal turrets at its corners all refer to the experience of Lāhawr. The interior of the lateral bays is remarkable for semidomes set below the apical dome, with two sets of pendentives. Other mosques at D́hākā follow the same format, as in that of K̲h̲ān Muḥammad Mird̲h̲a (1118/1706), with tall minars at the līwān corners, or the Sātgunbad̲h̲ mosque with octagonal corner towers.

The brick architecture of Sind is extensively clad in fine glazed tilework, owing much to Iranian influence, and apparently that of Harāt [q.v.] in particular. This is already apparent in the Dābgīr Masd̲j̲id at T́hat́t́ā (997/1588-9), of which the līwān remains in a ruined state, containing a square central compartment flanked by a rectangular one at each side, with arches connecting them between massive piers, and three deep īwāns, set in slightly raised pīs̲h̲tāḳs. The central dome, like the īwān below it, is notably larger than those either side, but all three are set on double octagonal drums of an Iranian type. The walls of the central compartment each house one well-shaped arch within another; at the west the interval contains an arched window set on either side of the buff carved sandstone miḥrāb. The tilework, floral, geometric and calligraphic, in cobalt and azure on a white ground, filled arch spandrels and soffits. The D̲j̲āmiʿ Masd̲j̲id of S̲h̲āhd̲j̲ahān (1057/1647) in the same city is unusual in plan, with repeated heavy piers forming the two aisles of the broad riwāḳs, and the three of the prayer hall, around a very deep court, focussed on a great pīs̲h̲tāḳ, with small subsidiary courts on each side of an east entry passing under two domes in series (cf. the Masd̲j̲id-i D̲j̲āmiʿ at Kirmān). The multiple bays are roofed by 80 small domes, with larger single ones over the central īwāns, backed to the west by a single shell dome replacing four bays in front of the miḥrāb’, this rises from intersecting pendentive work over a zone of 16 arches, pierced for a clerestory at the angles, and tiled throughout in mosaic (more than 100 pieces per sq. ft.) in ranks of wheeling stars. The smallest sound at the miḥrāb can be heard throughout the mosque, perhaps by virtue of its domes. In both these mosques the red brick is defined by white pointing which accents the arches. Further excellent tilework at the D̲j̲āmiʿ Masd̲j̲id of K̲h̲udābād has been badly damaged. The treatment of its façade shows stronger Lāhawrī influence in proportions and panelling; the external walls, however, are noteworthy for three superimposed registers of repeated blind arches, a few being pierced at the lower levels.

At Aḥmadābād, the mosque of Nawwāb Sardār K̲h̲ān (ca. 1070/1660?) combines a relatively orthodox Gud̲j̲arātī treatment of a three-bayed līwān, having three plain arches between narrow piers, a d̲j̲harōk̲h̲ā bay on each end wall, and balconied mīnārs framing the front, with features that seem to bridge the styles of Bīd̲j̲āpūr and Āgra. The three closely-spaced domes are bulbous, above torus mouldings, with steep mahāpadmas as in the Nagīna and Mōtī Masd̲j̲ids. The mīnārs, however, carry long foliations, lotus buds and the elongated, bulbed fmial of the later ʿĀdil S̲h̲āhī style, close to those at the similar and contemporary Mosque of Afḍal K̲h̲ān in the Dargāh of Gīsū Darāz at Gulbargā. The mosque, unlike its counterparts, is of brick and stucco. The mosque of Nawwāb S̲h̲ad̲j̲āʿat K̲h̲ān (1107/1695-6) has a five-arched front, with Gud̲j̲arātī merlons, and mīnārs placed to contain the central three bays, but the piers are panelled with rows of little niches, and a line of cartouches runs overhead, with three low domes of the Dāʾī Angā type; the mīnārs once more have foliations, but have lost their tops. In its ceiling, the domes jalternate with coved bays, as in the Mōtī Masd̲j̲id at Āgra, and it is finished with marble and polished plaster.

In general, it may be seen that whereas the enclosure of the court only achieves full architectural expression in cathedral mosques, or the later court mosques, the prayer hall is the subject of consistent architectural development. The particular structural means adopted in each region for enclosing the space become the vocabulary for a series of variations which in most cases go far beyond the immediate needs of the liturgy or of mere shelter, and can be recognised as successive resolutions of the need for balance, harmony, and unity at the chosen scale.

Bibliography

For general works, see hind. vii. Architecture. To these may be added Z. Desai, Indo-Islamic architecture. N. Delhi 1970

J. D. Hoag, Islamic architecture, New York 1977, 280-307, 364-88

R. A. Jairazbhoy, An outline of Islamic architecture, Bombay etc. 1972

B. Gray, ed., The arts of India, Oxford 1981. Works on mosques in general include E. La Roche, Indische Baukunst, II. Teil: Moscheen und Grabmäler, Munich 1921

Z. Desai, The mosques of India, N. Delhi 1971

and Y. K. Bukhari, The mosque architecture of the Mug̲h̲als, in Indo-Iranica ix/2 (1956), 67-75. For the early Arab mosques, see H. Cousens, The antiquities of Sind, Calcutta 1929, repr. Karachi 1975, 48 ff., and S. M. Ashfaque, The grand mosque of Banbhore, in Pakistan Archaeology, vi/1 (1969), 182-209. The Kuwwat al-Islām is still best described by J. A. Page in A guide to the Quṭb, Delhi, Delhi 1938, but see also M. C. Joshi, Some Nagari inscriptions on the Qutb Minar, in Medieval India—a miscellany, ii, Aligarh Muslim University 1972, 3-7, and S. K. Bannerji, The Qutb Minar: its architecture and history, in Jnal. of the United Provinces Historical Soc., x/1 (1937), 38-58. The Aŕhāʾī-din-kā D̲j̲hōńpŕā is described by M. Meister, The two-and-a-half-day mosque, in Oriental Art, xviii (1972), 57-63, and plan in ASIAR, ii, 1864-5

and see R. Hillenbrand, Political symbolism in early Indo-Islamic mosque architecture: the case of Ajmīr, in Iran, JBIPS, xxvi (1988), 105-18. For the mosque at Badāʾūn, see J. F. Blakiston, The Jami Masjid at Badaun and other buildings in the United Provinces, MASI, xix, Calcutta 1926, and A. Cunning̲h̲am, ASI, xi, 1880. The D̲j̲amāʿat K̲h̲āna is given a brief description by M. Zafar Ḥasan in A guide to Niẓāmu-d Dīn, MASI, x, 1922, 14-6. The early mosques at Pātan are described by J. Burgess and H. Cousens in Architectural antiquities of northern Gujarat, ASI, NIS, xxxii, London 1903, and those of Bharoč and Khambayāt in Burgess’s On the Muhammadan architecture of Bharoch, Cambay, Dholka, Champanir and Mahmudabad in Gujarat, ASI, NIS, xxiii, London 1896 (these two vols, being ASWI, ix and vi). The mosque at Fīrūzs̲h̲āh Kōt́lā is treated somewhat inadequately by Page in A memoir on Kotla Firoz Shah, Delhi (= MASI, lii) Delhi 1937

a few details and a plan of the Sand̲j̲ar Mosque are in M. Zafar Hasan, op. cit., 35-6. See J. Burton-Page, Indo-Islamic architecture: a commentary on some false assumptions, in AARP, vi (Dec. 1974), 15, for the Begampur Masd̲j̲id. The mosque at Irič is well illustrated in Blakiston, op. cit. pls. xxiiv, but without description. For the identification of the K̲h̲ayrpur mosque, see S. Digby, The tomb of Buhlūl Lōdī, in BSOAS, xxxviii (1975), 550-61. Sultanate work in general is admirably illustrated with measured drawings in T. Yamamoto, M. Ara, and T. Tsukinowa, Architectural remains of the Delhi Sultanate period (text in Japanese), 3 vols. Tokyo 1967. For Bengal, refer to A. H. Dani, Muslim architecture in Bengal, Dacca 1961

S. M. Ḥasan, Muslim monuments of Bangladesh, Dacca 1980

idem, Mosque architecture of pre-Mughal Bengal, Dacca 1979, and his unpublished Ph.D. thesis Development of mosque architecture with special reference to pre-Mughal Bengal, 2 vols., University of London 1965

some useful photographs and plans in Marg, xxviii/2 (March 1972). D̲j̲awnpur: A. Führer and E. Smith’s The Sharqi architecture of Jaunpur, ASI NIS, i, Calcutta 1889, is still unsurpassed

also F. Lehmann, The name and origin of the Aṭāla Masjid, Jaunpur, in IC, lii/1 (1978), 19-27

Cunningham, ASI, xi. Gud̲j̲arāt: besides the volumes by Burgess and Cousens given above, see Burgess, The Muhammadan architecture of Ahmadabad, 2 vols., ASI, NIS, xxiv, xxiii (= ASWI, viii and vii), London 1905, 1906, with excellent drawings and photographs. Mālwā: E. Barnes, Mandu and Dhar, ARASI, 1903-4, Calcutta 1906

G. Yazdani, Mandū, the city of joy, Oxford 1929

D. R. Patil, Mandu, N. Delhi (ASI) 1975, the current official guidebook

M. B. Garde, Guide to Chanderi, Gwalior 1928. K̲h̲āndēs̲h̲: no adequate sources other than general works, and Cunningham, ASI, ix, Central Provinces, 1873-5. Dakhanī architecture is reviewed by E. S. Merklinger, Indian Islamic architecture, the Deccan, 1347-1686, Warminster 1981, with a chronological catalogue and useful thematic treatment. For Bahmanī buildings, see G. Yazdani, The Great Mosque of Gulbarga, in IC, ii (1928), 14-21, and idem, Bidar, its history and monuments, Oxford 1948, with full descriptions and good plans. Barīd S̲h̲āhī: ibid. ʿImād S̲h̲āhī: see bibl. to gāwilgaŕh. Niẓām S̲h̲āhī: see bibl. s.v. ʿĀdil S̲h̲āhī: excellent coverage in Cousens, Bījāpūr and its architectural remains, ASI, NIS, xxxvii, Bombay 1916. Ḳuṭb S̲h̲āhī: for Golkond́a, s.v.

for Ḥaydarābād, see the Annual reports of the Archaeological Department, Hyderabad for 1916-17, 3 ff. and pls. ii-iii

for 1924-5, 2-4 and pls. iii-vi

for 1936-7, 2 ff. Kas̲h̲mir: W. H. Nicholls, Muhammadan architecture in Kas̲h̲mir, in ARASI, 1906-7, 164-70. Mug̲h̲al: for the K̲h̲ayr al-Manāzil, see ASI, NIS, xxii, 6 and pl. i, ARASI, 1903-4, 25-6 and pl. x-xi

for Fatḥpur Sīkrī, E. W. Smith’s superb Mughal architecture of Fatḥpūr-Sīkrī, ASI, NIS, xviii, Allāhābād 1894-7, now augmented by the intelligent discussion in S. A. A. Rizvi and V. J. A. Flynn, Fatḥpur-Sīkrī, Bombay 1975 (but beware error in description of domes, p. 74a). For Lāhawr, see Aḥmad Nabi K̲h̲an, Maryam Zamani Mosque, Lahore-Karachi 1972, and M. A. Chaghatai, The Wazir Khan Mosque, Lahore, Lahore 1975, with good plans, sections, and photographs

also J. Burton-Page, Wazir Khan’s mosque, in Splendours of the East, ed. Mortimer Wheeler, London etc. 1965, 94-101. For Āgra, see Nūr Bak̲h̲s̲h̲, The Agra Fort and its buildings, in ARASI, 1903-4, 185, and M. A. Husain, Agra Fort, N. Delhi 1956, 22 and 27

also R. Nath, Agra and its monumental glory, Bombay 1977, 35-8, and idem, The immortal Taj Mahal, Bombay 1972

The Fatḥpurī Masd̲j̲id is illustrated in ARASI, 1902-3, pl. xii. For the Dāʾī Angā mosque, see W. H. Nicholls, in ARASI, 1904-5, 20-22 and pl. iv

and for the Mōtī Masd̲j̲id, idem in ARASI, 1903-4, 26-7 and pl. xii-xiii, and M. Z. Ḥasan, Moti Masjid or the Pearl Mosque in the Lahore Fort, in Proceedings of the Pakistan History Conference, 2nd session, Lahore 1952, 8-16. For the Mōtī Masd̲j̲id at Āgra, see Nūr Bak̲h̲s̲h̲. op. cit., 181-4, and Nat̲h̲, op. cit. (1977), 39-42

for that at Dihlī, see G. Sanderson, A guide to the buildings, Delhi Fort, Delhi 1937, 55-7, and idem in ARASI, 1911-12, 13. Chaghatai gives comprehensive treatment to The Badshahi Masjid, history and architecture, Lahore 1972, withuseful comparative plans of the d̲j̲āmiʿ masd̲j̲ids at Āgra and Dihlī. The later mosques in Sind are reviewed in Cousens, The antiquities of Sind, ed. cit., though not in full detail, and by M. A. Ghafur in Muslim architecture in Sind area, Karachi 1961

for those in Aḥmadābād, see Burgess, op. cit. (ASWI, viii, 1905). Epigraphy is mainly available in EIM

a useful index is provided by V. S. Bendrey, A study of Muslim inscriptions, Bombay 1942, for 1907-38, supplementing J. Horowitz, A list of the published Mohammedan inscriptions of India in EIM, 1909-10, 30-144. Much material relating to mosques in Pakistan is set out by S. Mahmood in Islamic inscriptions in Pakistani architecture to 1707, unpubl. Ph.D. thesis, Edinburgh University 1981. Some of the principal mosque inscriptions are also quoted at length by R. Nath in Calligraphic art in Mughal architecture, Calcutta 1979.

