IN the fifteenth century Muhammadan building activity in India increased in the centres already established, especially at Gaur and Ahmadâbâd, and also extended to others, the chief of which were Dhar and Mandû in the province of Mâlwâ, and Jaunpur, about 40 miles north-west of Benares. Both in Mâlwâ and at Jaunpur there were marked developments in Muhammadan building craft, though in different directions. At Mandû the Indian builders began to extend the use of the arch structurally, just as they had done at Gaur in the previous century,1 so that their buildings assumed a more decidedly Saracenic or Persian appearance internally as well as externally. At Jaunpur they worked in an opposite direction—i.e. they took away the typical Arabic or Persian character of the arched screens in front of the lîwân, by combining the Hindu beam and bracket with the “Saracenic” arch. This happy combination of the two constructive principles continued to be the most common characteristic of Muhammadan building in India. It was essentially an Indian invention or adaptation—not a foreign one; that is, the Indian craftsmen were not being instructed by foreign builders, but were adapting the structural use of the arch, first forced upon them by their Muhammadan rulers, in the way which pleased themselves.

But in the province of Mâlwâ, as in Gujerat, the pre-Muhammadan buildings had been for many centuries largely built of stone, and consequently the arch had not been used structurally, even on a small scale, as it had been at Gaur, before the Muhammadan ascendancy. It was, however, inevitable that intelligent craftsmen, as Indians undoubtedly were, once they had accepted the arch as a structural necessity in front of the sanctuary of Muhammadan mosques, and finding it convenient for bridging over wide spans between columns, piers, or walls, should sooner or later begin to experiment with it in the interior of their buildings. This is what happened at Mandû and some other places in Mâlwâ. Fergusson, classifying Indian buildings as a student of architectural style rather than as a craftsman, assumes that at Mandû, as at Delhi and Ajmîr, Persian, Arabian, or Syrian builders introduced by the Muhammadan rulers were beginning to teach Indians the “true elements of architectural design,” according to Saracenic ideas.

Prima facie, the mosques and palaces of Mandû seem to afford strong evidence that this was the case; but the craftsmanship tells a very different tale. There is tile-work which might be Persian or have been imported from Gaur, but no evidence of the Arabian or Persian builder. The stone arches are built by Indian masons experimenting for themselves in this form of construction. The voussoirs of the arches are not divided with mathematical regularity as they would be by a Persian or Arabian mason skilled in arch construction, but are cut irregularly; the keystone, which to a skilled arch-builder is the principal one, being the smallest. Moreover, the form of them is not strictly Arabian or Persian, for the crowns are tipped up to give that suggestion of the sacred pîpal leaf which is typical of the arch in Indian-Buddhist and in Hindu shrines. The mihrâbs are only adaptations of local Hindu shrines (PI. XVIII). The domes are not crowned by the correct Saracenic finial, but by Buddhist-Hindu emblems—a sure sign that the masons were Indians. There are buildings at Mandû which show the transition from the old to the new Indian style, some of the columns in the interior being joined by beam and bracket and others by arches (PI. XIX.). This is an indication that Indian builders, being no longer bound by Hindu ritualistic traditions, were voluntarily adapting their craft to the new structural conditions, for foreign builders imported to instruct Indians would not have used Hindu methods and symbolism.

The difference in point of style between Mâlwâ architecture and the contemporary Muhammadan styles in Gujerat and Jaunpur is that at Mandû and other places in Mâlwâ the builders began to obtain the heights they wanted inside the mosques by joining the piers and columns with pointed arches, instead of by placing one column on the top of another, or by building two stories, as Hindus would have done. We may agree with Fergusson in appreciating the effect of simple grandeur and expression of power which they obtained in this way, without denying to Indian builders the credit which is their due.


