| Site home | 

< Previous  |  Main contents  |  Section contents | Next >

[[ Other sections:  || Archaeology|| Architecture||  Arts and Crafts||  Cultural history||  || Folklore||  || Indigenous||  || Language & Literature||  || Living traditions||  || State & Culture  ]] 

It’s eulogies are mentioned in inscriptions from Budh-Gaya and Nalanda, and in Tibetan translations of certain Sanskrit Buddhist works. In the Nalanda inscription of Vipulasrimitra it has been described as ‘a singular feast to the eyes of the world’. It occupies almost a square area of 276.6m from north to south and 276m from east to west. Like the Salban Vihara at Mainamati it’s entrance is projected from the north and it has a large temple in the middle of the courtyard surrounded by subsidiary buildings, but unlike the former it has large projecting portals in the center of each of the other three sides and has a narrow private doorway to the east of the gateway entrance. The outer walls of the monastery are 4.7m in thickness, and stand at present to a height from 3.6m to 4.5m. The number of the cells around the courtyard excluding those in the wing portals is 177-45 being on the northern side and 44 in each of the other three sides. The verandah infront of the cells appears to be equal in width to that of the Salban Vihara although in some place it was 2.75m wide. The original construction of the monastery was undertaken during the time of Dharmapala (c. 770-810 CE), but it continued to exist, through repairs, restorations and reconstructions as a monastic establishment down to the Muslim conquest. The crowning feature of this monument is it’s central shrine with the square solid block inside, the whole being seen from a long distance still today. Majestic as it was originally, but covered by debris and earth subsequently the name ‘Paharpur’ or hillock village was aptly given by local people carrying it to far and wide as one of the most impressive monuments of all time.

3.2.3 Temple

Like the stupa and the vihara we do not have any extant example of temple of the pre- Muslim period in Bengal with it’s upper part intact. Whatever idea of the temple architecture, therefore, we have are from a few surviving examples of West Bengal– belonging mostly from 11th century onwards, some available stone representations containing various deities in different museums, the manuscript illustrations noted above, and the excavated remains in the different archaeological sites of the country. The large number of temples which are now seen in various parts of Bangladesh and in West Bengal date from the 16th century onwards, and although are generally said


2 A comparable monastery identified with Vikramsila Mahavihara of Dharmapala is now being excavated at Antichak in Bihar.

to have been influenced by the Muslim architecture, are also known to contain some structural features reminiscent of older days. It is in the context of these examples that the temple architecture of ancient Bengal must have to be studied. A close examination of the various features of these temples reveal that the temples of pre-Muslim times followed the Northern and the Orissan styles of Indian architecture. The temples in Bengal were generally built on square, rectangular or cruciform plan with an ambulatory passage round the sanctum (garbhagrha), the most important part of the structure to contain images to be worshiped. The square temples often were the smallest of the three, and the cruciform the largest. In some cases pillared halls (mandpas) or small porches were constructed on the frontal side. The ornamentations of the temples were executed in carved brick and terracotta, and in carved and sculptured stones, only few examples of which are now seen in situ. For the convenience of study the temples of ancient Bengal may broadly be divided into three distinct types according to the form of the roof erected over the sanctum. They are: i. The Bhadra or Pida type in which the roof over the sanctum consists of a series of deminishing tiered stages crowned by the amalaka and finial; ii. The Rekha or Sikhara type characterized by a lofty curvilinear sikhara over the sanctum; and iii. The Bhadra type with a complex superstructure over the sanctum, the crowning member being either a stupa or a sikhara.


The Bhadra type is generally regarded as the earliest type of temples in Bengal. A model found at Chandraketugarh dated c.2nd century CE. consisting of a two- tiered pavilion sheltering a female divinity may be regarded as a proto-type of this style. A second example from the site is the 4.22m high remains of a massive Bhadra type of brick temple with entrant angles assigned to Gupta period. If this is the case this has been certainly an important discovery, and along with the ‘Govinda Swami temple at Baigram in Dinajpur District, dated 128 of the Gupta

era (477-78 CE)’ represent two of the

Sketch design of a Bhadra or Pida Deul

earliest temple structures ever known from the history of Bengal. The design is further represented by the Ashrafpur bronze votive stupa of the 7th century CE and several other stone representations containing deities such as the Uma-Mahesvara from Birol (Rajshahi) now in the Ashutosh Museum, the Surya from Baria (Rajshahi) in the Varendra Research Museum, and the Buddha from Madhyapada (Dacca) in the Dacca Museum. The developed type survives at present in a large number of simple brick temples of later dates in various parts of Bangladesh and west Bengal. The tiered roofs

in these temples are frequently seen in the curvilinear form representing the chauchala

bamboo-huts, one of the two most popular kinds of house-forms in the area.