(P.A. Andrews)

III. In Java.

In Java, the Arabic form masd̲j̲id is practically limited to religious circles. The Indonesian languages have developed the derivatives mesigit (Javanese, in Central and East Java), masigit (Sundanese, in West Java) and masèghit (Madurese, on the island of Madura and in part of East Java). In general, these terms are used only for the mosques in which on the Friday ṣalāt al-d̲j̲umʿa is held. Smaller mosques serving for the daily cult and religious instruction alone, are called langgar (Javanese), tadjug (Sundanese) and balé (in Bantěn).

Indonesian Islam has produced its own type of mosque, clearly to be distinguished from that of other Islamic countries. Since this type was probably first developed in Java, it can be termed the Javanese type of mosque. Its standard characteristics are the following: (1) The ground plan is a square one. (2) The massive foundations are raised. The Friday mosque is not built on piles, as is the case with the classical Indonesian houses and the smaller mosques mentioned above. (3) The roof is tapering, and consists of two to five storeys narrowing towards the top (4) An extension on the western or north-western side serves as miḥrāb [q.v.]. (5) At the front—sometimes also at the two lateral sides—is an open or closed veranda. (6). The courtyard around the mosque is surrounded with a stone wall with one or more gates. Another characteristic is that in Java the mosque stands on the west side of the alun-alun, the grass-covered square which is found in virtually all chief towns of regencies and districts. In Tjeribon, Indramayn, Madjalèngka and Tjiamis—all regions in West Java—even each dèsa has an alun-alun with a mosque at its west side.

In Java, the direction of the ḳibla [q.v.], is, however, not west but north-west, and so, in order to indicate the exact ḳibla, the miḥrāb or niche is sometimes built obliquely against the back wall. There are, however, also regions, like the Priangan, where the exact ḳibla is taken into consideration at the time of construction of the mosque.

The gate at the front which gives access to the courtyard surrounding the mosque is sometimes covered. The mosques of Central and East Java are characterised by their monumental entrance gates.

The veranda (Javanese: surambi, sěrambi, srambi ; Sundanese: těpas masd̲j̲id, těpas masigit) is not considered as belonging to the mosque itself, as is evident from the various purposes which it serves. It is the place where, at night, after the mosque has been closed, the ṣalāt is performed; where travellers and other people who have no home pass the night; where marriages are concluded; where in former times (see Raffles, The history of Java) religious courts were functioning; where sometimes religious instruction is given and where riyalat (Javanese; in Arabic riyāḍa = ascetic abstinence from sleep, food and sexual intercourse) is practised. It is also the place for religious meals (walīma) on feast days like Mawlid al-nabī and Miʿrād̲j̲ [q.vv.].

The walls of the mosque itself are rather low, but the roof tapers and ends in a sphere, on top of which is an ornament, called mastaka or mustaka in those regions where Javanese is spoken. It later times, this ornament was crowned by a crescent as the decisive symbol of Islam. This type of roof, in fact a piling-up of ever-smaller roofs, dates from pre-Islamic times and recalls the měru on Bali. In the present century, the cupola-shaped roof (Ar. ḳubba, [q.v.]), an imitation of the mosques in other Islamic countries, and in particular India, is competing with the traditional piledup roof of ancient Indonesia. Already before its restoration in 1935, the Masd̲j̲id Kemayoran in Surabaya diverged from the usual architectural pattern in that its base was not square but octagonal. In that year, two ḳubbas were constructed to the left and the right of the veranda. Another ḳubba, was added to the monumental minaret, which is said to be an imitation of the Ḳuṭb Minār in Dihlī [q.v.]. At the same period, the ḳubba, was also introduced into West Java. The use of the cupola-shaped roof became firmly established after Indonesia’s independence in 1949. Impressive, huge mosques, all with ḳubbas, have been constructed since that time. The Masd̲j̲id al-S̲h̲uhadāʾ in Yogyakarta and the Masd̲j̲id Istiḳlāl in Jakarta can be considered as examples of a new type of architecture applied to the mosque.

The interior of a mosque built in the ancient Indonesian style can be described as a closed hall, sometimes provided with pillars, of a sober character, reflecting the simplicity which is the characteristic of the masd̲j̲ids in Java. There are no pictures of man or animal on the walls, only sacred Arabic names and some religious texts like the s̲h̲ahāda [q.v.] and the ḥadīt̲h̲ in which the builder of a mosque is praised: “Allāh has built a house in Paradise for whoever has built a mosque for Allāh”. Since the floor of the mosque has to be clean, it consists of cement, tiles or marble. The grey colour of cement is occasionally alternated with rows of red tiles, indicating the rows (Arabic ṣaff) of the faithful when performing the ṣalāt. Mats are usually spread on the floor. In mosques which have not been constructed in the exact direction of the ḳibla, these mats are laid out in the right direction. Regular mosque-goers have their own small mat or rug (Ar. sad̲j̲d̲j̲āda). preferably one brought back by pilgrims to Mecca.

The miḥrāb at the rear side of the mosque is usually rather narrow, consisting of a small gate with a round arch. Sometimes the niche, or rather the extension, is large enough to contain the minbar on the right side. There are, however, also mosques with two or even three niches next to each other, each provided with a small gate. Occasionally, the miḥrāb is built out into a large pentagon with the minbar in the centre and the place of the imām to the left, the front side being fenced off by a wooden fencing with green and yellow sheets of glass and decorated with religious texts. Sometimes the miḥrāb is built out into a large, square place with the minbar in the centre, the place of the imām for the daily ṣalāt to the left, and to the right a small movable construction with an open front, this being the place of the regent of the region. The minbar (Javanese and Sundanese: mimbar, Javanese and Sundanese of Bantěn: imbar) is always found to the right of the mihrāb. Unlike other Islamic countries where the minbar is reached by a high flight of stairs, the minbar in Java is rather low. The height may vary from one to five steps, three steps being the average. Some minbars are very simple, but many others are conspicious for their woodcarving. As Islam permits, decorations consist of plants and flowers which sometimes look like pictures of men and animals. On closer inspection however, they prove to be representations of flowers and leaves of the lotus, arranged as wings and birds. Sometimes the naga (serpent) motive can be recognised on the arms of the minbar, as is the case in the holy mosque of Demak in Central Java and in the ancient, holy mosque of Kuṭa Deḍé in the same region.

Each mosque in Java possesses a drum, called běḍug, stretched with buffalo-skin. Before the ad̲h̲ān [q.v.] (Javanese and Sundanese: adan) this drum is beaten vigorously at least five times a day. The ad̲h̲ān itself is made either from the minaret (Javanese: měnara, Sundanese: munara) or, more often, in the mosque itself since not every mosque has its minaret. The muʾad̲h̲d̲h̲in. called modin or bilal, stands at the entrance of the mosque or on its roof.

The highest official of the mosque is the panghulu (thus in Sundanese; Javanese: pangulu ; Madurese: pangòlò, pangòlòh ; Malay, peng̲h̲ulu), often a learned man (Ar. ʿālim) who has studied theology and is a pupil of the pěsantrèn, the Indonesian religious school, or of the more modern madrasa ; he may even have studied in Mecca. Traditionally, the panghulus are highly-considered in Indonesian society. Sometimes the function is hereditary. One of his tasks is to supervise and coordinate the functions of the lower officials of the mosque: the imām, the k̲h̲atīb, the muʾad̲h̲d̲h̲in [q.v.] and the marbūṭ, the official who is responsible for maintenance. According to the linguistic area, these officials are called imām, kětib or ketip, modin or bilal, and měrbot, měrěbot or occasionally marbot.

In Java the mosque is also used for iʿtiḳāf [q.v.], especially during the last ten days of Ramaḍān.

Bibliography

H. Aboebakar, Sedjarah Mysdjid, dan amal ibadah dalamnja, 1955 [in Indonesian]

P. A. Hoessein Djajadiningrat, Islam in Indonesia, in Islam, the straight path, Islam interpreted by Muslims, ed. W. Morgan, 1958

G. F. Pijper, De Moskeeën van Java, and De Panghulu’s van Java, both in Studiën over de geschiedenis van de Islam in Indonesia 1900-1950, Leiden 1977

idem, The minaret in Java, in India Antiqua, a volume of oriental studies presented to Jean Philippe Vogel, Leiden 1947.

(G.F. Pijper)

IV. In the rest of South-East Asia.

That the traditional South-East Asian mosque originated in Indonesia and that it is formally sui generis cannot be disputed. Whether, as has been claimed, it developed in Java is less certain. Indeed, the history of Islam in Indonesia would suggest another possibility. The building in question was of wooden construction. It consisted of a simple structure on a square groundplan, erected on a substantial base. This distinguished it from the classic Indonesian house on stilts. The existence of internal pillars probably depended on its size. It had openings in the walls, probably closed with shutters, and an entrance in the east side, opposite the later miḥrāb. It is not known how the ḳibla was originally indicated, but some mark on the west wall seems likely. Above this groundfloor hall, which had relatively low walls, there were a number of upper storeys of decreasing area, up to a total of four: each individual storey, including the main hall, had its own roof, usually in palm thatch, with widespreading eaves. The upper stories contained loft-like rooms which were functional. The whole building was topped by a finial which, in later times, seems to have been crowned by a crescent. The whole building was enclosed within a wall which had a more or less elaborate gateway in the east side. Occasionally there was more than one gate. There is some evidence to suggest that the main structure was surrounded by an irregular moat which may have formed part of a stream which traversed the enclosure. There was no manāra ; the ad̲h̲ān was given either from the doorway of the mosque or from its top storey. This was probably preceded by the vigorous beating of a large skincovered drum, as is generally the practice today. A more simple structure, essentially a traditional Indonesian dwelling on stilts, serves as the model for a prayer hall which does not have the status of the mosque. It is still to be found in communities which cannot muster the requisite forty souls to constitute a congregation or, on occasions, as a supplementary building in a compound where it serves as a meeting place, a rest-house for visitors, an administrative centre as well as for ṣalāt when the mosque proper is closed.

This Indonesian prototype did not have the verandah, Javanese serambi, which is such a distinctive feature of the Central Javanese mosque. There is no evidence that this formed an original part of the mosque, from which it is, in fact, separated, both architecturally and dogmatically: shoes may be worn there. It seems to have derived from a royal building in pre-Islamic Central Java. Neither it, nor the externalised miḥrāb, belong to the original square mosque.

