2 The detailed accounts of Maḥmūd's campaign on Somnath are not, incidentally, well known. Baihaqī's Tārīkh-i Maḥmūdī, which recorded the events of the campaign, is lost, and the other authentic source of Maḥmūd's history, al-ʿUtbī's al-tārīkh al-yamīnī, ends before Maḥmūd's venture to Somnath. Later historians have used Baihaqī as their source, and give only summary – and sometimes exaggerated – accounts of the campaign. See Abū ʿUmar Minhāj al-dīn ʿUthmān b. Sirāj al-dīn al-Jauzjānī (Minhāj-i Sirāj), Ṭabaqāt-i Nāṣirī, ed. Habibi, Abd al-Hai (Tehran, 1984) (henceforth Minhāj-i Sirāj), I, 229–30Google Scholar; Muḥammad b. Khāwand Shāh called Mīr Khwand, Rauḍat al-ṣafā (Tehran, 1270/1853–54)Google Scholar, no page number, but under dhikr-i fatḥ-i sūmanāt bi dast-i Maḥmūd (henceforth Mīr Khwand); Ghiyāth al-dīn b. Humām al-dīn al-Ḥusainī known as Khwand Mīr, Tarīkh-i Ḥabīb al-siyar (Tehran, 1976), (henceforth Khwand Mīr) II, 382–3Google Scholar; Ḥamd'ullāh Mustaufī's Ẓafarnāma (completed in 807/1404–05) in Ẓafarnāma von Ḥamdallāh Mustaufī und Šāhnāma von Abu'l-Qāsim Firdausī (editorisch bearbeitet von Ḥamdallāh Mustaufī), Faksimile – Ausgabe der Handschrift der British Library (Or. 2833), (Iran University Press und Verlag der Österreichischen Akademie der Wissenschaften, 2 vols, Tehran, 1377/Vienna, 1999), I, 578–9. The Muslim historians of India repeat these accounts, often with some exaggeration. See ʿIsāmī, Futūḥ al-salāṭīn, ed. Agha Mahdi Husain (Agra, 1938), 33–40; Muḥammad Qāsim b. Hindū Shāh known as Firishta: Gulshan-i Ibrāhīmī known as Tārīkh-i Firishta (2 vols with addenda bound together, Lucknow, 1864) (henceforth Firishta), I, 32–4; Khwāja Niẓām al-dīn Aḥmad b. Muḥammad Hirawī, Ṭabaqāt-i Akbarī (Persian text, 3 vols, Biblioteca Indica no. 223, Calcutta: Asiatic Society of Bengal, 1927–35), I, 16–7.

3 For a select bibliography of the literary sources see Aliakbar Dehkhodá, Loghatnáme (Encyclopedic Dictionary) (14 vols), ed. Mohammad Moʿin and Jaʿfar Shahidi (Tehran, SH, 1372–73/1993–94), under the entry sūmanāt (); Dehkhodá's Dictionary is also available online at http://www.loghatnameh.com.

4 Shaikh Muṣliḥ al-dīn Saʿdī, Būstān in Kulliyāt-i Saʿdī, ed. Muḥammad ʿAlī Furūqī, (Tehran, 1363/1984), 374–7Google Scholar. For another edition see Būstān-i Saʿdī, ed. Iranparast, Nur'ullah (Tehran, 1977), 334–42Google Scholar. For the translations of this tale see The Bostan of Shaikh Sadi, tr. Ziaūddin Gulam Moheiddin Mūnshi, , revised Davies, Rochfort (Bombay, 1889), 245–51Google Scholar; Wisdom of the East, the Bustān of Sadi, tr. Edwards, A. Hart (London, 1911), 106–9Google Scholar. A translation of the concluding part of the tale is also given by Browne, Edward Granville, A Literary History of Persia (4 vols, London, 1909–24), II, 1906, 529–30Google Scholar.