At Jaunpur the principal buildings of the fifteenth century are the Atâla Masjid (PI. XX.), completed in 1408 during the Sultanate of Shah Ibrâhîm (1401-39), and the Jâmi’ Masjid, commenced in the time of Husain Shah (1452-78)—both noble structures with a strongly marked Hindu character though the exterior arches are without the pîpal-leaf keystone, and though pointed vaulting with ribs is introduced into some compartments of the interior. A very striking and original effect is produced by the treatment of the screen in front of the lîwân, which in these buildings is reduced to a single lofty arch, flanked by turret-like sloping buttresses2 which serve for minarets, and filled in with a subsidiary slightly recessed screen in which the “Saracenic” arch and the Hindu lintel are ingeniously combined. An archæologist or purist in style may think the combination strange and hybrid, but as architecture it is finely conceived.

The whole style of the building seems to be a reflection of the massive grandeur of Hindu tempks like the Sâs-Bahû or Padmanâbha temple at Gwalior (Pl. XXI), which resembles these mosques in being built in several stories and in being raised on a platform of masonry. All the domes at Jaunpur are surmounted by Hindu emblems, as is the case with nearly all Muhammadan mosques in India.


About the same time as the buildings already described were being constructed at Mandû and Jaunpur, Ahmadâbâd—now the capital of an independent Musulmân kingdom, and so called from the name of the second Sultan of the dynasty—was being adorned with a series of splendid buildings which, like other Muhammadan edifices of this period, bear striking testimony to Indian constructive genius. Ahmad Shah, being a Rajput himself, had no foreign prepossessions in architectural style, so that when he set about building a Jâmi’ Masjid3 soon after the commencement of his reign in 1411, his Indian builders were given an entirely free hand in the design of it. It so happened that about the same time, as Fergusson tells us, another independent Rajput chief, Kumbha Rânâ, of Mewar, a Hindu of the Jaina sect, was building a great temple at Rânpur, about sixty miles from Ahmadâbâd. A comparison between these two buildings is particularly useful as an illustration of my contention that Muhammadan and Hindu architects in India were, with rare exceptions, craftsmen of the same race, imbued with the same craft traditions and possessing an equal capacity for dealing with any constructive or purely artistic work which their rulers might be pleased to place in their hands.

The few instances in which it can be shown with certainty that Muhammadan rulers in India sent to foreign countries for architects or craftsmen by no means prove that India was unable to supply men of equal or superior capacity, though such cases might logically be taken to prove the ruler’s prejudice or ignorance. The only possible way of deciding this question judicially is to examine the buildings themselves for evidence of foreign design and craftsmanship, taking care to discriminate between the two, for a borrowed idea does not necessarily mean foreign brains or handiwork.

The term “Saracenic,” as applied to Muhammadan architecture in Gujerat, is even more misleading to the student than Fergusson’s classification is generally. There is not the least indication in any of these buildings of foreign design or handicraft. No other form of Muhammadan architecture in India, says Fergusson, is so essentially Indian: though generally he represents the Saracenic builder as the inspirer of the Hindu, he is constrained to admire this Indian style as being the most elegant of them all. Comparing the Hindu temple at Rânpur with the contemporary Jâmi’ Masjid at Ahmadâbâd, he feels instinctively that there is more poetry in the former, but, fearing that his artistic instinct may offend his academic conscience, he adds, “there is a sobriety about the plan of the mosque which after all may be better taste.”

Comparing the façade of the Jâmi’ Masjid at Ahmadâbâd (Pl. XXII) with the screens at Delhi and Ajmîr, it is easy to see how the fifteenth-century builders in Gujerat were trying to modify the thirteenth-century models which had been forced upon Indian master-craftsmen. They clearly felt with the Jaunpur builders, that, however beautiful the Ajmîr and Delhi screens might be in themselves, they were ill-fitted in structure for their purpose and artistically incongruous with the Hindu interior of the mosque. So instead of altering the structure of the interior in order to adapt the latter to the façade, as the Mandû builders tried to do, they Hinduised the design of the façade to make it fit the interior.

