If there had been no other examples of Cave Temples in India than those described in the preceding pages, the subject of its rock-cut architecture, though interesting to local antiquaries and those specially connected with Indian matters, would hardly have been deemed of sufficient importance to attract attention in Europe. The caves in Behar are too small and insignificant to claim a special notice, except from their bearing on the general history of the subject. Those in Orissa, though larger and more elaborately finished, are too much isolated in their character to be of much value, except when studied in connexion with more extensive groups; while those in the Madras presidency are interesting more from their bearing on the past history of Buddhist architecture in the north, and on the future of the Dravidian style in the south, than from any peculiar merit of their own. When, however, these eastern caves are taken in connexion with the whole subject, as we now know it, they become invaluable, as throwing light on the general history of cave architecture in India, and receive a reflected light from the western caves, which increases their importance to an extent they could hardly claim for themselves. When we turn to the Western caves the case is widely different. We there find at least one thousand excavations of various sorts and dimensions. Some of great size and of the most elaborate architecture, and all having a distinct meaning and bearing on the general history of architecture. When their story is carefully examined it appears that they are spread pretty evenly over more than 1,000 years of the darkest, though most interesting, period of Indian history, and throw a light upon it as great or greater than can be derived from any other source. In addition to these claims to attention the Western caves afford the most vivid illustration of the rise and progress of all the three great religions that prevailed in India in the early centuries of our era and before it. They show clearly how the Buddhist religion rose and spread, and how its form afterwards became corrupt and idolatrous. They explain how it consequently came to be superseded by the nearly cognate form of Jainism and the antagonistic development of the revived religion of the Brahmans. All this too is done in a manner more vivid and more authentic than can be obtained from any other mode of illustration now available.

With all these claims to attention it is hardly to be wondered at that the western caves have attracted the attention of the learned both in India and in Europe from a very early period of their connexion with the East, and that a detailed statistical account of them may still be considered as a desideratum, which it is hoped this work may to some extent at least supply.

It is not easy at first sight to account for the extremely rapid extension of Cave architecture in the West of India as compared with that in the East. Behar was the cradle of the religion that first adopted this monumental form, not any part of Western India, while it will probably be admitted the Buddhists were the first to introduce this form of architecture on both sides of the country. At the same time there seems no reason for supposing that Buddhism in any form existed in the West before missionaries were sent there by Aśoka, after the convocation held by him in the seventeenth year of his reign, as detailed above (ante, p. 17). Before this time there is certainly no evidence to show that the inhabitants of the Western Ghâts were dwellers in caves or used the rock for any monumental or religious purpose, but immediately afterwards they seem to have commenced excavating it and continued to do so uninterruptedly for a long series of years.

It has been suggested that as the Egyptian rock-cut temples are principally in Upper Egypt above the Cataract and in Nubia, that their comparative proximity to India may have been the cause of this form being adopted there. The distance of date, however, between the latest Egyptian and earliest Indian examples quite precludes this idea. Besides the fact that no similarity of any detail can be traced between them, and there seems no other country which could have influenced India in this respect. On the whole the explanation of the phenomenon is probably the prosaic fact that the trap rocks which overlie the country and form the hill sides everywhere in the West are exceptionally well suited for the purpose. They lie everywhere horizontally. Are singularly uniform in their conformation, and have alternating strata of harder and softer rocks which admit of caves being interpolated between them with singular facility, and they are everywhere impervious to moisture.

With such a material it is little wonder that once it was suggested, the inhabitants of the Western Ghâts early seized upon the idea of erecting permanent quasi eternal temples for the practice of the rites of their new religion, in substitution for the perishable wooden structures they had hitherto employed, and once the fashion was adopted we ought not to be surprised it became so generally prevalent nor that it continued in use so long.