Various origins have been proposed for the basic Indonesian mosque. It has been derived from: (1) the čańd́i, a temple of either Hindu or Buddhist intention, ultimately of Indian origin but modified by Indonesian religious concepts; (2) the traditional bamboo and thatch cockpit used in Bali for the quasi-ritual cock-fighting; (3) the multi-tiered sacred mountain which is of widespread significance in Indonesian religions (the Balinese temple with multi-tiered thatched roofs known as a meru, after the Indian sacred mountain, is an architectural example of this). The objection to (1) is that, quite apart from its possible unacceptability to Muslim teachers, the čańd́i does not occur in those parts of Indonesia where conversions to Islam first took place. The cockpit hypothesis appears to suffer from inherent implausibility. There is, however, good reason for holding the concept of the sacred mountain as one component in the undoubtedly complex origin of the Indonesian mosque. It differs so profoundly from mosques elsewhere in the Islamic world, not least in Cambay [see k̲h̲ambāyat ] and other parts of Gud̲j̲arāt [q.v.] from which the main impetus towards conversion seems to have come.

South-East Asia lies across the sea route from the Middle East and the Indian sub-continent to China and beyond. The Malay Peninsula and Sumatra mark the area where the monsoon system of the Indian Ocean meets that of the Pacific, and constitute a natural interchange point. For two millennia or more merchants have travelled and traded through this region. After the coming of Islam many of these travellers were Muslims, but, although there were without doubt Muslim communities in the ports and harbours of the region, some of whose members may have traded in the interior, there is no evidence at all for conversion to Islam among the local peoples. (Nor, incidentally, is there any evidence for mosques to serve the needs of such Muslim traders.) The first instances of such conversion comes at the end of the 7th/13th century. A hint in a Chinese source dated 683/1281 receives striking confirmation from Marco Polo who spent several months in Sumatra, on his way home from China ten years later. Of Ferlec (Perlak) he noted “the people were all idolaters, but, on account of the Saracen traders who frequent the kingdom with their ships, they have been converted to the Law of Mahomet”, adding that this was only the townspeople, those of the mountains being like wild beasts. The ruler of Samudra (Pasai), where Polo spent some months waiting for the wind to change, and who died in 699/1297, certainly died a Muslim for his tombstone, which was imported from Cambay, gives his name as Malik al-Ṣāliḥ. It was from this remote, in Javanese terms, area of Aceh that Islam spread to the Malay Peninsula, above all to Malacca, [q.v.], to the north coast of Java and thence to other parts of Sumatra, to the coasts of Borneo and to the sources for the much sought-after spices, by way of the ports of Sulawesi and Maluku. Over a period of some three centuries, Islam followed the trade routes and with it there went the Indonesian masd̲j̲id, with its tiered, overhanging roofs. More than a dozen have been identified, notably by De Graaf. What is noticeable is that it was precisely in areas which had not been strongly influenced by Indo-Javanese architecture of Hindu or Buddhist tradition that the mosque of this type developed. Its origins have to be sought in the socio-religious structures of northern Sumatra in the communal house which, as elsewhere in Indonesia, once constituted the men’s house. Now without windows or its original interior divisions, in Aceh it has become the meunasah which serves as a prayer house, a meeting place, and an administrative centre as well as a Ḳurʾānic school. It had the advantage that it had never housed idols, but this does not explain how the teachers from Gud̲j̲arāt and elsewhere were persuaded to permit the adoption of such an aberrant form of mosque.

Bibliography

Illustrations of many of the mosques are in François Valentijn, Oud en Nieuw Oost Indien, 5 vols., Dordrecht-Amsterdam 1724-6. See also H. J. de Graaf, De oorsprong der javaanse moskee, in Indonesie, I, 289-305

B. Schrieke, The shifts in political and economic power in the Indonesian Archipelago in the sixteenth and seventeenth centuries, in Indonesian Sociological Studies, i. The Hague 1955, 1-82, W. F. Stutterheim, De Islam en zijn komst in de Archipel, Groningen 1952.

(A.H. Christie)

V. In China.

The Chinese term is Ch’ing-chen ssu, lit. “Pure and True temple”. Ch’ing-chen chiao (“Pure and True Religion”) being a Chinese synonym for Islam. The first Muslim settlements in China, dating from the early centuries of Islam, were established either by the sea route along the southern and eastern coasts (Canton and Hainan Island in Kwantung Province; Chuan-chou in Fukien Province, Hang-chou in Chekiang Province, Yang-chou on the lower Yangtze in Kiangsu Province); or by the overland “Silk Road” route at the ancient city of Ch’ang-an (some miles south of present-day Sian, Shensi Province), T’ang dynastic capital between 618-906 (corresponding approximately to the first three centuries Hid̲j̲rī).

Chinese Muslim tradition holds that numbers of mosques were established in these and several other cities by Saʿd b. Abī Waḳḳāṣ and various Companions of the Prophet or itinerant holy men during the first century, quite probably during the Rās̲h̲idūn caliphs’ period. Pending further archaeological excavation, however, most of these oral traditions must be treated with caution, and according to Leslie (op. cit., in Bibl., 40), but few sites “merit serious consideration”, the most important of which are:

1. Canton (the Huai-sheng mosque and Kuang-t’a minaret). This mosque, claimed by Muslim tradition as the first and oldest in China, may well date back to T’ang times, but the earliest extant reference dates from ca. 603-/1206, whilst the earliest mosque inscription (in Chinese and Arabic) records the re-building of the Huai-sheng ssu in 751/1350 after its destruction by fire seven years before. The presence of a mosque in Canton in 755/1354 is attested by Ibn Baṭṭūṭa.

2. Chuan-chou (the Sheng-yu mosque), also sometimes claimed as the earliest mosque in China, though Leslie considers this to be “a priori, less convincing” than the claim of the Huai-sheng ssu. The mosque inscription of 710/1310-11 (in Arabic) dates the first building of the mosque to 400/1009-10, commemorating a restoration which took place over three centuries later. It claims that the Sheng-yu ssu was the first mosque “in this land”, and calls it “The Mosque of the Companions” (al-Aṣḥāb).

3. Hang-chou (the Chen-chiao or Feng-huang mosque), ascribed by late Ming (11th/17th century) inscriptions to T’ang times, though Leslie rejects these unsubstantiated claims in favour of a Sung Dynasty establishment, Hang-chou being the capital of the Southern Sung (ca. 521-678/1127-1279), and by Yuan times “the greatest city in the world” (according to Marco Polo), with a substantial Muslim population living in its own ward (Ibn Baṭṭūṭa, Odoric).

4. Ch’ang-an (the Ch’ing-chiao or Ch’ing-ching mosque), which differs from those other mosques listed so far in that its foundation is ascribed to the arrival of Muslim soldiers travelling overland, rather than sailors coming by sea. Undated epigraphic evidence and long-established tradition date this mosque to the early T’ang (late Umayyad) period, but this remains inconclusive, and Leslie suggests that “until further evidence is forthcoming its is better to reject a T’ang date and query a Sung one, whilst taking for granted a Yuan [Mongol] presence”.

Leslie continues by providing “Desultory Notes” for numerous other cities in Eastern. Central and Northern China (49-53), before concluding that many thousands (or even tens of thousands) of Muslims, mostly of Persian and Arab origin, were resident in China during T’ang Dynasty times, though little definitive evidence exists for the number of mosques which had been established during this early period of Chinese Islam. It is clear, however, that most of these Hsi-yu jen or “Westerners” were semi-permanent or permanent residents, many of whom would have intermarried freely with the indigenous Chinese population, thereby giving rise to a nascent Chinese-speaking, increasingly Sinicised Muslim population which would, by Ming times, develop into the Hui Chinese Muslim community. Certainly by T’ang times, the distinction was already being made between “foreigners” and “native-born foreigners”. S̲h̲arīʿa law requires the establishment of congregational mosques wherever communities of more than forty adult male Muslims are gathered together; the presence of many small mosques along the Chinese coast and (to a lesser extent) in the interior may, therefore, be taken for granted by late T’ang/Sung times. Doubtless, except in the more important coastal towns such as Canton (Khanfu) and Chuanchou (Zaitun) these mosques would have been fairly insubstantial buildings, long since altered beyond recognition or destroyed; thus, definitive proof of the extent of mosque-building in China during this early period will depend upon future archaeological excavations.

The Yüan period (ca. 678-770/1279-1368) was characterised by a substantial expansion of Islam in the central and western parts of China, most particularly in Yunnan, where Sayyid Ad̲j̲all S̲h̲ams al-Dīn Buk̲h̲ārī (who conquered and subsequently adminstration. istered the former Nan-ch’ao area for the Mongols) is credited with establishing two mosques in the region. Sayyid Ad̲j̲all and his family may be seen as the archetypical example of Muslims in service under the Mongols—by whom they were employed as soldiers, administrators and financial middlemen—and from Yuan times the central focus of Islam in China moved definitively away from the southern coastal ports towards the north and west. Certainly, the oldest mosques in Yunnan and the north-west are likely to have been established during this period, a trend which was continued under the Ming Dynasty (ca. 771-1054/1368-1644) which is also known as a period of Sinicisation for the Chinese Muslim community— indeed, it may be that the Chinese-speaking Hui Muslim community emerged as a separate and distinct entity (paralleling, for example, the Swahili [q.v.] in East Africa and the Mappila [q.v.] of southern India) during this period.

It is probable that the mosques of the Hui (Chinese-speaking) Muslims, which are scattered throughout China but are particularly numerous in the provinces of Kansu, Ningsia, Tsinghai and Yunnan, evolved in their characteristic form during this period. Certainly under the Ming, the nascent Hui community expanded greatly as a result of intermarriage, overt (and, perhaps more frequently, covert) missionary work, and their success in the fields of military and commercial venture. Wherever Hui settled in any numbers, ḥalāl establishments (caravanserais, restaurants, inns), mosques and attendant madrasa s soon followed. As Israeli notes (op. cit. in Bibl, 29), many mosques constructed during the Ming period were built in a style reminiscent of indigenous Chinese temple architecture, either eliminating the minaret altogether, or eschewing the distinctive styles associated with the mosques of Central Asia, South Asia and the Middle East in favour of Chinese-style pagodas. As a result of this architectural development, the muezzin could no longer call the faithful to prayer in the usual way, but stood inside the mosque instead, calling the ad̲h̲ān behind the main mosque entrance. “And when one entered the mosque, one was struck by the traditional Muslim flavour; cleanliness and austerity. Except for the Emperor’s tablets that were mandatory in any house of prayer, there was no sign of Chinese characters or Chinese characteristics. On the walls there were Arabic inscriptions of verses from the Qur’an and the west end (qibla) was adorned with arabesques. Once the believers were inside, they put on white caps, shoes were taken off, elaborate ablutions were ritually performed, and the prayers began in Arabic, with heart and mind centred on Mecca. When prostrating themselves before the Emperor’s tablets, as required, the Muslims would avoid bringing their heads into contact with the floor... and thus did they satisfy their consciences in avoiding the true significance of the rite—this prohibited worship was invalid because it was imperfectly performed” (Israeli, op. cit., 29).

Israeli defines this combination of external Sinicisation of mosque building and internal Islamic orthodoxy as a manifestation of the dichotomy of Chinese Islam. Certainly, the functions of the mosque remained immediately recognisable in their Islamic purpose. Thus, besides the area set aside for prayer, the interior of larger Chinese mosques is generally divided between lecture hall, dormitory, conference rooms, community leaders’ offices, and the “dead man’s room” for washing and otherwise preparing deceased Muslims for burial. Amongst the bestknown and most beautifully decorated of these traditional Chinese mosques are the Niu-chieh ssu (Ox Street mosque) in Peking, and the Hua-chueh ssu in Sian.

By contrast with the Sinicised Hui Chinese mosques scattered throughout “China Proper”, the mosques of the periphery are often very different. Thus the mosque architecture of Sinkiang conforms closely to that of neighbouring Western Turkestan, whilst in the far north-east (Heilungkiang Province), an area formerly much influenced by Russian culture, mosque may sometimes outwardly resemble Orthodox churches. In this context, an informative trilingual study illustrating many of the best-known mosques in China and clearly depicting the different architectural forms has recently been published by the China Islamic Association (op. cit. in Bibl., 1981).