5 A reference to pre-Islamic Arabian society, religion and culture which included the worship of the deity manāt.

6 Farrukhī, Dīwān-i Ḥakīm Farrukhī Sīstānī, ed. Dabir-Siyaqi, M. (Tehran, 1335/1956), 69–71Google Scholar (henceforth Farrukhī). Farrukhī's contemporary, the poet Sanā’ī, also alludes to the story in “Abu'l-majd Majdūd b. Ādam al-Sanā'ī al-Ghaznawī”, Ḥadīqat al-ḥaqīqa wa sharī ʿat al-ṭarīqa (ed. Razavi, Mudarris) (Tehran, no date but c. 1983), 512Google Scholar:

7 Abū Saʿīd ʿAbd al-Ḥayy b. al-Ḍaḥḥāk b. Maḥmūd Gardīzī, Zain al-akhbār, ed. ʿAbdul-Hayy Habibi, (Tehran, 1347/1968), 190 (henceforth Gardīzī)Google Scholar.

8 Alberuni's India (Taḥqiq ma li'l-Hind) (ed. Sachau, E. C.) (Arabic text, London, 1887), 252; (English tr. Sachau, London, 1888), II, 102–3Google Scholar. Sachau's translation is given here.

9 For one of Maḥmūd's ventures to north India plundering cities on his way to Qanūj and sacking the temples see Muḥammad b. ʿAbd al-Jabbār al-ʿUtbī, Al-tārīkh al-yamīnī, in Aḥmad b. ʿAlī Manīnī, Al-fatḥ al wahabī ʿalā tārīkh Abī Naṣr al-ʿUtbī li'l-Shaikh al-Mutanabbī (Cairo, 1286/1869–70), II, 270–8Google Scholar; Abū Sharaf Nāṣiḥ b. Ẓafar Jurfādiqānī, Tarjuma-yi tārīkh-i yamīnī (Tehran, 1978), 379–86Google Scholar (henceforth Jurfādiqānī).

10 Khwand Mīr, II, 383.

11 Maḥmūd's booty from the campaign to Qanūj was, according to Gardīzī's account (184), over twenty million dirham, 53,000 slaves and over 350 elephants; and according to Jurfādiqānī (386) three million dīnār and so many slaves that the market value of slaves fell to between two and ten dirham.

12 For an extensive historical study of Maḥmūd's campaign to Somnath and its echoes in later sources see Davis, Richard H., Lives of Indian Images (Princeton: Princeton University Press, 1997), 88–112, 209–21Google Scholar; Davis, Richard H., “Memories of broken idols”, in Bierman, Irene A. (ed.), The Experience of Islamic Art on the Margins of Islam (Reading: Garnet, 2005), 133–68Google Scholar. For a thorough study of Somnath's history through Indian sources and the repercussions of Maḥmūd's episode on communal tensions in modern Indian society see Thapar, Romila, Somnatha: The Many Voices of a History (New Delhi: Penguin Books India, 2004)Google Scholar; see also Thapar, Romila, “Somanatha and Mahmud”, Frontline 16/8 (April 23, 1999), 121–7Google Scholar. For a fresh and valuable study of the concept of the “infidel” in Islam and its interpretation and effect in Muslim India see Flood, Finbarr Barry, Objects of Translation: Material Culture and Medieval “Hindu–Muslim” Encounter (Princeton and Oxford: Princeton University Press, 2009)Google Scholar. This work also discusses the episode of Maḥmūd's campaign to Somnath, see pp. 77–87.

13 The temples of Somnath and other towns of Saurashtra were studied early in the twentieth century by Henry Cousens, who was not primarily concerned with the Muslim edifices, but reported a few of the major mosques in Saurashtra and briefly noted other remains. See Cousens, Henry, Somanātha and Other Mediaeval Temples in Kāṭhiāwād (Calcutta: Archaeological Survey of India, Imperial Series, XLV, 1931)Google Scholar. Monuments of some of the other towns of Saurashtra are studied by Burgess, James, Report on the Antiquities of Kaṭhiawād and Kachh (London: Archaeological Survey of India, Western Circle, II, 1876)Google Scholar. Many Muslim edifices in Saurashtra (Kathiāwād) – some of considerable age – are omitted from these works, an example is the mosque of Abu'l-Qāsim al-Iḍḥajī, noted below. The present paper considers only the major early sultanate mosques at Somnath, but in Somnath and elsewhere in Saurashtra there are many other Muslim edifices awaiting study.