Disliking the regularity of the Ajmîr and Delhi screens, they broke up the horizontal lines by dividing the façade into five compartments instead of three; and by increasing the height of each successive compartment from the ends towards the centre of the façade, they gave the whole design the pyramidal lines which are characteristic of Hindu temple-structures. The lofty “Saracenic” arches of the screen were reduced in number to three instead of seven—one on each side of the great central arch—the ten smaller Hindu arches of the adjoining compartments being formed by bringing five of the interior rows of columns on the north and south of the lîwân out to the line of the façade and linking them together below the capitals with brackets in Hindu fashion, in the same way as most of the small arches were formed in the buildings at Gaur. The keystones of the three main arches have, as usual, the symbolism of the pîpal leaf worked into them.

The beautiful minarets which are so characteristic of this and other mosques in Gujerat have none of the Saracenic feeling of the Qutb Minâr at Delhi, but are entirely Hindu in style, being only adaptations of the splendid Rajput Towers of Victory at Chitor (Pl. XXIII). Unfortunately the Jâmi’ Masjid lost the upper part of its minarets by an earthquake in 1819, and the unity of the whole design of the façade was thus sadly broken. But even when this is taken into consideration one feels that the difficulty of harmonising the Saracenic façade with the Hindu interior was not overcome quite so successfully in the Jâmi’ Masjid as in some of the later buildings in Gujerat, particularly the Rânî Rupâvati Masjid at Mirzapur (Plate LIV), which has also lost the upper part of its minarets. Fergusson’s observation that as the style progressed it became more and more Indian, rather than Saracenic, may be noted in this connection.

The beauty of the Jâmi’ Masjid and of most of the Gujerat buildings of this century lies, however, mostly in their interior structure and decoration, into which no trace of the Saracenic element enters. Even the most sacred symbol of Islâm, the mihrâb, is so completely transformed that, except for a small pointed arch, as much Hindu as Saracenic, it is only a replica of the door of a Hindu shrine.

Plate XXV, the interior of this building, and Plate L, showing the interior of another Gujerat building, the Jâmi’ Masjid at Champanîr, will help the reader to realise the decorative richness and noble structural design of these early Gujerat mosques, though the Champanîr mosque is really about half a century later than the Ahmadâbâd building.

The plan of the lîwân at Ahmadâbâd (fig. 11) will show the disposition of the columns and arrangement of the domes; the sections (Pl. XXIV.) will explain the structure of the interior. There are fifteen large domes, each supported on eight columns according to the usual Hindu design, and built up in horizontal courses by gradually changing the octagonal base into a circle. The large domes are linked together by a flat roof and by a number of smaller domes of similar construction supported on four columns each. The longitudinal section of the lîwân follows the pyramidal lines of the exterior, the great central dome in front of the main entrance, together with its four smaller connecting domes, being raised up above the adjacent ones so as to admit a diffused light through clerestory windows. A similar arrangement obtains in the next adjacent aisles on the longitudinal section. Fergusson observes of this arrangement that “the necessary amount of light is introduced, as in a Byzantine dome, but in a more artistic manner. The sun’s rays can never fall on the floor, or even so low as the head of anyone standing there. The light is reflected from the external roof into the dome, and perfect ventilation is obtained, with the most pleasant effect of illumination without glare.” He might have added that the arrangement was not a Saracenic invention, but a long-standing tradition in Indian temple-building of that part of India; being only a slight modification of the similar idea which is carried out in the lighting of the splendid chapter-house at Ajantâ (Cave XIX.). None of the structural Buddhist monasteries of the same period are extant, otherwise we should doubtless have discovered in them the exact prototypes both of the Rânpur temple and of the Jimi’ Masjid at Ahmadâbâd.