Bibliography

K. Himly, Die Denkmäler der Kantoner Moschee, in ZDMG xli (1887), 141-74

G. Phillips, Two mediaeval Fuh-Kien trading ports: Chuan-Chow and Chang-Chow, in T’oung Pao, vii (1896), 223-40

G. Arnaiz and M. van Berchem, Mémoire sur les antiquités musulmanes de Ts’iuan-tcheou, in TP, xii (1901), 677-727

Cl. Huart, Le texte turc-oriental de la stèle de la mosquée de Péking, in ZDMG, lvi (1902), 210-22

W. Bang, Über die Mandschu Version der Viersprachigen Inschrift in der Moschee zu Peking, in Keleti Szemle, iii (1902), 94-103

H. Saladin, Monuments musulmans de Chine et d’Extrême-Orient, in Manuel d’art musulman, Paris 1907, 579-83

“N.”, Les Mosquées de Pekin, in RMM, ii (1907), 570-73

E. Blochet and A. Vissière, Épigraphie musulmane chinoise, in ibid., v (1908), 289-93

R. Ristelhuber, L. Bouvat, F. Farjenel, Études chinoises, in ibid., iv (1908), 512-30

anon., Liste des mosquées de Pékin, in ibid., vi (1908), 699

M. Broomhall, Islam in China: a neglected problem, London 1910, 83-120, 183-90

Le Commandant D’Ollone, A. Vissière, E. Blochet, et alii, Recherches sur les Musulmans chinois, Paris 1911, passim

G. Cordier and Vissière, Études sinomahométanes: deuxième série. Renseignements envoyés par M. G. Cordier sur la chambre funéraire et le temple commémoratif du Seyyid Edjell à Yun-nan-fou, in RMM, xv (1911), 60-9

eidem, (Note on two photographs of the Mosque of Seyyid Edjell Omar), in ibid., xxv (1913), 306

Vissière, Études sino-mahométanes: deuxième série. VII. L’Islamisme à Hang-tcheou, in ibid., xxii (1913), 1-84

Cordier, Études sino-mahométanes: troisième série. Les Mosquées de Yun-nan-fou, in ibid., xxvii (1914), 141-61

Vissière, Inscriptions sino-mahométanes de Foutcheou, in ibid., xxvii (1914), 162-73

A. Von le Coq, Buried treasures of Chinese Turkestan, London 1928 (photographs)

Ito Chuta, S̲h̲ina kenchiku sos̲h̲oku (“Chinese architectural decoration”), Tokyo 1941, i, 107-9

ii, plates 42-8

A. Hutt, The Central Asian origin of the Eastern Minaret form, in Asian Affairs, viii 3 (June 1977), 157-62 (with reference to Sinkiang)

R. Israeli, Muslims in China

a study in cultural confrontation, London 1980

China Islamic Association, The religious life of Chinese Muslims, Peking 1981

D. D. Leslie, Islam in traditional China, Canberra 1986, esp. 40-68 (plus extensive Chinese language bibliography, glossary, etc.; a uniquely competent and useful reference work).

(A.D.W. Forbes)

VI. In East Africa.

In East Africa the mosque is commonly spoken of in Swahili as msikiti, pl. misikiti, but msihiri, misihire in the Comoro Islands; and cf. Swahili sijida, the act of adoration, and verb sujudu “to prostrate oneself”, from Ar. sad̲j̲ada. Nineteenth-century traditional histories claim the setting up of Muslim cities on the eastern African coast in the 7th and 8th centuries A.D. Of this there is no earlier literary evidence, but a mosque is mentioned in the Arabic History of Kilwa named Kibala (possibly a Bantu form from ḳibla) as existing on that island ca. 950 A.D. In spite of recent excavations at Kilwa [q.v.] by H. N. Chittick, there has so far been no positive identification of a mosque of this period. The first reliable evidence is from inscriptions. Cerulli reports one in the Friday Mosque at Barāwa, Somalia, dated 498/1104-5, while on Zanzibar Island there is the well-known Friday Mosque at Kizimkazi [q.v.] which has an inscription dating its foundation to 500/1107. The inscription is certainly of Sīrāfī provenance, which does not argue that Zanzibar was much Islamicised at this period. The 4th/10th century Kitāb ʿAd̲j̲āʾib al-Hind of Buzurg b. S̲h̲ahriyār of Rāmhurmuz contains, however, the tale of the conversion of an eastern African king of a place of which no identification is given; he was followed by his people. In the same century al-Masʿūdī, who visited eastern Africa, speaks of the people and their sovereigns as pagan. By the 6th/12th century al-Idrīsī says that “the people, although mixed, are actually mostly Muslims”, which would accord with the epigraphical evidence.

Between 1962 and March 1964 the greater number of known mosques, from mediaeval times to the 18th century, both standing and ruined, were planned and photographed by P. S. Garlake. He omitted, however, an important series of foundation inscriptions of mosques at Lamu [q.v.], some twenty in all, and ranging from the 14th to the 19th century. He rightly says that “the most sensitive indicator of change and development in style and decoration is bound to be the mosque rmhrab”: he distinguishes a clear and unbroken development of style and technique from the early classic miḥrāb with a plain architrave of the 14th and 15th centuries; a developed classic miḥrāb in which the plain surfaces of the architrave are broken by decoration; a neo-classic miḥrāb of greater elaboration, both this and the foregoing in the 16th century; a simplified classic miḥrāb restricted to northern Kenya, and a derived classic miḥrāb on the Tanzanian coast in the 18th century, in which, however, there were new developments that led to multifoliate arches of an elaborate character. The dating of some of these miḥrābs derives from inscriptions, but is based to a great extent upon the evidence of imported pottery and Chinese porcelain, the latter coming to be used as a decoration by insetting it into the architrave of the miḥrāb.

All the 19th century Swahili settlements in eastern Africa are on the edge of the shore: Gedi, two miles from the Mida creek, is the sole exception. Some earlier mosques, however, are found on cliffs or headlands, where they may have been placed to serve as mariners’ marks. Some of them are still of special veneration for seafarers. The population in these places was on the whole small, and only at Kilwa [q.v.] and at Mogadishu [see maḳdishū ] was the need felt for mosques of more than modest size. Throughout the coast from Somalia to Mozambique, the only available building material of a permanent character was coralline limestone, obtained either from old raised beaches or directly from coral reefs. Mouldings, arches, and all features wherever precision was required, were of finely dressed coral blocks. A fine concrete, whose aggregate was coral rubble, was used for circular and barrel vaults. The method of burning it has survived to this day. From it also was made the plaster which in the 18th century was used to decorate not only the miḥrāb but also elaborately decorated tombs. There was a limited répertoire of mouldings, used also on tombs, and—more sparingly—in domestic architecture. The planning of all buildings, religious and domestic, was restricted by the span of the timber rafters, always of mangrove wood, which never exceeds 2,80 metres or approximately 9 feet. Even the vaulted buildings conform to this as to a fixed and unalterable convention. Thus even in the Great Mosque of Kilwa, with its five aisles and six bays, there is a sense of constriction rather than of spaciousness. Walls may be built of dressed coral limestone but quite commonly of coral rubble plastered over. Piers occur in mosques in Kenya and Pemba during the 14th to 16th centuries, but not in the south. After the 13th century in Tanzania, columns alone are found, some square and some octagonal. Generally, these were of dressed coral, but occasionally, as at Kaole (southern mosque) and in the northern muṣallā of the Great Mosque of Kilwa, wooden columns fitted into coral sockets were used. Because of the difficulty imposed by the length of the rafters, the master-builders—for only rarely can architects have been employed, and perhaps only for the Fak̲h̲r al-Dīn Mosque at Mogadishu—in seeking to erect a building of a particular breadth, frequently encumbered the perspective of the miḥrāb by constructing a central arcade of pillars. This clumsy feature (which occurs quite unconnectedly in certain mediaeval European churches) appears not only in two-aisled mosques such as those of Tongoni and Gedi but also in the four-aisled Friday Mosque of Gedi and the original North Mosque which forms part of the Great Mosque of Kilwa.

Minarets [see manāra. 3. In East Africa] are very rare, and minbars [q.v.] have certain idiosyncratic features. In all, the mosques of the eastern African coast have a distinct regional character of their own, deriving in earlier times from the common use of ogival or returned-horseshoe arches, and in later times from the elaborate plaster decoration of the miḥrāb and its architrave.

Bibliography

H. N. Chittick, Kilwa, 1975, describes the Kilwa mosques, bringing up to date P. S. Garlake, The early Islamic architecture of the East African coast, Nairobi 1966, with its numerous plates and plans and exhaustive bibliography up to that date

G. S. P. Freeman-Grenville, Some preliminary observations of medieval mosques near Dar es Salaam, in Tanganyika Notes and Records, no. 36 (1954), is wholly superseded by the finding of better evidence for date

see also J. S. Kirkman, Men and monuments of the East African coast, London 1964, and Fort Jesus, Oxford 1974

and, for inscriptions, G. S. P. Freeman-Grenville and B. G. Martin, A preliminary handlist of the Arabic inscriptions of the eastern African coast, in JRAS (1973).

(G.S.P. Freeman-Grenville)

VII. In West Africa.

In Muslim West Africa, the smallest hamlet has its mosque, and the quarters of an individual town compete with one another in the construction of cultic sites. In most villages, the mosque is situated in the middle of the public square, near the tree which is the traditional place for bargaining and discussion (“palaver”); it is generally constructed in the style of a large shed, roofed with zinc plates and bamboo partitions or with banco or with moulded clay, and has the appearance, in the majority of cases, of the most attractive building in the locality, often surrounded by bushy trees. The mosque is regarded with pious respect, and is kept clean. Volunteers, often women of a certain age, accept responsibility for maintenance, cleaning and the supply of drinkable water for the faithful.

In towns, the mosque is a more substantial building and it dominates the neighbourhood with its minaret or minarets. Sometimes, as in the case of the Great Mosque of Dakar, it has only one, while that of Touba, the most important centre of the religious brotherhood of the Murids, has three, of which the tallest, known as the “Lamp” (Fall) measures 83 m. In fact, it is the modern mosques which possess minarets; the most ancient have none, but still dominate their surroundings with cubic pillars. In small villages, the floor of the mosque is covered with matting or with fine sand which is sifted every day. In the urban setting, oriental carpets cover the floor. A palisade of bamboo or zinc plates or even a cement wall forms an enclosure within which a spacious courtyard is set out, to enable those worshippers who cannot pray at the times when the mosque is crowded to perform their religious duties. On the left side of the larger mosques, the place reserved for women is separated from that where the men pray by a metal grill.

The imām leads the prayer standing in a niche (minbar) in the ḳibla wall. The Great Mosque is furnished with a throne, a kind of raised dais where the imām takes his place to preach his sermon and to harangue the faithful, first in Arabic and then in the local language.

All the other facilities, including lavatories and taps for ablutions, are located on the exterior. In a corner of the courtyard there is a hut for the washing of corpses.

Each imām is served by a nāʾib or deputy who officiates in his absence. Two or more muezzins make the call to prayer from the tops of the minarets. In the larger mosques loud-speakers have been installed, to relay either the call to prayer or the sermon of the imām. The majority of imāms receive no monthly salary. The imām of the Great Mosque of the Senegalese capital is one of the few who receives regular payment and occupies an official residence; more often, the imām and his family are accommodated in the mosque.

The architectural style reproduces especially that of the Mag̲h̲rib. It is thus that the Great Mosque of Dakar, inaugurated by King Ḥasan II, was built under the supervision of a Moroccan architect, as was the Islamic Institute which adjoins it. However the ancient mosques of northern Senegal, including those of Halwâr, Ndioum, Guédé and Dialmath, are in the Sudanese style of the mosque-institutes of the towns of Mali (Djenné, Mopti, Timbuktu, etc.) and of the land of the Sahel (cf. J. Boulègue, Les Mosquées de style soudanais au Fuuta-Tooro (Sénégal), in Notes africaines, 136 (Oct. 1972), 117-19). This is a style characterised by its massive buttresses exceeding the height of the roof, in a rounded, conical form, with a small cubic minaret; the whole is constructed in brick made from dried earth and covered with a facing of the same material and ochre or beige in colour. The walls are very thick. An elaborate system of ventilation maintains a freshness similar to that provided by airconditioning.