14 Cousens, Somanātha, 33, pl. 22 does not mention the larger reservoir, but notes the temple and the smaller tank as Bhalkeśvara Talāv or Bhalka Tīrth (the tank of the arrows) and remarks: “It is a pool of slimy water surmounted with rough stone steps, which may or may not be very old; certainly the little temple, beside it, is of no great age”. It should be noted that in India the term talāo or tālāb usually refers to large-sized reservoirs and in Gujarat smaller step-wells or stepped tanks – known in north India as bā’olī – are referred to as wav.

15 Burgess, James, Muhammadan Architecture of Ahmadabad, Part I: AD 1412 to 1520 (London: Archaeological Survey of India (New Imperial Series), XXIV, Western India, VII, 1900), 52–3, pl. 65Google Scholar.

16 Based on the square sarvatobhadra or circular nandiāvarta diagrams, see Ananthalwar, M. A. and Rea, Alexander (eds) and Iyer, A. V. Thiagaraja (comp.), Indian Architecture, I, Architectonic or the Silpa Sastras (Delhi: Indian Book Gallery, 1980), 141–3Google Scholar. Alexander Kinloch Forbes, who visited the town when more of the walls were preserved, reports that “the walls form an irregular four sided figure of which the south side follows the line of the bay for some time, within a few feet of high-water mark”. See Forbes, A. Kinloch, “Puttun Somnath”, Journal of the Royal Asiatic Society, Bombay Branch, 8, 1864–66 (Bombay and London, 1872), 51Google Scholar. What remains of the walls and their curvature, as well as the built settlement, still circular in form, indicates that an ancient circular plan cannot be ruled out. Indian towns with nandiāvarta plan are not usually laid on a perfect circle. See for example the arrangement of the town of Warangal in Michell, George, “City as cosmogram: the circular plan of Warangal”, SAS 8, 1992, 1–18Google Scholar.

17 A. Kinloch Forbes, “Puttun Somnath”, 50.

18 Farrukhī, 71–2; Gardīzī, 191; Mīr Khwand (see note 2); Minhāj-i Sirāj, I, 229.

19 Abū Isḥāq Ibrāhīm b. Muḥammad al-Iṣṭakhrī, Masālik wa mamālik (Persian text, ed. I. Afshar, Tehran, 1961), 147, 151Google Scholar, Al-masālik wa al-mamālik (Arabic Text) (Cairo, 1961), 102, 104–5Google Scholar; Ḥudūd al-ʿālam min al-mashriq il'l-maghrib (Tehran: ed. M. Sotoodeh, 1962), 66Google Scholar; Abū ʿAbd'ullāh Muḥammad b. Aḥmad al-Muqaddasī, Aḥsan al-taqāsīm fī maʿrifat al-aqālīm (Leiden, 1906), 477, 484, 486Google Scholar; Ḥauqal, Ibn, Ṣurat al-arḍ (Leiden, 1872), 227–8, 232–3Google Scholar. For a recent study of the historic ports of western India see Lambourn, Elizabeth, “India from Aden: Khutba and Muslim urban network in late thirteenth century India”, in Hall, Kenneth R. (ed), Secondary Cities and Urban Networking in the Indian Ocean Realm c. 1000–1800 (Lanham MD and London: Lexington Books, 2008), 55–97Google Scholar.

20 Abū ʿAbd'ullāh Muḥammad b. Muḥammad al-Idrīsī, Opus Geographicum (Naples and Rome, 1971), 185Google Scholar.

21 Indian Antiquary 8, 153.

22 The shrine is at the north of a sizeable enclosure and consists of three old chambers, but none of architectural merit. The tomb of Mangrolī Shāh is in the west and the chamber to its east was originally a colonnaded portico, which has been walled up. To the west of this chamber is a small mosque with a single miḥrāb and two columns with corresponding pilasters on the north and south walls surmounted by lintels supporting the flat roof of the prayer hall. The mosque is of considerable age and has a colonnaded entrance portico much in the style of the buildings of the maritime settlers, but has been many times restored making it difficult to establish if the building originally dates from prior to the Muslim conquest of the region. To the south-east of the complex is another small mosque, which judging from the style of its miḥrāb seems to date from the time of the Gujarat sultanate. There are also some modern buildings in the enclosure.