At the same time it may be observed that under the circumstances the amount of labour expended in excavating a rock-cut temple in so suitable a material is probably less than would be required to erect a similar building in quarried stone. If we take, for instance, even such an elaborate temple as the Kailasa or Elurâ, it will be found that the cubic contents of the temples left standing is about equal to the amount of material quarried out of the pit in which it stands. It is at the same time evident that it would be much less expensive to chip and throw out to spoil this amount of material, than to quarry it at a distance and carry it to the temple, and then hew it and raise it to the place where it was wanted. The amount of carving and ornament being, of course, the same in both cases. It is not so easy to make a comparison in the case of a Chaitya cave or a vihara, but on the whole it is probable that excavating them in the rock would generally prove cheaper than building them on the plain. If this is so, it is evident that the quasi eternity of the one offered such advantages in such a climate over any ephemeral structure they could erect elsewhere, that we ought not to be surprised at its general adoption. The proof that they exercised a wise discretion in doing this lies in the fact that, though we have in the west of India nearly a thousand rock-cut temples belonging to the Buddhist, Brahmanical and Jaina religions, we have only one or two structural examples erected in the same region at the very end of the period of time to which these caves belong.

There are in Western India upwards of fifty groups of rock-excavations, belonging to the three great sects,—Buddhists, Brahmans, and Jains,—and of these the great majority are within the limits of the Bombay Presidency, or on its immediate borders. Besides these there are a few insignificant groups in Sindh, the Panjâb, Beluchistân, and Afghanistân.

Geographically the Cave-Temples are distributed very irregularly, but the principal localities in which they exist may be enumerated as follows :—

1. In the province now known as Kâṭhiâwaṛ—, the ancient Saurâshṭrâ, forming the peninsular portion of Gujarat, between the Gulfs of Khambay and Kachh,—there are about half-a-dozen groups of caves scattered along the ranges of hills that run parallel to its southern coast. In these groups there are about 140 separate excavations.

2. In the islands of Salsette and Elephanta close to Bombay there are at least 130 caves,—all within 9 miles north or south of the head of the Bombay harbour at Trombay, where stood the old town of Chêmula—probably the great mart known to the early Alexandrian merchants as Semylla or Timula.1

3. Not quite 80 miles from Bombay as the crow flies, a little to the north of east, is the old city of Junnar—probably the Tagara of Ptolemy and the Periplus,—round which are several groups containing not less than 120 separate caves, while at Harischandragaḍ, Pulu Sonala, and Nânâghat, about 16 miles to the west of it, there are together about 25 more.

4. About 50 miles east of Bombay and 42 south-west of Junnar is Kârlê, where there exists one of the finest Buddhist Cave Temples in India, and within a radius of little more than 20 miles from it are about 60 caves, several of them of special interest.

5. A line drawn southwards from Poona nearly parallel to the Western Ghâṭs or Sahyâdri Hills, passes through groups at Śirwal Wâi, and Karhâḍ, embracing about 80 caves.

6. Along the Konkan, on the western side of the same range, between the hills and the sea, at Kuḍâ, Mhâṛ, Chipalun, &c. the number of caves may be estimated at 60 more.

7. Within a distance at most of 50 miles from the railway leading from Bombay to Nagpur, and lying almost in a straight line between Nasik and Pâtur, 20 miles east of Akola, are the important groups of Nâsik, Ankai, Elurâ, Aurangabad, and Ajaṇṭâ, with others of less note, numbering about 150 caves.

8. About 250 miles E.S.E. from Bombay, and 130 W.N.W. from Haidarabad is the small village of Karusâ, where, and at Dhârasinwâ, 40 miles to the west, and Kalyâna—the old Châlukya capital 30 miles south-east from it,—there are about 120 caves, some of considerable dimensions, though others are small and insignificant.

9. On the north of the Narmadâ in Malwa are the groups at Bâgh, Dhamnâr, and Kolvi—neither of great importance; and, lastly, far to the south, on the banks of the Malaprabhâ in Belgaum district are the caves of Bâdâmi and Aiholê, architecturally among the most interesting Brahmanical groups in India, especially as affording a fixed date, by which that of others can be compared.

This brings up the total to about 900 caves, and there are a few of little note scattered in ones and twos over the same area, so that we may safely estimate the total of known caves in the West at over 900; besides many which have not yet been visited by any European, and of which consequently no record exists.

These are divided primarily into three classes according to the sects by whom or for whose use they were hewn out, viz., Buddhists, Brahmans, and Jains. The earliest examples we have belong to the Buddhists, and date from the middle of the third century B.C., but excavations belonging to this sect extend from that date down to near the end of the seventh century of our era, thus ranging through between nine or ten centuries. They are also the most numerous class, fully 75 per cent of the whole being Buddhist caves.