As the temple built by Kumbha Rânâ (Pl. XXVI) lies in a sequestered valley in Jodhpur far away from the beaten track, it has not attracted so much attention as the famous shrines of Mount Âbû and has not yet been properly photographed, so it is difficult to add to what Fergusson has given for the purpose of showing that Kumbha Rânâ’s temple and Ahmad Shah’s mosque belong to exactly the same school of architectural design. But one interesting point may be noticed, which might be puzzling to Fergusson’s readers the fact that several of the domes of the Hindu temple are on the exterior “Muhammadan”—i.e. they are not sculptured in the Hindu style, but are brought to an even surface by cement and fine plaster in the same way as the domes of Muhammadan mosques. It is possible that in this particular instance the domes may be modern restoration, but it is a fact that soon after the Muhammadan conquests began, the Hindu temple-builders in Northern India began to treat the exterior of their domes in the same way as their craft brethren, the Muhammadan builders, were doing. It would be quite wrong to take this as a proof that the Muhammadans were teaching a superior art to the Hindus; it was simply that the latter sheltered themselves from the fury of their oppressors by observing the same law of protective imitation by which nature provides for the protection of the weak against the strong. The Brahmans were trying to protect their temples and to make them less offensive to Muhammadan susceptibility by making less conspicuous the anthropomorphic symbolism which Islâm denounced as “idolatry.” At the same time the teaching of Islâm was not without its influence upon Hinduism, inasmuch as both Jaina and Saiva teachers began to discountenance the use of images in religious ritual, as the Vedic rishis before the days of Buddhism had done. Idolatry, in the Puritan acceptance of the word, had never been and is not now a part of Brahmanical religious teaching.

The result of this was that in Northern India Hindu and Muhammadan buildings could no longer be distinguished by their domes, for they were often exactly similar. This, however, applies only to the pavilions and to the porch, or mandapam, in front of Hindu shrines, for neither the curvilinear spire of the northern styles nor the pyramidal structure which surmounted the shrine containing the image or sacred symbol in Dravidian temples was ever reproduced in Indian mosques. Of course the entire absence of figure sculpture, and generally of animals also, from Muhammadan buildings gave them a distinctive character, quite apart from the more frequent use of arches and differences in planning. What they lost in human interest and in plastic beauty they gained in charm of colour, in fine combinations of geometric and floral patterns, and in rich material. To many Europeans with “classical” predilections they will be more pleasing and correct in taste, owing to the greater restraint in plastic treatment which the law of Islâm imposed upon Indian craftsmen. On the other hand, those who can enter into the spirit of the great Gothic masters will feel not less admiration for the imaginativeness and wider artistic range which are shown in Hindu temple decoration of the same period.

Throughout the fifteenth century we find the Indian Muhammadan builders pursuing their own aims on these lines, often using foreign models in decorative design, as good craftsmen in all countries use them, not imitatively, but to increase their stock of artistic material. As regards structural design and craftsmanship, it would be difficult to name a single Indian Muhammadan building in this century which could be called foreign to India in the same sense as St. Mark’s at Venice was foreign to Italy, or as both Gothic and Renaissance architecture were originally foreign to England.

In several fine mosques at Gujerat and Jaunpur continued experiments were made in the design of the façade, though no important variation was made in the interior: the mosque of Muhâfiz Khan at Ahmadâbâd is one of the most successful in this respect, and one of the few which has its minarets still intact. The Jâmi’ Masjid at Dholkâ is another good example which Dr. Burgess supposes to be not later than 1485. The Alif Khan Masjid, otherwise known as the Brick Masjid, is dated by the same authority at about 1450: it is especially interesting in the present day‒when one of those many foolish or cynical reasons urged for neglecting the Indian building-craft is that it is necessarily extravagant—as showing what beautiful work Indian builders have done in brick and plaster as well as in more precious materials. It is necessary to observe in this connection that comparatively few Indian buildings usually classed as stone are constructed entirely of solid masonry. The main walls are generally of brick faced with stone, sometimes marble. The framing-in of the doorways of Alif Khan’s mosque (Pl. XXVII) is an adaptation of the design of the doorways of Hindu shrines.