23 The inscription was first reported in Annual Report of Indian Epigraphy (1954–55), C. 168, without giving its text. An ink rubbing of the epitaph, with a description of the tombstone, but without mention of the content of the inscription, is also given in Chhabra, B. Ch., Sircar, D. C. and Desai, Z. A., “Inscriptions from Mantai Tirukeśwaram, Mannar District and from the tomb of Mangrolī Shāh at Veraval” in Epigraphical research, Ancient India 9, 1953, 228–9Google Scholar, pl. 113 b. The lower part of the ink rubbing paper seems to have been folded and the ink smudged, giving the impression that the lower part of the stone was cracked and the three last lines damaged, but as can be seen from our photograph there is no damage to this part of the stone. This ink rubbing has also been reproduced in Lambourn, Elizabeth, “Carving and communities: marble carving for Muslim patrons at Khambhāt and around the Indian Ocean rim, late thirteenth–mid-fifteenth century”, Ars Orientalis 34, 2004, 102Google Scholar, but the content of the inscription is not given. For a bibliography of the inscription see Desai, Z. A., Arabic, Persian and Urdu Inscriptions of West India (New Delhi, 1999), 203Google Scholar, inscription no. 1886. Desai gives the date as 1 Rabiʿ II 699/26 December 1299.

24 The letters of the date are given without any dots and as the words for seven (سبع), nine (تسع), seventy (سبعین) and ninety (تسعین) are similar, other combinations of the date such as 679, 697 and 699 could also be suggested, but judging from the form of the letter sīn in other parts of the inscription the reading given above seems more likely.

25 Porter, Venetia, “Three Rasulid tombstones from Ẓafār”, JRAS, 1988, 32–44Google Scholar; Yatim, Othman Mohd. and Nasir, Abdul Halim, Epigrafi Islam terawal di nusantara (Kuala Lumpur, 1990), 21–3, 32, 36, pls. 4.6–8, 4.18Google Scholar; Gallop, Annabel Teh, Early Views of Indonesia, Drawings from the British Library (London and Jakarta, 1995), 54–5, fig. 42 and pl. 14Google Scholar; Lambourn, Elizabeth, “The decoration of the Fakhr al-dīn mosque in Mogadishu and other pieces of Gujarati marble carving on the East African coast”, Azania 34, 1999, 61–86, particularly pls. 4–5CrossRefGoogle Scholar; Shokoohy, M., Muslim Architecture of South India, the Sultanate of Maʿbar and the Traditions of the Maritime Settlers on the Malabar and Coromandel Coasts (Tamil Nadu, Kerala and Goa) (London and New York: Routledge Curzon, 2003), 138–9, 248, pl. 5.2Google Scholar; Lambourn, Elizabeth, “Carving and communities”, Ars Orientalis 34, 2004, 101–35Google Scholar.

26 Translation given from Arberry, Arthur J., The Koran Interpreted (2 vols. London and New York, 1955), I, 71–2Google Scholar; for another translation see Muhammad Ali, Maulana, Translation of the Holy Quran (Lahore, 1938), 52Google Scholar.

27 The two inscriptions have been separated. The Sanskrit version is kept in Harasiddha Mātā at Veraval and the Arabic version is now set into a wall of the Qāḍī Masjid, a fairly recent structure at Veraval. For the Sanskrit text see Hultzsch, Eugen, “A grant of Arjunadeva of Gujarat dated 1264 AD”, Indian Antiquary 11, 1882, 241–5Google Scholar; Sircar, Dinesh Chandra, “Veraval inscription of Chaulukya-Vaghela Arjuna, 1264 A.D.”, Epigraphia Indica 34/4, 141–50Google Scholar; see also Burgess, James and Cousens, Henry, Revised List of Antiquarian Remains in the Bombay Presidency (Bombay, 1897), 251–2Google Scholar. For the Arabic version see: Bhavnagar Archaeological Department, Corpus Inscriptionum Bhavnagari: Being a Selection of Arabic and Persian Inscriptions Collected by the Antiquarian Department Bhavnagar State (Bombay, 1889), 28–30Google Scholar (wrongly attributes the inscription to the Gujarat sultan Maḥmūd I); Desai, Z. A., “Arabic inscriptions of the Rajput period from Gujarat”, Epigraphia Indica Arabic and Persian Supplement (EIAPS), 1961, 10–15, pl. 2Google Scholar. For a bibliography of the inscription see Desai, Z. A., Arabic, Persian and Urdu Inscriptions of West India (New Delhi, 1999), 203Google Scholar, inscription no. 1885; also see Patel, Alka, “Transcending religion; socio-linguistic evidence from Somnatha-Veraval inscription”, in Parker, Grant and Sinopoli, Carla (eds), Ancient India in Its Wider World (Ann Arbor: University of Michigan Press, 2008) 143–64Google Scholar; Lambourn, Elizabeth, “India from Aden”, in Hall, Kenneth R. (ed.), Secondary Cities and Urban Networking in the Indian Ocean Realm c. 1000–1800, 76–7Google Scholar.