Religious function. In West Africa, the principal function of the mosque is still religious; each quarter possesses several, and in this context a genuine rivalry prevails between quarters or between members of different brotherhoods. It is thus that the mosque of the Tid̲j̲anīs is found alongside those of the Murīds [see murīd ], of the Ḳādirīs [see ḳādiriyya ] or of the Hamallites [see ḥamāliyya ]. The faithful fill the mosques without regard for their particular affiliation. The Tid̲j̲ānīs organise gatherings in the mosque after morning and evening prayers to recite, in chorus, the litanies (d̲h̲ikr) peculiar to their religious order. This ritual is performed around a carpet and in darkness. But on Fridays or at times of canonical festivals, great crowds of Muslims are seen streaming towards the mosques clad in their splendid boubous or flowing robes.

Special prayers for the dead are also offered in the mosque. In this case, the bier is placed before the faithful, who pray upright without bowing or sitting. After these funeral rites, the parents of the deceased arrange a ceremony of recitation of the Ḳurʾān “for the repose of his soul”.

The veneration of which the mosque is the object inspired Cheikh El-Had̲j̲i Malik Sy (1853-1922), founder of the zāwiyya tid̲j̲āniyya of Tivaouane, to compose a poem in Arabic consisting of forty verses in rad̲j̲az style and revealing the details of a whole system of etiquette. Cheikh Aliou Faye, the chief marabout of the Gambia, revised and embellished his master’s poem, entitling his version Tabs̲h̲īrat al-murīd or “The way of success for the disciple”. The following are a few of the verses:

Whosoever wishes to enter Paradise without punishment and without the need to give an exact account of his actions at the Resurrection, should build a mosque for God the Merciful, and he will be granted one hundred and thirty palaces in Paradise.

Every believer who enters this mosque to pray will obtain a pleasant dwelling in Paradise.

A mosque may be built in any place, even in the square of a church or a or a synagogue.

There it is forbidden to grow crops, to dig wells, to sew and to compose [profane] poetry.

There it is forbidden to eat garlic, leek, onion, to shave, to cause an injury to a human being, to cut the nails, to cast lice or fleas and to kill them.

To tie animals, confine the mentally ill, to allow a criminal to enter and be seated.

All mosques are of equal worth, with three exceptions: those of Mecca, Medina and Jerusalem, which are the best mosques.

Social function. Besides this predominant religious role, the mosque also performs a very important social function. It is there, in fact, that, under the patronage of the imām, marriages are contracted between the parents of the betrothed parties. The father or guardian of the prospective bride gives her hand to the father or guardian of the suitor and receives the dowry. This function is so important that when information is sought regarding the marital status of a female person, the question is asked: “Have the men gone to the mosque for her?” (in Wolof: Ndax demn̂anu jaka ja?). As a form of pleasantry and to tell a girl that she is nubile, the remark is made: “I shall go to see the imām about it.” Parents or guardians may be accompanied to the mosque by other parents and friends who act as witnesses. The relatives of the suitor bring the dowry which they entrust to the imām; the latter gives it to the father or guardian of the prospective bride and recites the sacramental formula. In the presence of all, the imām blesses the couple. Cola, non-alcoholic drinks or delicacies are distributed.

Even though, since the promulgation of a “Family Code” in Senegal, for example, some ten years ago, marriages must be contracted before the mayor or the representative of the public authorities, it is considered that, without the mosque playing a part, the matrimonial union is not valid. Thus the imām in fact represents the municipal magistrate.

Often the elders of the village hold meetings not under the traditional tree, but inside or in the courtyard of the mosque at any hour of day or night to discuss public matters; finance for the sinking of wells, construction of a market, division of the produce of common land, preparations for the reception of distinguished guests, etc. In this case, the mosque represents a kind of national assembly where all the affairs of the village community are the object of wide and democratic debate.

Sometimes the mosque performs the role of a tribunal where disputes between members of the village are laid public and closely examined. Solutions are always formed on the basis of the S̲h̲arīʿa, or of local custom, or of both. These may be disputes between spouses, between two dignitaries, between two families, between herdsmen and stock-breeders, between a representative of the state and local landowners, between traditional chiefs and religious leaders. Sometimes the division of bequests is performed in the mosque under the supervision of the imām.

Some mosques provide places of lodging for strangers. It is in this way that travelling Muslims are accommodated. Furthermore, any person who is regarded as having lived a pious life and who has contributed to the building of the mosque, is buried there after his death. Such is the case of Cheikh Ahmadou Bamba Mbacke, Cheikh El-Had̲j̲i Malik Sy, Cheikh Ibrahima Niasse, Cheikh Ahmadou Anta Samb, and Bouh Kounta respectively at Touba, Tivaouane, Kaolack, Kébémer and Ndiassance (Senegal).

Many other men renowned for their piety or for their work in the service of Islam are entombed within or in close proximity to the mosque.

Economic function. The economic function of the mosque is explained by the fact that the temporal is always closely linked with the spiritual. Thus, for example, the sums raised from legal alms (zakāt) are in most cases entrusted to the imām of the mosque who, as an expert in the matter, ensures that they are distributed to those entitled to them. Sometimes cattle are led to the mosque to be slaughtered by the imām, who distributes the meat to the needy. Every Friday, a whole army of beggars is seen flocking to the mosques, attracted by the prospect of receiving charity from the wealthier believers. The same spectacle is witnessed during the major Islamic feasts of Tabaski and Korite.

The imām received a gratuity for his services when marriage is celebrated. Even though the sum is by no means considerable, it is important for the imām who is not salaried. In the course of one Sunday afternoon he may preside over several marriage ceremonies. Furthermore, numerous mosques receive requests for readings of the Kurʾān in exchange for a certain sum, the amount being left to the discretion of the customer.

Mosques which incorporate tombs receive a profitable income as a result of daily, weekly, monthly and annual pilgrimages or on the occasion of major Islamic feasts. This applies in the case of the mosque of Touba during the well-known feast of Magal, which commemorates the departure into exile (in 1895) of Cheikh Ahmadou Bamba Mbacke, founder of the brotherhood of the Murīds, and that of Tivaouane at the time of the Mawlūd [see mawlid ].

Cultural function. Although the mosque in West Africa fulfils a considerable economic role, its function in the cultural sphere is more striking. In the majority of cases, the courtyard of the mosque is the setting for a Ḳurʾānic school. Sometimes dozens of young children, boys and girls, are seen squatting in a halfcircle before their master, who sits either on the ground, on a sheepskin rug or reclining on a couch, holding a cane. Each pupil places on his knees a tablet on which the lesson to be learned is inscribed in ink made from soot from cooking-pots. In the evening, after twilight and before the meal, a large fire is lit and the verses to be learned are read by the light of the flames. By this educational method, in the shadow of the mosque, many scholars arrive at the point where they can recite the entire Holy Book by heart.

The mosque also serves as a high school and university when, having memorised the Ḳurʾān, the pupils become students and learn the other Islamic sciences: exegesis, ḥadīt̲h̲, theology, mysticism, Muslim law and even literature, history, logic, astronomy, rhetoric, etc.

It is also in the mosque that lectures are held on various subjects relating to religion, as well as educational lectures given by scholars or distinguished guests from other Muslim countries. In the mosque, throughout the month of Ramaḍān, marabout exegetes expound and comment on the Ḳurʾān before an audience, either to recall the teaching of the Holy Book or to instruct the faithful. On the “Night of Destiny” nobody sleeps, and reverent vigil is held in the mosque. Also in the mosque, particularly at Tivaouane, the sanctuary of Tid̲j̲ānism in Senegal, the head k̲h̲alīfa of the disciples of the brotherhood founded by Aḥmad al-Tid̲j̲ānī (1737-1815 [q.v.]) expounds and comments on the Burda of al-Buṣīrī (608-ca. 695/1212-ca. 1295 [q.v. in Suppl.]).

Political function. Finally, the mosque performs in West Africa a political function which is far from insignificant, because the region contains a very substantial percentage of Muslims. This figure is increasing as a result of large-scale conversion to Islam of followers of other religions (Christianity and animism). Islam has enjoyed a revival of activity under pressure exerted both from the interior of this zone and, to a lesser extent, from the exterior. In Senegal, for example, the quite recent appointment of M. Abdou Diouf to the post of chief magistrate has had a considerable influence in this domain, to such an extent that, unlike his predecessor, the head of state, accompanied by the presidents of the National Assembly and the Economic and Social Council, participates behind the senior imām in the prayers conducted on the occasion of major festivals. In his k̲h̲uṭba, the latter invariably affirms his loyalty to the authorities and invites the believers present to pray, with him, for the President of the Republic and the members of his government, whom he mentions by name, appealing to God to “perpetuate their rule and assist them, giving peace, health and long life to them, to their families and to Senegal”.

This account of the activity of the present President of the Republic of Senegal applies to the other Muslim Heads of State of West Africa.

The imām often uses the occasion of the Friday Prayer to draw attention in his k̲h̲uṭba to themes of concern to the government such as the misappropriation of public funds, corruption, juvenile delinquency, drugs, prostitution, the degradation of morals, the urgent need to combat bush-fires and desertification.

After this survey of the functions of the mosque in West Africa, it may be affirmed that it performs a multifarious role in this region by virtue of its status as the supreme place of prayer.

Bibliography J. M. Cuoq, Les Musulmans en Afrique, Paris 1975, 103-271, gives information and bibliographies concerning religious life in West Africa

see also, in particular, J. Schacht, Sur la diffusion des formes a”architecture religieuse musulmane à travers le Sahara, in Travaux de l’Inst. de Rech. Sahariennes, xi (1954), 11-27.

(A. Samb)

Cite this page

Pedersen, J., Hillenbrand, R., Burton-Page, J., Andrews, P.A., Pijper, G.F., Christie, A.H., Forbes, A.D.W., Freeman-Greenville, G.S.P. and Samb, A., “Masd̲j̲id”, in: Encyclopaedia of Islam, Second Edition, Edited by: P. Bearman, Th. Bianquis, C.E. Bosworth, E. van Donzel, W.P. Heinrichs. Consulted online on 10 January 2017 <http://dx.doi.org/10.1163/1573-3912_islam_COM_0694>

Encyclopaedia of the Qurʾān / Mosque

---

Article Table of Contents

Early history and constituent parts

Types of mosques

Decoration of mosques

Bibliography

A Muslim place of prayer (q.v.). The English word “mosque” derives, via the French mosquée, the Old French mousquaie, the Old Italian moschea and moscheta, and the Old Spanish mezquita, from the Arabic word masjid, meaning a place of prostration (sajda, see bowing and prostration) before God. The word masjid (and its plural masājid) appears twenty-seven times in the Qurʾān, fifteen times in the phrase al-masjid al-ḥarām, “the holy mosque,” where it presumably refers to the sanctuary surrounding the Kaʿba (q.v.) in Mecca (q.v.). The word masjid is used once in the phrase al-masjid al-aqṣā, “the furthest mosque” (q 17:1). In Muḥammad's lifetime this probably referred to a place of prayer in heaven (see ascension), although later commentators have universally understood this phrase to refer generally to the sanctuary of Jerusalem (q.v.) and specifically to the mosque erected at the south end of the Temple Mount. Other uses of the word masjid in the Qurʾān indicate that it could be applied to any place where God was worshipped, whether in Islamic or pre-Islamic times, as, for example, the tomb of the Seven Sleepers mentioned in q 18:21 (see men of the cave). Later authors agreed that the concept of the masjid was not specific to Islam; the historian al-Ṭabarī (d. 310/923) mentions that King David (q.v.), for example, had a masjid (Pedersen, Masdjid). The word muṣallā, referring specifically to a place for ṣalāt, or prayer, appears only once in the Qurʾān (q 2:125), where God made the maqām Ibrāhīm, “station of Abraham,” in Mecca “a place of prayer” (see abraham). Whereas any place where ritual worship (q.v.) is performed would technically be a muṣallā, the word has taken on a special meaning in Islam as a large undifferentiated space, usually outside the city, where the extraordinary ṣalāts are performed (see ritual and the qurʾān). These include the festival prayers (see festivals and commemorative days) marking the end of the holy month of Ramaḍān (q.v.) and the tenth day of Dhū l-Ḥijja, when animals are also slaughtered (see slaughter; consecration of animals) in commemoration of Abraham's (Ibrāhīm's) sacrifice (q.v.), as well as the extraordinary prayers for rain. In later times, particularly in Persian-speaking lands, the festival muṣallā is normally known by the Persian name ʿīd-gah, “festival place.” The word muṣallā has also taken on a secondary meaning in some regions of the Muslim world as “the covered part of a mosque.” Jāmiʿ, a third word commonly applied in later times to congregational mosques, does not appear in that form in the Qurʾān but derives from the later usage masjid al-jāmiʿ, “congregational mosque,” which itself was also transformed into such phrases as the Persian masjid-i jumʿa, “Friday mosque” and the Urdu jamaʿ masjid, “congregational mosque.” In modern usage, the word masjid (Turkish mescit) is sometimes used to refer to a small mosque for daily prayer, while the word jāmiʿ (Turkish cami; Ottoman cāmiʿ) is understood to refer to a congregational mosque for communal worship on Friday (see friday prayer).