Of the second or Rekha type, there are a number of known extant examples, both in stone and brick, but all of them now in West Bengal. They are generally dated from the 8th century CE. onwards. The best specimen of this type is the brick Siddhesvara temple at Bahulara, dated 11th century, in the Bankura District. It consists of a straight and perpendicular sanctum crowned by a curvilinear sikhara, all tastefully decorated with niches and scroll patterns. A Jaina Temple at Charra and a Brahmanical temple at Tuisama near Budhpur in Purulia District are two further examples of the period. Two more examples from Telkupi in the same district date c. 13th century, and are distinguished from the earlier in

Sketch of a Shikhara or Rekh Deul

external decorative designs. Decaying examples of the type from the 14th century through 15th

century are quite a few, and are represented by the extant Jaina examples at Deoli in Purulia District and at Harmashra and Deulbaria in Bankura District. The twin examples of Sallesvara and Sandesvara at Dihar in Bankura District, and examples at Barakar in Burdwan and at Banda and Para in Purulia District are of the same varying type but with a occasional ‘best expressions’ as the one at Barakar. In present Bangladesh an excellent miniature specimen of this type in stone of the Pala Period may be seen in the Brahma temple from Nimdighi (Rajshahi), now in the possession of the Department of Archaeology in the National Museum. Another miniature example in bronze has been found in Jhewari (Chittagong) and is now in the Indian Museum, Calcutta.

The third complex type also does not survive in Bangladesh; in fact, as far as known, not in any part of West Bengal either. But the basements and various terrace-walls which have been discovered in the archaeological sites of Mahasthan, Mainamati and Paharpur clearly prove that at one time there were stupendous structures over them, and that these must have been of this type. The cruciform (chaturmukha or sarvatobhadra) ground plans of the monastery temples at Mainamati and Paharpur and the re-entrant structural remains over them are so similar to the ground plans of several extant examples now in Burma, Cambodia and Indonesia3, that a parallel has frequently been drawn with the suggestion that most probably these structures in South-East Asia were largely influenced by those widely known great example of


3 As for example the Ananda Temple at Pagan (1080 CE) in Burma, the temple of Angkor Vat (between 1112-1151 CE) in Cambodia and the Kalasan temple (c. 778) in Java.

Bangladesh. The nava-ratna temple at Kantanagar in Dinajpur and the pancha-ratna temple at Puthia in Rajshahi, both dated 18th century CE, although are on square foundations, may be regarded because of the superstructures over them as the successor-temples of these types. It also needs to be mentioned that temples with superstructures crowned by stupas or sikharas are also seen in the illustrations of the Astasahasrika Prajnaparamita and in stone representations containing deities such as those of Buddha from Mahakali (Dacca) now in the Dacca Museum; and Jina Rsabhantha from Suruhar (Dinajpur) now in the Varendra Research Museum.

Of the several temple ruins discovered at Mahasthan, the two large ones unearthed from the mounds locally known as the Govinda Bhita temple is situated on the bend of the Karatoya outside the northern wall of the citadel, and is now cleared of the external debris. It’s picturesque setting and commanding height, according to archaeologists, mark it as one of the most important ancient temple remains in the country. The temple is surrounded by a massive enclosing wall, 1.8m in thickness, and consists of two distinct sets of buildings, conveniently called eastern and western. The most important feature of the temple was that it consisted of successive terraces formed by a large number of cells around a central rectangular platform at the top which appears to have formed the foundation of the high superstructure. On the western side are the remains of a long porch, 9m long and 1.5m projected, in two successive levels the earliest being dated 6th century and the latter 8th/9th century CE. The curious cellular construction of the temple was a remarkable development and must have served as an important method for providing a high and solid foundation for an imposing shrine at the top.

The temple of the Lakshindarer Medh is situated to the west of the Gokul village, about 2 km to the south-west of the citadel. It is also an imposing structure of cellular construction, cruciform in plan and surrounded by a central shrine of complex outline. In the center is an octagonal plinth which originally must have carried a stupa. The stupa, however, was replaced by a square shrine, about 8.1m a side, and a porch on the west during the Sena period (c. 1095-1228). In the center of the shrine a stone-slab of 50 cm by 45 cm was discovered and it bore twelve shallow holes with a larger hole in the centre containing a tiny gold-leaf about an inch square. The gold-leaf bore, in repousse relief, the figure of a recumbent bull which strongly indicates that the overlying shrine was a Xiva temple.

There are a number of other smaller temples within the citadel. All of them are either rectangular or square in plan and are sometimes surrounded by circumambulatory passages. A temple at the Bairagi Bhita has a porch in the middle of the northern side. It is likely that all these temples together with the similar ones in Mainamati and Paharpur to be described hereafter belonged to either type i or type ii.