The next, in order of time, are those of the Brahmans, whether Śaiva or Vaishṇava, which range from about the fourth to the eighth century of our era, or perhaps later. Of the whole, about 18 per cent of the excavations are Brahmanical, but a large proportion of them are of very considerable dimensions, but, except at Karusâ, and some scattered caves in the Sâtârâ district, few of them are small, whereas among the early Buddhist caves there are many which are insignificant.

Lastly, there are the Jaina Cave-Temples, which are much less numerous than those of either of the preceding sects, and of which the earliest may belong to the fifth or sixth century, and the latest perhaps to the twelfth, they are the least numerous of all, not exceeding four per cent of the whole.

We may thus estimate their numbers as follows :—

If to these we add the Eastern caves, described in the first part of this work, it may safely be assumed that the Rock-cut Temples of India, known at the present day, amount to more than a thousand separate excavations.

All such excavations, it will be understood, were for religious purposes, some being temples—Chaityas, or Halls devoted solely to worship, others monasteries, or Vihâras consisting of a hall for assembly, sometimes with an inner shrine for worship, and with cells for monks; some were Dharmaśâlâs, with or without cells, where councils or assemblies were held; while in the more complete Buddhist establishments there were, first, the temple; secondly, one or more monastic halls with surrounding cells; and occasionally also separate dwellings, or hermitages for ascetic monks.

For purposes of description, these works may be classified as follows :—

I.—Buddhist Cave-Temples may be divided into two great classes: first, those which were executed, so far as can be judged from style or inscriptions, before the Christian era or during the first century after it. These belong to the Hinâyâna sect and are generally plain in style, and are devoid of images of Buddha for worship.

II.—Buddhist Cave-Temples belonging to the Mahâyâna sect of a date subsequent to the year A.D. 100, after which images of Buddha first began to appear. These images gradually in the course of time supersede the earlier dâgoba or relic-shrine, until, in the latest examples, the personages represented become numerous, and the pre-eminence of Buddha himself seems to have been threatened by the growing favour for Avalôkitêśwara Bôdhisattwa, who, in Nepal, under the better known name of Padmapâni, had become the favourite divinity of the populace.

III.—The Brahmanical Caves: The Brahmans were probably first led to excavate Cave-Temples in imitation of the Buddhists, and as a means of pressing their candidature for a larger share of popular favour. Their works are very similar to the later Buddhist Vihâras, only without the side cells for monks—such being unnecessary in what were meant only as places of public worship for a religion in which monasticism was not an element. The shrine is usually in the back wall of the Vaishṇava temples, but in those of the S'aiva sect it is generally brought forward into the cave with a pradakshiṇâ or passage for circumambulation round about it.

IV.—The Jaina Caves are the least numerous, but among them are one or two very fine ones. They also are on the plan of the Buddhist Vihâras, sometimes with cells in the walls, but more distinguished by numerous figures of their Tîrthaṅkaras or Jinas, who hold the same place in their system as the various Buddhas do in that of the Buddhist sect. The Jains are now divided into two sections; the Svetâmbaras or white-robed community, who are of more recent origin than the Buddhists,3 and the Digambaras or naked Jains, who are generally understood to be an older sect than the followers of Buddha. It is to this latter division that all the Jaina caves belong, and as yet, with the exception of a small late group in the extreme south of the Peninsula, they have been found only in the Dekhan and Rajputana, or in the region ruled over by the Râṭhoṛs or Balharâs and Châlukyas.

• 1. Ptolemy (Geog. VII. i. 6; VIII. xxvi. 3), and (I. xvii. 4); and the author of the Periplus Mar. Æryth. (§ 53); see below, p. 205.)
• 2. The Jaina excavations in the rock at Gwalior extend down to the 14th and 15th centuries, but as these are not included in the limits of the Bombay Presidency, they are omitted in the above enumeration, but will be noticed further on, after those in the west have been described. They consist of upwards of 50 separate excavations, but all of very modern date.
• 3. Stan. Julien's Mem. sur les Cont. Occ. I., 163, 164.