The Qurʾān gives absolutely no indication of what, if any, form a masjid should take, and perfectly valid worship may be performed after ablution (see cleanliness and ablution) virtually anywhere, using only the most minimal markings on the ground or a mat or rug. When Muslims gather in groups for communal worship, they line up in rows facing the qibla (q.v.), or direction of prayer, and repeat a series of prayers and, following the imām (q.v.) or prayer leader, perform a series of prostrations. Starting from these modest beginnings, over the centuries Muslims have built praying-places of great power and beauty that count among the finest examples of world architecture (see art and architecture and the qurʾān). This article will discuss the history and development of such structures and their constituent elements as they evolved over the centuries.

Early history and constituent parts

It is generally accepted by both Muslim and non-Muslims alike that the simple house erected by the prophet Muḥammad after he emigrated from Mecca to Medina (q.v.) in 622 (see emigration) played a key role in the evolution of the mosque. According to later accounts, this building was a roughly-square building with mud-brick walls approximately 100 cubits (ca. fifty meters) to a side. Several doors led to the interior, which comprised an open court with several small rooms along the eastern wall in which the Prophet and his wives lived (see wives of the prophet). Porticoes supported on palm-trunks and thatched with palm fronds running along the north and south walls provided shade for the activities of Muḥammad, his family and his Companions (see family of the prophet; people of the house; companions of the prophet). Until Muḥammad broke with the Jews (see jews and judaism) in Rajab or Shaʿbān 2/December 623-January 624 and the qibla, or direction of prayer, was changed from Jerusalem to Mecca (q 2:136), the northern portico, known as the mughaṭṭa, or “covered area,” was used for prayer and the southern portico, known as the ṣuffa, “row (of columns),” was used for accommodating Muḥammad's dependents and guests, who were known as the ahl al-ṣuffa (see hospitality and courtesy; community and society in the qurʾān). After the qibla was shifted to Mecca, these arrangements were reversed, with the larger covered area on the south and the smaller one on the north. Following this precedent in later times, the mosque might provide temporary lodging for travelers and scholars (Grabar, Formation, 105-38; Hillenbrand, Islamic architecture, 30-128).

On Fridays, the Prophet would lead noon (q.v.) congregational worship in the court, his position marked by his lance (ʿanaza) thrust in the ground. He would address the community of believers from a raised seat, or pulpit, made from tamarisk wood, which was moved into position as needed. Although the minbar is not mentioned in the Qurʾān, the Prophet's seat was derived from the pre-Islamic judges' seat and symbol of authority (Becker, Die Kanzel). The minbar is the only common feature of the later mosque to have been used by the Prophet. The earliest minbars had only two or three steps, but the earliest example to survive is a teakwood specimen with many steps from the ninth century in the Great Mosque of Qayrawān, now in Tunisia.

Following Muḥammad's death in 632, he was buried under the floor of one of the living rooms to the side of the court. The Prophet had disapproved of any monumental commemoration of the dead (see death and the dead) but his grave was eventually surrounded by a low screen and covered with a dome as the building was expanded. Although the grave became the focus of popular veneration, it was never allowed to become a focus of prayer (see intercession). In later times, bodies were not buried under mosques but tombs of important individuals were sometimes erected adjacent to them. In other cases, when the tomb of a particularly holy figure in some cemetery became the focus of popular veneration, a mosque might eventually be built to accommodate worship there.

Muḥammad's immediate successors, the caliphs (see caliph), continued to use the house-mosque in Medina and it was expanded to accommodate the increased numbers of believers. As a result of the rapid expansion of the faith throughout Arabia and into Iraq (q.v.), Syria (q.v.) and Egypt (q.v.), believers resident in these regions needed places for communal worship. In some regions, such as Syria, existing churches (see church) provided sufficient and suitable space, and they were either appropriated or divided between the Christians and Muslims. In other regions, such as Egypt or Iraq, where suitable buildings were lacking in the required places, new structures were erected. In Jerusalem, according to the European pilgrim Arculf, the Muslims had erected a massive but rather crude structure at the southern end of the Temple Mount (see aqṣā mosque) by ca. 50/670 (Creswell, Muslim architecture; id., A short account). According to much later accounts, the first mosque in Egypt was built at Fusṭāṭ; it was a small structure measuring 50 × 30 cubits (25 × 15 meters) with a very low roof supported on multiple columns or piers. In Iraq, where the new towns of Kūfa and Baṣra were founded in 19/640, the first mosques were marked out by a ditch or low wall and orientated towards Mecca. The Mecca-facing, or qibla, part of the mosque might be covered with a palm-thatch roof supported on multiple columns or piers to provide shade. As Muslim power was consolidated in the following decades, these makeshift and temporary structures were rebuilt with more durable materials, but the many-columned (“hypostyle”) system of supports was maintained.

The second and fourth caliphs, ʿUmar (q.v.) and ʿAlī (see ʿalī b. abī ṭālib), were murdered in mosques and the third, ʿUthmān (q.v.), was murdered while reading the Qurʾān, so it was thought necessary to provide some sort of protection for the ruler when he attended the mosque. This screened enclosure, which allowed the ruler to be seen but not approached, was known as the maqṣūra. The eighth/fourteenth-century historian Ibn Khaldūn ascribed the introduction of the maqṣūra to the first Umayyad caliph Muʿāwiya (r. 41-60/661-80) or to one of his successors, Marwān I (r. 64-5/684-5), but the sources are in some disagreement about the date. The first examples were made either of brick or wood, and the oldest to survive is the magnificent wooden example from the fifth/eleventh century also in the Great Mosque of Qayrawān. In later centuries, when the Islamic rulers participated with less frequency in Friday worship, the maqṣūra came to serve less of its original practical function, although it and the area immediately around the miḥrāb and minbar remained the focus of the mosque's interior decoration. The Ottoman sultans later introduced a royal loge, hünkar mahfili, into their mosques. Unlike the centrally-placed maqṣūra, the Ottoman loge was placed to the side of the mosque and in some instances, such as the Selīmiye mosque (Selīm II, r. 974-82/1566-74) in Edirne, raised on the second floor (see Fig. viii).

With the establishment of the Marwānid branch of the Umayyad family and the shift of the capital from Arabia to Syria, the caliphs ʿAbd al-Malik (r. 65-86/685-705) and al-Walīd (r. 86-96/ 705-15) embarked on an ambitious program of construction in the major cities of the realm. The “sacred mosque” (al- masjid al-ḥarām) in Mecca and the Prophet's mosque in Medina, which had already been enlarged several times to accommodate larger numbers of Muslims, were completely rebuilt, as was the mosque of Damascus, which the Muslims had heretofore shared with the Christians of that city. The Umayyad mosque of Medina is known only through later texts (Sauvaget, Mosquée omeyyade) but the Damascus mosque, despite a disastrous fire in the late nineteenth century, survives largely intact. Built within the walls of a Roman temple enclosure, the Damascus mosque was one of the most ambitious architectural projects of the time. Like the Dome of the Rock in Jerusalem, the mosque used the forms and motifs of late antique architecture, such as basilical halls with columns, arches, gables, domes, mosaics and marble revetments, to create a new Islamic architecture. On the south, slightly less than half the enclosed space was covered as a prayer hall; the rest was left open as a courtyard.

The most notable feature of the Damascus as well as the Medina mosque was the introduction of the miḥrāb, a semicircular niche in the center of the qibla wall (Whelan, Origins). Although the origins and meaning of the miḥrāb remain a matter of intense speculation, the form seems to have had a commemorative function, to judge from a slightly earlier silver dirham decorated with a niche enclosing an upright that has been interpreted as the Prophet's spear (Miles, Mihrab and ʿAnazah). In any event, the miḥrāb immediately became a distinguishing feature of virtually all mosques (Papadopoulo, Le mihrab) and the minbar was, from an early date, placed to the right of the miḥrāb.

At Damascus, the area immediately in front of and beside the miḥrāb, which presumably comprised the caliph's maqṣūra, was architecturally emphasized by a massive gabled bay and dome (see Fig. i). Comparable but more modest forms were used to emphasize the miḥrāb area at Medina. The interior walls of the Damascus and Medina mosques were decorated with mosaics and inscriptions (see epigraphy of the qurʾān); the surviving mosaics at Damascus show a riparian landscape with houses and pavilions which perhaps depicts paradise (q.v.) as it is described in the Qurʾān, but few, if any, other mosques had such specific decoration.

The disposition of some mosques, such as those in Aleppo, Diyār Bakr and Ḥarrān, may have been based more or less closely on the example of the Damascus mosque, but the Umayyads do not seem to have established a standard mosque type. Rather, the Umayyad idea of a mosque appears to have comprised a rather flexible association of constituent parts, which should include (in decreasing order of importance): orientation towards Mecca, a miḥrāb in the qibla wall, open and covered spaces, arcades surrounding the courtyard, domes or raised roofs in the area near the miḥrāb, and a maqṣūra. Some or all of these features can be seen in smaller mosques of the Umayyad period, such as at Qaṣr al-Ḥayr al-Sharqī, Jabal Says, etc.

The ʿAbbāsids, who seized power from the Umayyads in 132/749, appear to have had no doctrinal objection to the mosque as it had evolved under Umayyad patronage — although the ʿAbbāsids did claim that the Umayyads' excessive elevation of the minbar was wrong. The ʿAbbāsids consequently ordered minbars reduced in size but as the Qayrawān minbar (mid-third/ninth century) still has many steps, the order seems not to have been effective. Literary sources indicate that the ʿAbbāsids established mosques in the second half of the second/eighth century at Baghdād, their new capital in Iraq, as well as in other cities. None has survived intact, but they do not appear to have deviated from the Umayyad norm in any significant way. By the early third/ninth century, however, many ʿAbbāsid mosques began to have a single tower located next to the entrance in the wall opposite the miḥrāb.

These towers are traditionally understood to have been places from which the call to prayer (adhān) was given by the muezzin (muʾadhdhin) but there is no evidence to suggest that these towers were erected for this purpose (Bloom, Minaret). In early Islamic times, the first call to prayer was normally given from the doorway or the roof of the mosque; Shīʿīs in particular continued to follow this practice (see shīʿism and the qurʾān). er the later Umayyads, sev eral mosques had a small structure on the roof, known epigraphically as miʾdhana (place for the adhān), which was presumably meant to protect the muezzin from the elements. Only the mosque of Medina had four towers in Umayyad times, and these do not seem to have been used for the call to prayer. If, in later times, towers were often used for the call to prayer, the tower seems to have been introduced into the mosque simply to indicate its presence from afar. Such an interpretation would coincide with the most common Arabic name for these towers, which is manāra, “a place or thing that gives light” (cf. Hebrew menorah), whence (via the Turkish minare) the English word “minaret.”