28 Quran, IX, 18. Translation from Arberry, I, 209; also see translation from Maulana Muhammad Ali, 193.

29 Ulugh Khān's campaign on Somnath is described by ʿAlā al-dīn Khaljī's court poet Amīr Khusrau Dihlawī, Khazā’in al-futūḥ (ed. Syed Moinul Haq. Aligarh: Aligarh Muslim University, no date but c. 1927), 50–53; also see Ḍiyā' al-dīn Barnī, Tārīkh-i Fīrūz Shāhī (Calcutta: Bibliotheca Indica, no. 33, Asiatic Society of Bengal, 1862) (henceforth Barnī), 251; Yaḥyā b. Aḥmad b. ʿAbd'ullāh al-Sihrindī, Tārīkh-i Mubārak Shāhī (Calcutta: Bibliotheca Indica, no. 254, Asiatic Society of Bengal, 1931) (henceforth Tārīkh-i Mubārak Shāhī) 76; Firishta, I, 103.

30 In 795/1392–93 and near the end of the Tughluq period Muḥammad b. Fīrūz Shāh's army commander, Ẓafar Khān, took over Gujarat and put an end to what little had been left of any resistance. He remained loyal to the Tughluqs to the bitter end, but after their demise claimed independence and established the Gujarat sultanate. See Sikandar b. Muḥammad known as Manjhū b. Akbar, Mir’āt-i Sikandarī (ed. Misra, S. C. and Rahman, M. L., Baroda, 1961), 6–20Google Scholar; Khwāja Niẓām al-dīn Aḥmad b. Muḥammad Hirawī, Ṭabaqāt-i Akbarī (Calcutta: Persian text, 3 vols, Biblioteca Indica no. 223, Asiatic Society of Bengal, 1927–35), III, 1935, 82–5; Firishta, I, 153; II, 178–80.

31 Burgess, James, The Muhammadan Architecture of Ahmadabad, Part I, A.D. 1412 to 1520 (London: ASI, New Imperial Series, XXIV, ASWI, VII, 1900)Google Scholar; The Muhammadan Architecture of Ahmadabad, Part II, with Muslim and Hindu Remains in the Vicinity (London: ASI, New Imperial Series, XXXIII, ASWI, VIII, 1905)Google Scholar; Burgess, James and Cousens, Henry, Architectural Antiquities of Northern Gujarat (London: ASI, New Imperial Series, XXXII, ASWI, IX, 1903)Google Scholar.

32 See for example; Rajan, K. V. Soundara, Ahmadabad (ASI, New Delhi), 1980Google Scholar; Michell, George and Shah, Snehal (eds), Mediaeval Ahmadabad (Bombay: Marg 39/3, 1988) particularly John Burton-Page's chapter on mosques and tombs, 30–119Google Scholar; Lambourn, Elizabeth, “A collection of merits: architectural influences in the Friday Mosque and the Kazaruni tomb complex at Cambay, Gujarat”, SAS 17, 2001, 117–49Google Scholar; Patel, Alka, Building Communities in Gujarat: Architecture and Society during the Twelfth through Fourteenth Centuries (Leiden and Boston: Brill, 2004)Google Scholar. For vernacular and domestic architecture see Pramar, V. S., A Social History of Indian Architecture (Delhi and Oxford: OUP, 2005)Google Scholar and for waterworks see Jain-Neubauer, Jutta, The Stepwells of Gujarat in Art-Historical Perspective (New Delhi, 1981)Google Scholar and Hegewald, Julia A. B., Water Architecture in South Asia: A Study of Types, Developments and Meaning (Leiden, Boston and Cologne: Brill, 2002)Google Scholar.