The mosque and tombs at Sarkhej, near Ahmadâbâd, which also belong to the middle of the fifteenth century, are chiefly remarkable for the development they show in the use of pierced stone trellises which had been employed in Hindu temples for many centuries previously. This was an application of indigenous craft which afterwards became a fine art as exquisite as Persian tile-work, and constituting one of the chief glories of Indian mosques of the sixteenth and seventeenth centuries.

Among the important Muhammadan tombs built in this century, that of Sayyid Usmân, near Ahmadâbâd, is interesting from a structural point of view from a new departure which was made in the supports of the dome; the base of the latter, instead of being octagonal, was transformed into a dodecagon, and greater massiveness was given to the supports by joining two or four pillars into single piers—a practice which became common in later Muhammadan buildings and gave them a distinctive character. It was built, according to Fergusson, in 1460.

Another fine tomb of about the same date, that of Sayyid Mubârak, Minister of Mahmûd Begarah, who reigned at Gujerat from 1459 to 1511, is almost unique among the buildings of the province, because the builders, desiring to plan it on a grander scale than usual by increasing the distance between the supports of the roof, took to using arches throughout the building, including the double corridors which surrounded the enclosure of the tomb, as well as in the exterior screens which form the four façades and in the entrance porch. Here, also, for the first time apparently, clerestory windows with pointed arches were introduced into the octagonal base of the dome, giving the structure a distinctly Byzantine appearance. Here, certainly, the casual observer might say, Saracenic builders have been at work. But a careful study will show the Indian masonic tradition carried right through the whole building. The arches are put in by Indian craftsmen,4 for they have the symbolism of the pîpal leaf in the keystones. The piers are in the form of four square pillars grouped together, a design which a Hindu builder would adopt when a wider spacing than usual necessitated an increase in the traditional size of the roof supports. A Saracenic master-builder, accustomed to wide spaces between the piers, would not think of a large pier as four small pillars combined. The domes are all of Indian construction and with Indian symbolism. There is no trace anywhere of foreign suggestion or supervision. All that the Saracenic or Byzantine appearance of the building proves is that, given similar conditions and similar constructive problems, skilled craftsmen in all parts of the world arrive at similar results, though they may choose different ways of working.

At Gaur and at Mandû the buildings of the fifteenth century show little variation on those of the preceding century. The Dakhil Gate and other entrances to the Fort at Gaur and the Eklâkhi Masjid or tomb at Panduâ, ascribed to about the middle of the century, are examples of the beautiful brickwork with moulded and carved decoration which was one of the master-crafts of Bengal until quite recent times.

In Mâlwâ there was great building activity throughout the century, a number of palaces being constructed by the Sultans of that province at Mandû, and a very fine mosque, the Jâmi’ Masjid, which was finished by Mahmûd Shah in 1454. The style of these buildings has already been described.

  • 1. As there was easy communication by sea between Gaur and the west coast of India, it is extremely probable that craftsmen from Gaur assisted in the building of Mandû and other Muhammadan cities in the neighbouring provinces.
  • 2. The slope inwards of these buttresses is perhaps the Indian craftsman’s reminiscence of early Buddhist methods of construction when the walls of buildings were sloped inwards to counteract the thrust of vaulted roofs. The sloping architraves sometimes found in the doorways of modern buildings in Sikhim and Tibet are undoubtedly derived from this ancient practice. Some early Pathân tombs in India show the same slope, e.g. Tughlaq Khan’s tomb at Delhi.
  • 3. A Jâmi’ Masjid is the mosque in which the principal or Friday services are celebrated: hence it might be called a “cathedral mosque” to distinguish it from others.
  • 4. It should be noted that Mahmûdâbâd, the place of Sayyid Mubârak’s tomb, is close to the old Hindu city of Dabhoi, some of the remains of which are shown in Plates II and III. Doubtless the Muhammadans, as they were wont to do, had drawn many Hindu craftsmen into their service from there.