Although this work is principally intended to illustrate the splendid series of caves in Western India, there are four or five groups in the Bengal and Madras presidencies a knowledge of which, if not indispensable, is at least extremely useful to enable us to understand the history of the cave architecture on the Bombay side of India. It is true that with the exception of the Mahâvallipur group they cannot pretend to rival the western caves either in splendour or extent, but the Katak caves present features of great beauty and are interesting from their originality. The greatest historical interest, however, centres in tho Behar series, which, though small, are important for our purposes, having all been excavated during the existence of the Great Mauryan dynasty, and being, therefore, tho earliest caves, so far as we at present know, excavated for religious purposes in any part of India.
The Barabar caves are situated in an isolated range of granite hills on the left bank of the Phalgu river about 16 miles due north from the town of Gâya. They are seven in number, and though differing in plan, are all similar in character and evidently belong to same age. Their dimensions are inconsiderable. The largest, called the Nagarjuni cave, is a plain hall with circular ends measuring 46 feet by 19 feet 5 inches, and though two others, the Sudama and Lomas Rishi, are nearly as large, they are divided into two apartments, and consequently have not the same free area.1
Fortunately there is no difficulty whatever with regard to the date of these caves; six out of the seven have inscriptions upon them, all in the oldest form of the Pali alphabet, identical with that found on Aśoka's lâts. More than this, the inscription on the Sudama cave states that it was excavated in the 12th year of that monarch or B.C. 252, and is therefore the earliest here. The latest is the Gopi or Milkmaid cave, in the Nagarjuni hill, which is dated in the reign of Daśaratha, the grandson of Aśoka, in or about B.C. 214. The whole group is therefore comprehended within about 40 years, and was commenced apparently within 80 years after Alexander's visit to India.2
The only cave in this group that has no ancient inscription upon it is the Lomas Rishi, but it is not difficult to see why this was the case. It is the only one which has any architectural magnificence externally, and was consequently selected by two kings, Sârdula Varmâ and Ananta Varmâ, sons and grandsons apparently of Yajῆa ŚrÎ of the Andra dynasty in the third or fourth century of our era, to adorn it with their inscriptions and to announce its conversion to the purposes of the Brahmanical faith.3 Before doing this they no doubt carefully obliterated the more ancient inscription, which at that time was in all probability perfectly legible and easily understood. Whether this is, or is not the true explanation of the absence of an inscription in the lat characters in this cave, is of very little importance. It is so absolutely identical both in dimensions and disposition with the Sudama cave, which we know was excavated in the 12th year of Aśoka, that there can be no doubt as to its age. Its architecture alone, if it may be so called, would be sufficient to settle this point. As may be seen from the annexed woodcut it is as essentially wooden as any other cave facade in India. Whether it is more so man the cave at Bhâjâ quoted above (woodcut No. 1), is difficult to determine on its merits alone. If we had any Chaitya caves in Behar which admitted of direct comparison it might be possible to do so, but when these eastern caves were excavated, the bold expedient had not occurred to any one of sinking a cave at right angles to the face of the rock, deep into its bowels, and leaving one end entirely open for the admission of light. All the Behar caves have their axis parallel to the face of the rock, and their entrances are placed consequently on one side, so as to act as windows to light their interiors as well as for entrances. Another peculiarity of the eastern caves is that no real woodwork was used in their decoration, while all the early Chaitya caves in the west, were adorned with wooden ribs internally, whose remains are to be seen at this day, and their facades were, as at Bhâjâ, entirely constructed in teak Wood. It may be that the roofs of the buildings copied in the caves at Behar were framed in bambu, without wooden ribs, like the huts of the present day, and consequently they could neither be easily repeated nor imitated in the rock. But be this as it may, these differences are such that no direct comparison between the styles adopted in the two sides of India, could be expected to yield any very satisfactory results. It is consequently fortunate that in Aśoka's time, as we know from the example at Bharhut, it was the fashion to inscribe everything. At Bharhut there is hardly a single person, nor a Jataka, or historical scene, which has not a name or a description attached to it, and this seems also to have been the case with these caves. Before the time when the gateways at Sanchi were erected, in the first century of our era, this good custom seems to have died out. All the rails there are inscribed with the names of their donors, but they are earlier than the gateways. They too, however, have also the names of their donors engraved on them, but unfortunately nothing to help us to discriminate what the subjects are which are represented in the sculptures.
One characteristic which is constant both in the early caves in the eastern and western sides of India is that all the doorways have jambs sloping inwards. This could only have arisen from one of two circumstances: either it was, as at Mycenae and in all the early Grecian buildings in pre-Hellenic times, for the sake of shortening the bearing on the lintel. The Pelasgi had no knowledge of the principle of the radiating arch, and used only small stones in their architecture generally. It consequently, though awkward, was a justifiable expedient. In India it arose, as already pointed out, from a totally different cause. It was because the earliest cave diggers were copying wooden buildings, in which the main posts were placed sloping inwards, in order to counteract the outward thrust of their semicircular roofs. Though tolerable, however, while following the main lines of the building, the sloping jambs of the doorways were early felt to be inappropriate to stone constructions, and the practice in India died out entirely before the Christian era.4
Although so differently arranged that it is difficult to institute any direct comparison between them and the western Chaitya or Church caves, it seems almost certain that none of the Barabar caves were meant as residences, but were intended for sacred or ceremonial purposes. The one most like a Vihara, or residence, is the Nagarjuni cave, called “the Milkmaid's cave,” but even there a great hall 46 feet long, with rounded ends, and only one small door in the centre of one side, seems too large for the residence of one hermit, and it has none of those divisions into cells which are universally found in all western Viharas.5 At the same time it must be confessed that our knowledge of Buddhist ceremonial in the age of Aśoka does not enable us to say what kind of service would be appropriate to such a hall. It may, however, have been a Dharmaśâlâ or hall of assembly for the congregation; a form of building which was probably usual with the Buddhists in all ages of their supremacy.
The case is somewhat different with the Karna Chopar cave, a rectangular hall measuring 33 feet by 14, which was excavated in the 19th year of Aśoka. But here a vêdi or stone altar at one end clearly indicates a sacred purpose. On the other hand, there can be doubt but that the Sudama and Lomas Rishi caves, which are so nearly identical in form, were real Chaityas. Instead, however, of the circular dagobas, which in all instances occupy the centre of the apsidal inner “termination of the western caves, its place is here taken by a circular chamber evidently meaning the same thing. It is difficult for us now to decide at the present day whether it was inexperience which prevented the early cave diggers from, seeing their way to leave a free standing dagoba in their halls, or whether it was that in structural buildings of that age a wooden or metal dagoba or relic shrine stood in a circular chapel, and they copied that.7. Reports, vol. i. p. 49.