33 Shokoohy, M., Bhadreśvar: The Oldest Islamic Monuments in India (Leiden and New York: Brill, 1988), 11–33Google Scholar. The shrine is also noted in Flood, Objects of Translation, 47–8, but the information on this shrine and some other early Muslim edifices given in chapter 6, including the Ghurid remains in Hansi, the Shahi mosque in Khatu and the Chourasi Khamba mosque in Kaman is based on the present author's primary published reports.

34 Ibid., 42–9. Abu'l-Qāsim was a shipmaster and the chief (ṣadr) of the Muslim merchant community in Junagadh.

35 Marshall, John H., “The monuments of Muslim India”, in Haig, Wolseley (ed.), The Cambridge History of India (Cambridge, 1928), III, 630Google Scholar; Welch, Anthony and Crane, Howard, “The Tughluqs: master builders of the Delhi sultanate”, Muqarnas 1, 1983, 128Google Scholar; Michell, George and Zebrowski, Mark, The New Cambridge History of India, 1/7, Architecture and Art of the Deccan Sultanates (Cambridge: CUP, 1999), 63–4CrossRefGoogle Scholar;

36 Marshall, Cambridge History of India (1928), III, 68. Brown, Percy, Indian Architecture (Islamic Period) (Bombay, 1942, revised edition, 7th reprint, 1981), 60Google Scholar.

37 The problem with the dated mosques in Mangrol is that the inscribed panels were moved from one building to another in the nineteenth and twentieth centuries, sometimes more than once. The attribution of the inscriptions to a building cannot therefore be determined with certainty. Mosques with columns composed of two superimposed shafts are likely to be earlier than the time of Fīrūz Shāh.

38 Cousens, Somnāth, 64–5; Desai, Z. A., “Khalji and Tughluq inscriptions from Gujarat”, EIAPS, 1962, 24–6Google Scholar. The inscription is now fixed on the qibla wall of the Jāmiʿ, but Desai reports that it was once in the Bohra Masjid, and may not have originally belonged to the Jāmiʿ. He also implies that the Jāmiʿ may be earlier than the inscription. Cousens mentions that “the mosque was built by Shams Khān Vazir to Firuz Shāh in 1364” without giving his source. There was no such personage in Fīrūz Shāh's court but there were two personages called Shams al-dīn operating in Gujarat. One was Malik Shams al-dīn Abū Rajā’, Deputy Governor of Gujarat who could be the founder of the mosque. According to Shams-i Sirāj, “when Malik Shams al-dīn arrived at the territory of Gujarat he founded many things there” (ملک شمس الدین چون در اقطاع گجرات رفت در گجرات نیز چیزهای بسیار بنیاد نهاد). In 877 he was replaced by Shams al-dīn Dāmghānī, who rebelled against Fīrūz Shāh and a year later was killed by his own centurions (amīrān-i sada). For the affairs of Gujarat see Shams-i Sirāj ʿAfīf, Tārīkh-i Fīrūz Shāhī (Calcutta: Bibliotheca Indica, no. 119, Asiatic Society of Bengal, 1891), 454–5, 500–2 (henceforth Shams-i Sirāj); Tārīkh-i Mubārak Shāhī, 132.

39 Cousens, Somnāth, 65–6, mentions that “the mosque was built in 1401 by Jāfar Khān, at the time of Muḥammad Tughlak” without giving his source. This account seems to be confused as the two Tughluq sultans by this name were Muḥammad b. Tughluq (725–752/1325–51) and Nāṣir al-dīn Muḥammad (792–5/1390–93). Cousens might have meant the last Tughluq sultan Maḥmūd Shāh (795–816/1392–1414), but no inscription of this sultan has ever been attributed to the Rāvalī Masjid. Desai, Z. A., “Khalji and Tughluq inscriptions from Gujarat”, EIAPS, 1962, 30–32Google Scholar, reports another inscription of the time of Fīrūz Shāh with the date quoted above. He notes that the inscription was originally fixed on the wall of a tomb near the Rāvalī Masjid, but was said to have come from the mosque. While the inscription refers to the construction of a mosque, there is no firm evidence that it belonged to the Rāvalī and it is likely that the mosque is earlier than all the suggested dates.