Like the mosque itself, there was no particular shape the minaret needed to take: the square and battered third/ninth-century tower of the Great Mosque of Qayrawān was modeled on a nearby Roman lighthouse (see Fig. ii), while the contemporary towers attached to the mosques of al-Mutawakkil and Abū Dulaf at Sāmarrāʾ in Iraq are helicoidal spirals, a form invented by ʿAbbāsid builders. In Syria, square stone towers became common, while in Iran, cylindrical and polygonal towers of astonishing height showed off the talents of Iranian builders, particularly in the eleventh and twelfth centuries.

As ablution (wuḍūʾ) is required before ritual worship, many mosques are known to have been provided with facilities for washing, although few such installations have survived the centuries. The ninth/fifteenth-century Egyptian historian al-Maqrīzī (d. 845/1442) preserved a description of the late third/ninth-century ablution pavilion in the courtyard of Ibn Ṭūlūn's mosque in Cairo (which was, incidentally, surmounted by a place for the muezzins). A magnificently-decorated sixth/twelfth-century ablution pavilion has been preserved from the original Almoravid mosque of Marrakesh, although the mosque itself has not survived (Meunié and Terrasse, Nouvelles recherches; see Fig. iii).

Other elements of mosque furniture include the dikka, a platform sometimes found in congregational mosques (for one example, see Fig. vi). They are used on Fridays by muezzins giving the third call to prayer before the ṣalāt in the mosque, as well as by “repeaters” (muballigh) to enable the entire congregation to hear in the pre-loudspeaker age. Reading-stands (kursī) held the large manuscripts of the Qurʾān that were often presented to mosques as pious gifts; some also provided a seat for the reader. Most kursīs were made of wood, elaborately carved and inlaid with colored woods and bone or ivory, but other materials were used. Perhaps the largest is the stone reading stand in the courtyard of the mosque of Bībī Khānum in Samarqand (see Fig. v). It is thought that it was made to hold the enormous manuscript of the Qurʾān whose pages measure over 1 × 2 meters, of which several leaves survive. From an early date, lamps (see lamp) and candlesticks were installed in mosques to provide light (q.v.) at night. Some were made of metal elaborately decorated with piercing and inlaying (Behrens-Abouseif, Metal lamps), while others, particularly in the Mamlūk period, were made of glass enameled with intricate inscriptions and designs (Wiet, Lampes et bouteilles; for one example of the latter, see Fig. iv of material culture and the qurʾān). Some enameled glass Mamlūk lamps were inscribed with the Light Verse (q 24:35), a particularly felicitous choice. Mats or carpets often covered the floors of mosques to prevent the worshiper from getting dusty. The Prophet is said to have initiated the custom of praying on a carpet, although some Muslims reject this practice as a later innovation.

Types of mosques

During the first centuries of Islam, Muslims carried the flexible idea of the hypostyle mosque from its homeland in Syria and Iraq throughout the Muslim world. Before ca. 400/1000, hypostyle or “Arab-type” mosques were erected everywhere from Spain and Morocco in the west to Iran and central Asia in the east. While all share general features of planning and orientation, Muslim patrons and builders were sufficiently flexible to accommodate local traditions of construction and decoration. Thus, the mosque of Cordoba in Spain, begun by Umayyad emigrants from Syria in the late second/eighth century, used a distinctive two-tiered system of supports and mixed construction of recycled stone and brick to support tile-covered gabled roofs. The third/ninth-century builders of the mosque at Qayrawān continued local Tunisian traditions of fine ashlar construction, using recycled antique stone columns and capitals to support a flat timber roof. In most of Iran, brick, whether sun-dried or fired, became the major material of construction for supports and coverings; the scarcity of wood had led Iranian builders in previous centuries to develop ingenious techniques for covering large spaces with brick vaults. In some areas of Iran and Anatolia, on the other hand, where timber continued to be available after the Muslims first settled the region in the late fifth/eleventh century, builders developed a timber-framed hypostyle mosque. Two examples are the small village mosque at Abyāna (Iran; before 1103 c.e.) or the Eşrefoǧlu mosque at Beyşehir (696/1296), although this structure, like many other Anatolian mosques, is enclosed in stone walls. In the Maghrib the hypostyle type of mosque became typical, and its popularity excluded virtually all other types.

The inherent flexibility and adaptability of the hypostyle mosque made it so popular in such a wide variety of situations over such a long period that the name “Arab-type” mosque, which is sometimes used, is patently unsuitable. The plan, with some variation, is found for example in an early mosque in west Africa (e.g. the eighth/fourteenth-century Djingere-Ber mosque at Timbuktu) as well as a modern one in the same region (the Great Mosque of Mopti built in 1935). It is also found in east Africa (e.g. the Great Mosque, Kilwa Kisawāni; begun in the sixth/twelfth century), India (the Quwwat al-Islam mosque, Delhi; begun 592/1196), China (e.g. Yangzhou mosque, Jiangsu Province; begun 673/1275[?]) and southeast Asia (e.g. Masjid Agung, Demak; founded 881/1477); and modern architects continue to exploit its structural possibilities, as in the Mosque of the King (Marbella; 1981) or the King ʿAbd al-ʿAzīz mosque (Casablanca; 1983; Frishman and Khan, The mosque).

Most mosques surviving from the early centuries of Islam are large structures intended for congregational worship, but several smaller mosques also survive from the second/eighth century and later. Found in such cities as Toledo (Spain), Qayrawān and Sūsa (Tunisia), Fusṭāṭ (Egypt), Termez (Uzbekistan), and Balkh (Afghanistan) as well as in rural areas of Arabia and central Asia, this type of mosque is characterized by nine square units arranged in a 3 × 3 grid, with four columns or piers supporting the roof. The germ of the type is found in the Umayyad mosques of Khān al-Zabīb, Umm al-Walīd and Qaṣr al-Ḥallābāt. The widespread popularity of this “nine-bay” plan suggests that it, like the hypostyle mosque, was diffused from some central source (King, Nine bay).

The hypostyle mosque, whether large or small, may have been the most common type in early Islamic times, but literary evidence suggests that other types of struc tures were also used where available.

In Iran, where free-standing domed chambers and vaulted halls had been essential elements of the architectural vocabulary in pre-Islamic times, it is possible that pre-existing domes and īwāns (a barrel-vaulted hall open at one end) would have been adapted for use as mosques, much as how, in Syria, basilicas had been transformed into mosques. The crucial archaeological evidence for this transformation is lacking, however, perhaps because many of these structures were built of mud-brick, and fell into dust once abandoned. The undated domed mosques at Yazd-i Khwāst and Qurva may be recycled older buildings. The prayer hall of the Iranian mosque at Nayrīz (perhaps begun 363/973) is a large single īwān which has a miḥrāb at one end and is open to the courtyard at the other.

In the late fifth/eleventh century, however, the Saljūq rulers of Iran, who made their capital at Iṣfahān, transformed the third/ninth-century hypostyle mosque of that city by removing many of the columns in the area immediately in front of the miḥrāb and inserting a huge freestanding brick dome in their place. The building of the Iṣfahān dome in 479-80/1086-7 by the powerful vizier Niẓām al-Mulk during the reign of Malik-shāh has been shown to have been directly inspired by Malik-shāh's restoration of the maqṣūra dome in the Damascus mosque, which had been destroyed by fire a few years before (Blair, Surveyor). While the idea of the great dome may have been inspired by the precedent of Damascus, the form this new dome took was dependent on earlier buildings in central Iran, such as the freestanding domed tomb in Yazd known as the Twelve Imāms (429/1038). Indeed, the Iṣfahān dome was initially freestanding within the hypostyle hall, but such a structure was both structurally unstable and visually awkward, so a massive īwān was soon erected in front of it. This īwān physically and visually linked the courtyard to the maqṣūra dome, and the flanking hypostyle halls were connected to the new parts. The īwān's court facade was embellished by a high rectangular frame, known as a pīshṭāq, which was, perhaps at a later date, flanked by slender towers. By the early sixth/twelfth century, three other īwāns had been erected at the midpoints of the three other sides of the court, so that the mosque had taken on an entirely new aspect. In place of the hypostyle halls surrounding the spacious courtyard, there was now a massive arched īwān in each of its four sides; that on the qibla side terminated in a huge dome over the miḥrāb.

The prestige of the mosque in Iṣfahān (see Fig. iv), which was the Saljūq capital, coupled with the inherent flexibility of the four-īwān plan, which had been used for centuries on a much smaller scale in residential and palatine architecture, led builders throughout the Saljūq realm to copy the developments at Iṣfahān. Older hypostyle mosques were transformed by the addition of a dome and one or more īwāns (e.g. Ardistān), and new mosques were built from scratch using the new plan (e.g. Zawāra). Indeed, by the eighth/fourteenth century, the four-īwān mosque had become the Iranian mosque type, and such plans were used in various scales from the modest (Varamīn, 722/1322 and later) to the mammoth (Tīmūr's mosque of Bībī Khānum in Samarqand, 802-3/1399), where even the lateral īwāns were domed. The four-īwān type continued to be the most popular in later periods in Iran and central Asia, as for example in the Kalān mosque (920/1514) in Bukhārā or the Masjid-i Imām (formerly Masjid-i Shāh) erected by the Safavid Shāh ʿAbbās between 1021/1612 and 1040/1630 in Iṣfahān.

As Iranian cultural norms were prevalent in most of the eastern Islamic lands in the period after ca. 650/1250, the Iranian combination of an īwān leading to a dome over the miḥrāb, or even the four-īwān type plan, was widely disseminated, although, like the hypostyle mosque before it, its features were often creatively reinterpreted, as at the Ūlū Cāmiʿ (621/1224) at Malaṭya in southeast Anatolia, where the courtyard is very small, and the mosque of Baybars I (665/1266) in Cairo, where the īwān is a sort of hypostyle hall and the dome was built of wood.

The Iranian type of four-īwān mosque was used on the Indian subcontinent, as for example at Thatta (1054/1644), but the essential features of the plan were more commonly adapted to create a new Indian type of mosque more suited to the climate. Like earlier Hindu temples, this type of mosque is often raised on a high plinth. It comprises a vast walled court with minarets set at the exterior corners and massive portals on the main and lateral axes. On the exterior, these are approached by flights of steps and are topped with small minaret-like towers and open pavilions. Within the court, which often has no surrounding arcade, the prayer hall occupies most but not all of the qibla wall's width. Projecting into the court, the fairly shallow prayer hall comprises a central īwān-like portal leading to a dome chamber over the miḥrāb and minbar. This īwān and dome unit is usually flanked by smaller versions of it, other domed bays and towers linked together behind a screen-like facade. An early example is found at the Atala mosque at Jaunpur (810/1408; see Fig. vii); a later one is the Badshahi mosque at Lahore (1084/1673), said to be the largest mosque in the subcontinent.

Another distinctive type of mosque, characterized by a single large dome, developed in Anatolia, particularly under the Ottomans (r. 1281-1924 c.e.). After the region was opened for Muslim settlement follow ing the Battle of Manzikert in 463/1071, the first mosques erected, such as the ʿAlāʾ al-Dīn mosque in Konya (550-617/1155-1220), were hypostyle structures. Perhaps in response to the severe Anatolian winter climate, these mosques had no courtyard, or only a vestigial one. Most early Anatolian mosques were, therefore, closed halls resting on a multitude of columns, sometimes fronted by courtyards akin to the forecourts of some Byzantine churches. Indigenous Byzantine experience with building domed and vaulted churches in stone, combined with the knowledge of Iranian traditions of building brick vaults and domes brought by the Saljūqs, undoubtedly encouraged local builders to experiment with the covering of mosques with domes and vaults; the subsequent history of the mosque in Anatolia, particularly under the Ottomans, is dominated by the desire to create a unified prayer space under a domical covering.

Scholars have debated the origins of the mature type of Turkish domed mosque. Some see its origins in the simple, single domed mosque preceded by a portico, such as the Mosque of Ḥacī Özbek at Iznīk (734/1333), while others see it in the more complex organization of domed and vaulted elements characteristic of the Bursa-type or zāwiya (Turk. zaviyelī) mosque usually associated with a dervish cloister, in which a domed central space precedes a vaulted or domed īwān-like prayer hall and is flanked by corresponding īwān-like spaces to the left and right which could be used for teaching, etc. (see ṣūfism and the qurʾān). Other vaulted or domed chambers in the corners, often furnished with fireplaces, could be used for lodging itinerant dervishes. This completely covered mosque type, well suited to the harsh Anatolian climate, is exemplified by the Yeşil Cāmiʿ (815-22/1412-19) at Bursa.