The temples of Mainamati appear to have been made in two plan-types: one a simple square cella at the back with an oblong mandapa or a corridor like approach in front, and two a cruciform variety. The square form with a mandapa as is noticed in the

Charpatra Mura, a Hindu temple, apparently a North Indian variety is a new type differing basically from other Buddhist shrines of Mainamati. It is seen to have been evolved by assimilating many of the elements and features of the local Buddhist architecture of the Khadga and early Deva dynasty. On the evidence of the discovery of four copper-plate inscriptions (and hence charpatra-mura) within the temple dedicated to ‘Ladaha Madhava’ (Visnu) and other architectural details, the temple has been described as a Chandra monument built by two Chandra kings and a later Deva King. The temple was probably rebuilt in it’s present form. If this is the case, the temple must have represented a later type than the cruciform one, and perhaps may be regarded as one of the earliest known Hindu temples in present Bangladesh. The other example at Itakhola Mura, dated between late 7th and early 8th century CE is a pre­ cruciform type, and although represents a somewhat North Indian type, is unique by itself in the sense that the leading nave-like corridor in the middle flanked by three more corridors on the right and left shows the characteristic of a church plan unseen in any other example in Mainamati.

The examples of the cruciform plan shrine as is seen in the midst of the courtyards of

Salban Vihara, Bhoja Vihara and Ananda Vihara are of the same type and appear to have been a common variety built in the 8th century.

The cruciform temple of the Salban Vihara was built along with the monastery by King Bhava Deva and as is apparent from the remains must have carried a gigantic superstructure over it. It measures 51.8m from arm to arm and has an entrance from the north through a stepped terrace. The entrance leads to an ambulatory passage, 2.1m wide, which was carried round the entire building communicating with the chapels in the projecting arms of the cross. The basement walls of both the eastern and western projections are now in a good state of preservation and are seen to have been ornamented with a horizontal row of sculptured terracotta plaques depicting mythology, folk-lore and everyday life of the country. The Terracotta plaques are crowned by two rows of horizontal carved bricks, the lower in chequered and the upper in petal designs both forming a sort of cornice.

The plan of the temple was subsequently altered in two phases to form an oblong structure with a square entrance hall in front and a pillared central pavilion surrounded by other rooms at the back. It is likely that at these periods also the temple carried superstructures over it to make it harmonious with the plan below.

The central temple at the Salban Vihara is flanked by other subsidiary temples remains of which two on the western side and a third on the north-west corner outside the monastery have been excavated. Of the two within the monastery the smallest was square in plan with re-entrant angles at the corners, and the other was rectangular containing an entrance hall infront and a small cella at the back both being surrounded by an ambulatory passage. The last one, much larger than these two, is square in plan, and was provided with a rectangular columned terrace in front and an ambulatory walk round the sanctum.

The central temple of the Bhoja Vihara is of similar cruciform type to that of Salban Vihara. At one time it’s basement was profusely decorated with carved plaques and ornamental bricks. The centre of this structure was occupied by a 1.8m square blind box-chamberthe same arrangement to be found at Paharpur, but on a smaller scale here. The central temple of the Ananda Vihara, only partly excavated is of similar type but larger in scale in conformity with the size of the vast monastery and perhaps needs no elaboration.

By far the largest and the most important of all the known temples in Bangladesh is that which has been discovered in the center of the courtyard of the Somapura Vihara at Paharpur. Built by the Pala ruler Dharmapala towards the end of the 8th century CE, it like the one of the Salvan Vihara in Mainamati took on the ground the form of a cross measuring 108.3m by 95.45m and consisted above of at least three raised terraces. At present the height of the ruins is about 21m above the ground, but originally, it is estimated, must have been above 30m. The approach to the temple was by a stairway on the northern side leading to an ambulatory passage carried round the entire structure and communicating with the four chapels in the projecting arms of the cross. The central square platform at the top which was so long regarded as the floor of a cell enshrining an image of exceptional sanctity is now considered by the archaeologists as providing only a structural member to add to the solidity of the foundation of the lofty walls. There is no evidence that the hollow square pile has any access to it’s inner part from the ante-chambers. Like the Mainamati example the basement walls of this temple are also decorated with horizontal rows of terracotta plaques depicting the same subject matters, and also by the insertion of stone bas- reliefs at most angles of the projections and at intervals in specially built recesses in the middle of the walls. Inspite of the fact that the upper part of the temple has now vanished, a picture of it in it’s heyday is not unimaginable when we think of it’s vast ground plan, the various terraces and the square pile over it. On the analogy of Pagan and other South-East Asian temples, it has been suggested that the square masonry pile in the center at one time supported a tall curvilinear sikhara, an impressive spire seen from a long distance. The splendour of Paharpur temple is now gone, but it’s frame remains, and in it is to be seen, until we know further, the culminating point of the pre- Muslim architecture of Bangladesh.

Other mentionable temples at Paharpur are the remains of a miniature model of the central temple on the eastern part of the monastery courtyard, and the temple of Tara in the Satyapir Bhita. The latter consists of two parts, the sanctum in the north and a pillared hall on the south around which there is the usual circumbulatory corridor.

[[ Other sections:  || Archaeology|| Architecture||  Arts and Crafts||  Cultural history||  || Folklore||  || Indigenous||  || Language & Literature||  || Living traditions||  || State & Culture  ]]