40 Cousens, Somnāth, 65; Z. A. Desai, “Khalji and Tughluq inscriptions”, 23–4.

41 The mosque is first reported here. Z. A. Desai, “Khalji and Tughluq inscriptions”, 27–30, calls the mosque Junī Jail kī Masjid, because of its proximity to the town's prison, but the mosque is known as Chishtīwālā, and according to its inscription was built by the efforts of one Khwāja Muḥammad b. ʿAlī for Khwāja Farīd al-dīn Kalān, a disciple of Shaikh Naṣīr al-dīn Chishtī, who is said to have been later buried in the mosque.

42 The Chishtīwālā Masjid is closely similar both in plan and in scale to the Jāmiʿ of Veraval built during the reign of Muḥammad b. Tughluq and dated 1 Ramaḍān 732/27 May 1332. For the Jāmiʿ of Veraval see Cousens, Somnāth, 34; Husain, Mahdi, “Six inscriptions of Sulṭān Muḥammad bin Tughluq Shāh”, EIAPS, 1957–58, 38–9Google Scholar.

43 Burgess, James, The Muhammadan Architecture of Ahmadabad, Part I, A.D. 1412 to 1520 (London: ASI, New Imperial Series, XXIV, Western India, VII, 1900), 17–19, pls. 3, 11–18Google Scholar.

44 Ibid., 53, pl. 52.

45 Desai, Z. A., “Inscriptions of the Gujarat sultans”, EIAPS, 1963, 24–6, pl. 8Google Scholar.

46 Ibid., 30–32. The name suggests that it may have been a shrine, but the inscription refers to the construction of a mosque. Another inscription from a lost mosque is also known (ibid., 50). The original location of the mosque is unknown and the worn inscription is hardly decipherable, but its Persian text is an indication that it belonged to a sultanate mosque, as the pre-sultanate inscriptions of the maritime communities are all in Arabic, even when set up by Persian-speaking merchants, as we have seen in the case of the inscriptions of Abu'l Qāsim b. ʿAlī at Junagadh and Nākhudā Fīrūz b. Abī Ibrāhīm in Somnath.

47 Cousens, Somnāth, 28–9, pls. 10–11.

48 Cousens, Somnāth, 13, notes the gate briefly and gives a photograph, but does not mention the Idrīs or the Qalandarī mosque (noted below), both near the gate.

49 During the sultanate period these galleries were known as mulūk khāna, but the Emperor Jahāngīr notes that the Mughals called them shāh nishīn (royal chamber). See Shams-i Sirāj, 80; Sikandar b. Muḥammad Manjhū, b. Akbar, Mir'āt-i Sikandarī, ed. Misra, S. C. and Rahman, M. L. (Baroda, 1961), 38Google Scholar; Muḥammad Jahāngīr Gūrkānī, , Jahāngīr nāma or Tūzuk-i Jahāngīrī, ed. Hashim, Muhammad (Tehran, 1980), 242Google Scholar.

50 Shokoohy, Mehrdad and Shokoohy, Natalie H., “The Karao Jāmiʿ Mosque of Diu in the light of the history of the island”, SAS 16, 2000, 55–72Google Scholar.

51 The mosque is first reported here.

52 Desai, Z. A., “Inscriptions of the sultans of Gujarat from Saurashtra”, EIAPS, 1953, 61–2Google Scholar.

53 For the illustration of a tenth-century image of Śiva Naṭarāja now housed in the mosque see Davis, Richard H., Lives of Indian Images (Princeton: Princeton University Press, 1997), 92Google Scholar, fig. 14. Davis, who seems to be unfamiliar with the principles of architectural planning of mosques and temples, reiterates, regrettably, the view of Hindu extremists and notes (p. 289, note 27): “The structure housing the museum had also experienced shifts in identity. Built originally during Kumārapāla's time as a temple to the Sun god Sūrya, it served as a Jāmi Masjīd [sic.] in later mediaeval times, before being appropriated and transformed into a secular archaeological site museum”.

54 Tārīkh-i Mubārak Shāhī, 76. Barnī, 251, describes the destruction of the icon.