In the ninth/fifteenth and tenth/six teenth centuries the Ottoman sultans sponsored a series of immense domed congregational mosques which combined various trends, including the single domed space (as at Iznīk), the completely covered space (as at Bursa), the large dome before the miḥrāb (as in Iranian mosques) and the open arcaded forecourt (as in the Great Mosque of Damascus or Byzantine churches). These buildings were the centerpieces of large religious and charitable foundations. The earliest examples, such as the Üç Şerefeli mosque in Edirne (841-51/1437-47), have vast central domes with low subsidiary spaces, but the best known examples represent creative responses to the models of Byzantine architecture, primarily the great church of Hagia Sophia in Istanbul, which had been founded by the emperor Constantine in the fourth century c.e. and rebuilt in the sixth century by the emperor Justinian. Immediately after the conquest of Constantinople in 857/1453 by the Ottoman emperor Meḥmed II the church was converted into the city's first congregational mosque by the addition of a miḥrāb, minbar and minarets. The Süleymāniye mosque in Istanbul (964/1557), the masterpiece of Ottoman architecture by the great architect Sinān, for example, follows the model of Hagia Sophia by using an immense central dome buttressed at either end by semi-domes. The many windows create a highly centralized space, which contrasts with the strongly directional space of the prototype. Unlike many mosques of earlier times, in which the exterior facade was often neglected in favor of a focus on the interior or courtyard, the exteriors of the imperial Ottoman mosques were clearly meant to be monumental, presenting cascades of domes and semi-domes punctuated by slender minarets. The central mass is often enveloped in the domes of the adjacent religious and charitable foundations, which frequently included the tomb of the founder.

The power and prestige of the Ottoman empire in the Balkans, north Africa and the Near East encouraged the construction of similar if somewhat simpler mosques (and complexes) in the capital cities of the Ottoman empire. Sometimes these structures incorporated local architectural motifs and techniques, such as the striped masonry used in the Sulaymāniyya complex at Damascus, completed in 962/1554-55, or the ogee windows of the tenth/sixteenth-century mosque of Sinān Pasha in Cairo. In the nineteenth and twentieth centuries c.e., the Ottoman type of domed mosque has been adapted throughout the world and has become one of the most popular designs for new mosques. These can range from banal copies of Sinān's masterpieces in reinforced concrete to inventive reinterpretations in modern materials, such as Vedat Dalakoy's State Mosque (1970-86) in Islamabad, Pakistan, or Skidmore, Owings and Merrill's Islamic Cultural Center (1987-91) in New York (Holod & Khan, Contemporary mosque).

Distinct types of mosques have also developed in other areas of the Islamic world. In Indonesia, for example, the Java-type mosque is set in a courtyard surrounded by a stone wall. It has a veranda on the front, and is characterized by a square plan, raised foundations, tapering roof two to five stories high, and a projecting miḥrāb. In China, many mosques, particularly those built under the Ming emperors, are reminiscent of indigenous Chinese temples. Other distinct styles have also evolved in east and west Africa.

Decoration of mosques

Virtually all media of Islamic art are represented in the decoration of mosques, with the one proviso that Islam forbids the representation of animate beings (i.e. humans and animals) in such situations where they might be taken as objects of devotion, as in mosques. Thus, one rarely, if ever, finds pictorial or sculptural representations in mosques, whose main decoration has consisted of inscriptions, often from the text of the Qurʾān, and vegetal and geometric designs. In the early period, the interior decoration of major Umayyad mosques, such as those in Mecca, Medina, Jerusalem and Damascus, adapted the techniques and many of the motifs of Byzantium, so colored and gold glass mosaic was popular. In later times, carved and molded plaster, carved and painted wood, glazed ceramic tile and marble paneling were commonly used media of decoration.

In many times and places, the interior of the mosque has been more important than the exterior, and the decoration has been concentrated largely on the inside. This may have resulted from the idea that the mosque is centered around the courtyard, and consequently the building is planned and intended to be seen from the courtyard outwards. The result is that exteriors were neglected, often irregular, and hardly distinguished from the surrounding urban fabric. At first, doorways were simple affairs with no great decoration, but eventually they became places of some importance, perhaps following the lead of the portals to the “sacred mosque” (al- masjid al-ḥarām) in Mecca, which were monumentalized at an early date, and eventually projected from the mosque wall itself. The earliest fancy portal to survive is the Bāb al-Wuzarāʾ, “Vizier's Gate” (241/855-6) at the Great Mosque of Cordoba, in which the doorway has been embellished with inscriptions and carving (Brisch, Zum Bāb al-Wuzarāʾ). Congregational mosques of the Fāṭimid period emphasize the doorways, probably because of the Fāṭimid aversion to building minarets. The triple-arched portal of the Great Mosque at Mahdiyya (308/921), the first Fāṭimid mosque in Tunisia, is modeled on a late antique triumphal arch (Bloom, The origins). Archaeologists working in Iṣfahān in the 1930s discovered a large portal in baked brick and carved plaster. On the basis of its style, it has been identified as the portal of the Jurjir mosque built by the Būyid vizier Ibn ʿAbbād in the third quarter of the fourth/tenth century (Blair, Monumental inscriptions, 52-3).

Qurʾānic inscriptions in mosques were usually prepared as coherent programs of decoration, although few complete programs have survived and the meaning is usually inferred from surviving fragments. The mosaic inscriptions in the Great Mosque of Damascus, for example, are known to have focused on eschatological texts about the day of judgment (sūras 78 and 79 of the Qurʾān; Finster, Die Mosaiken; see eschatology; last judgment; apocalypse), while the stucco inscriptions carved in the walls of the mosque of al-Azhar in Cairo (founded 363/972) used verses such as q 21:101-7 to describe the paradise that awaited the true believers (see belief and unbelief). Other decorative programs seem to have been somewhat less selective: medieval sources assert that the carved wooden friezes running under the ceiling of the mosque of Ibn Ṭūlūn in Cairo (265/879) repeated the whole text of the Qurʾān, and the inscriptions in the mosque of al-Ḥākim, also in Cairo, contain the opening verses of many different chapters (Bloom, Mosque of al-Hakim).

Nevertheless, specific qurʾānic verses were inscribed in appropriate situations. The most common qurʾānic text used for inscriptions in mosques is q 9:18, which states that God's mosques should be reserved for good Muslims who believe in God, pray, pay alms (see almsgiving), and worship God alone (Dodd and Khairallah, The image; Blair, Mosque inscriptions). The text, one of the few in the Qurʾān to actually mention mosques and what should be done within them, quickly became popular in congregational mosques. Other qurʾānic citations commonly found in mosques include the Throne Verse (q 2:255; see throne of god), which extolls God's majesty and is often used around domes, the Light Verse (q 24:35-8), which describes God as the light (q.v.) of the heavens (see heaven and sky) and the earth (q.v.), and Sūrat al-Fatḥ (“Victory,” q 48; see victory). Although the Light Verse is often associated, because of a mistranslation, with the common motif of a lamp in a niche, certain verses were often associated with specific parts of mosques. Miḥrābs often contained verses q 17:78-9, which mention prayer and vigil (see vigils).

Qurʾānic verses were also selected to give a particular inscription a specific ideological position. An inscription once on the exterior of the mosque of al-Ḥākim (380-405/990-1013) quotes q 28:5, which uses the word aʾimma, the plural of imām, the title by which the Fāṭimid rulers styled themselves. Similarly, an inscription on one of the same mosque's towers cites q 11:73, containing the phrase ahl al-bayt (“people of the house”), which the Fāṭimids interpreted as a direct reference to themselves as the descendants of the Prophet (Bloom, Mosque of al-Hakim). Similarly, the portal of the Jurjir mosque in Iṣfahān (ca. 350/960) is inscribed with q 3:18, in which the use of the word qisṭ may have been chosen to advertise the building's function as a Muʿtazilite foundation (Blair, The octagonal pavilion; see muʿtazilīs). As sectarian struggles increased over the course of the fourth/tenth and fifth/eleventh centuries, such pointed uses of qurʾānic inscriptions only increased, but in later times verses with more general application came to be expressed in monumental tiled and painted inscriptions.

Jonathan M. Bloom

Bibliography

C.H. Becker, Die Kanzel im Kultus des Alten Islam, in C. Bezold (ed.), Orientalische Studien Theodor Nöldeke zum siebzigsten Geburtstag (2. März 1906) gewidmet von Freunden und Schülern, 2 vols., Gieszen 1906, i, 331-51

D. Behrens-Abouseif, Mamluk and post-Mamluk metal lamps, Cairo 1995

S. Blair, The monumental inscriptions from early Islamic Iran and Transoxiana, Leiden 1992

id., Mosque inscriptions, in Eric M. Meyers (ed.), The Oxford encyclopedia of archaeology in the Near East, 5 vols., New York 1997, iv, 58-60

id., The octagonal pavilion at Natanz.

A reexamination of early Islamic architecture in Iran, in Muqarnas 1 (1983), 69-94

id., Surveyor versus epigrapher, in Muqarnas 8 (1991), 66-73

J.M. Bloom, Minaret. Symbol of Islam, Oxford 1989

id., The mosque of al-Hakim in Cairo, in Muqarnas 1 (1983), 15-36

id., The origins of Fatimid art, in Muqarnas 3 (1985), 20-38

K. Brisch, Zum Bāb al-Wuzarāʾ (Puerta de San Esteban) der Hauptmoschee von Córdoba, in Studies in Islamic art and architecture in honour of professor K.A.C. Creswell, Cairo 1965, 30-4

K.A.C. Creswell, Early Muslim architecture. Volume I, Oxford 19692

id., A short account of early Muslim architecture. Revised and enlarged by James W. Allan, Aldershot 1989

E.C. Dodd and Sh. Khairallah, The image of the word. A study of qurʾānic verses in Islamic architecture, Beirut 1981

B. Finster, Die Mosaiken der Umayyadenmoschee, in Kunst des Orients 7 (1970), 83-141

F.B. Flood, The great mosque of Damascus. Studies in the making of an Umayyad visual culture, Leiden 2001

M. Frishman and H. Khan, The mosque. History, architectural development and regional diversity, London 1994

O. Grabar, The formation of Islamic art, New Haven 1973

R. Hillenbrand, Islamic architecture. Form, function and meaning, Edinburgh 1994

R. Holod and H. Khan, The contemporary mosque. Architects, clients and designs since the 1950s, New York 1997

G.R.D. King, The nine bay domed mosque in Islam, in Madrider Mitteilungen 30 (1989), 332-90

J. Meunié and H. Terrasse, Nouvelles recherches archéologiques à Marrakech, Paris 1957

G.C. Miles, Mihrab and ʿanazah.

A study in early Islamic iconography, in G.C. Miles (ed.), Archaeologia orientalia in memoriam Ernst Herzfeld, Locust Valley, NY 1952, 156-71

A. Papadopoulo, Le mihrab dans l'architecture et la religion musulmanes. Actes du Colloque International tenu à Paris en Mai 1980 publiés et pourvus d'une étude d'introduction générale, Leiden 1988

J. Pedersen, Masdjid, in ei2, vi, 644-77

J. Sauvaget, La mosquée Omeyyade de Médine. Étude sur les origines architecturales de la mosquée et de la basilique, Paris 1947

E. Whelan, The origins of the mihrab mujawwaf. A reinterpretation, in ijmes 18 (1986), 205-23

G. Wiet, Lampes et bouteilles en verre émaillé. Catalogue générale du Musée Arabe du Caire, Cairo 1912

Cite this page

Bloom, Jonathan M., “Mosque”, in: Encyclopaedia of the Qurʾān, General Editor: Jane Dammen McAuliffe, Georgetown University, Washington DC. Consulted online on 07 January 2017 <http://dx.doi.org/10.1163/1875-3922_q3_EQCOM_00125>