Râjâgṛiha, or Râjgir as it is now popularly called, was the capital of Magadha or central India during the whole period of Buddha's ministrations in India. It was the residence of Bimbasara, during whose reign he attained Buddhahood, and of Ajâtaśatru, in the 8th year of whose reign he entered into Nirvâṇa, B.C. 481, according to the recently adopted chronology (ante, p. 24,25). It is quite true that he resided during the greater part of the 53 years to which his mission extended at Benares, Ṥravasti, or Vaisaka (Lucknow1), but still he frequently returned to the capital, and the most important transactions of his life were all more or less connected with the kings who then reigned there. Under these circumstances it is hardly to be wondered at that Râjgir was considered almost as sacred in the eyes of his followers, as Jerusalem became to the Christians, and that such pilgrims as Fa Hian and Hiuen Thsang, naturally turned their steps almost instinctively to its site, and explored its ruins with the most reverent care. Long before their time, however, the old city had been deserted. It never could have been a healthy or commodious city, being surrounded on all sides by hills, which must have circumscribed its dimensions and impeded the free circulation of air to an inconvenient extent. It consequently had been superseded long before their time, in the fifth and seventh century, by a new city bearing the same name but of much smaller size just outside the valley, to the northward. This, however, could never have been more than a provincial capital. The seat of empire during Aśoka's reign having been transferred to Palibothra (Patna) on the Ganges, which we know from the accounts of Megasthenés was an important city in the days of his grandfather Chandragupta. At the same time, any ecclesiastical establishments that might have been attracted by the sanctity of the place, mast have been transferred to Nalanda, between 6 and 7 miles due north from the new city, where there arose the most important monastic establishment connected with Buddhism that, so far as we know, ever existed in India.2 Fortunately for us Hiuen Thsang has left us a glowing description of the splendour of its buildings, and of the piety and learning of the monks that resided in them. With this, however, we probably must remain content, inasmuch as some excavations recently undertaken on the spot have gone far to prove that all the remains now existing belong to buildings erected during the supremacy of the Pâla dynasty of Bengal (765 to 1200 A.D.). The probability is that all the vihâras described by Hiuen Thsang were erected wholly in wood, which indeed we might infer from his description, and that the monastery was burnt, or at least destroyed, in the troubles that followed the death of Ṥíladitya in 650 A.D.,3 and they consequently can have no bearing on the subject we are now discussing.

Under the circumstances above detailed leading to the early desertion of Rajgir, it would of course be idle to look now for any extensive remains of the buildings, if it ever had any, in stone or any permanent material, and equally so to expect any extensive rock-cut Vihâras or Chaitya caves in the immediate vicinity of such an establishment as that at Nalanda. Practically we are reduced for structural buildings to the Jarasandha-ka-Baithak, above described (woodcut No. 2), and for rock-cut examples to one cave, or rather pair of caves, known as the Son Bhandar or Golden Treasury.

The larger of these two caves is very similar in plan to the Kama Chopar cave at Barabur and nearly of the same dimensions, being 34 feet by 17 feet.4 Its walls are perfectly plain to the height of 6 feet 9inches, and thence rise to 11 feet 6 inches in the centre of a slightly pointed arch. The doorway is towards one end and has the usual sloping jambs of the period, the proportion between the lintel and sill being apparently as 5 to 6, which seems to be somewhat less than the proportion at Barabar.5 This doorway is balanced towards the other end of the cave by a window nearly 3 feet square, which is a decided innovation, and the first of its class known to exist in India. A still greater advance in cave architecture is the existence of a verandah 8 feet deep, extending along the front, and at one end some way beyond the cave. It existence is quite certain from the mortice holes still remaining in the rock into which the ends of the rafters were inserted, as shown in the woodcut. Its having been added here is specially interesting, as it certainly is, like the window, an improvement, and almost as certainly an advance on the design of the Barabar caves, and as clearly anterior to that of the Katak caves, where the verandahs are, as a rule, cut in the rock, with massive pillars in stone forming part of the original design.

As will be explained in the subsequent pages of the work, nearly all the ornamentation of the Chaitya caves in the West down to the Christian era was either a literal copy of wooden construction, or was executed in wood itself, generally teak, attached to the rock and in very many instances, as at Bhâjâ, Bedsa, Karlê, and elsewhere, the actual woodwork still remains where it was fixed some 2,000 years ago. From the representations of buildings at Buddha Gaya and at Bharhut and from the front of the Lomas Rishi cave quoted above (woodcut No. 3) we know that precisely the same mode of decoration was employed in the eastern caves, that was usual in the western ones, but in none of the Behar caves have we any evidence of wood being so employed except in the verandah of this cave and in one or two doubtful instances at Katak. One example may not be considered as sufficient to prove a case, but as far as it goes, this seems to be a first attempt to remedy a defect that must have become apparent as soon as the Barabar caves were completed. With very rare exceptions all the caves on both sides of India have verandahs, which were nearly indispensable, to protect the openings into the interior from the sun, but in nearly all subsequent excavations these were formed in stone, and became the most ornamental parts of the structure.

The other Son Bhandar cave is situated at a distance of 30 feet from the larger one and in all respects similar except that its dimensions are only 22 feet by 17. The roof has almost entirely fallen in, and only one mortice hole exists to show that it had a wooden verandah similar to that in front of the other cave.

Between these two caves a mass of rock is left standing in order to admit of a flight of steps being cut in it, leading to the surface of the rock above the roof of these two caves. Whether this led to an upper storey either in woodwork or brick, or whether there was not a dagoba or shrine on the upper platform, can only be ascertained when some one visits the spot after having his attention specially directed to this object, from its analogy with what is found in other places. From the arrangements of some of the Katak caves, I would rather expect to find the remains of an upper storey. But it may be very difficult to determine this, for whether it was a stupa or dwelling, if in brick, it may have been utilised long ago. As before mentioned, General Cunningham seems to think that a Vihârain brick, but with granite pillars, existed in a corresponding situation above the Vapiya and Vadathi caves at Barabar.6 If he is right in this, which seems very probable, it would go far to establish the hypothesis of the existence of a second storey over the Son Bhandar cave.

There seems to be nothing except its architecture by which the age of this cave can be determined. Kittoe, indeed, says “there are some rude outlines of Buddhas carved upon it,” and there is also a handsome miniature Jain temple much mutilated,7 which he gives a drawing of. The Buddhas I fancy are much more likely to be Jaina Tirthankaras, which are so easily added when there is so much plain surface, and as the “temple” shows that the cave was afterwards appropriated by the Jains, nothing is more probable than that they should ornament the walls by carving such figures upon them. Broadley is more distinct. “Outside the door,” he says, “and 3 feet to the west of it, is a headless figure of Buddha cut in the rock, and close to it an inscription in the Ashoka character."8 But as neither Cunningham nor Kittoe saw either; and they do not appear in Peppe's photograph from which the woodcut is taken, we must pause before accepting his statement. On the whole, therefore, taking the evidence as it stands, there seems no good reason for doubting that the Son Bhandar caves belong to the Great Mauryan dynasty, B.C. 319 to 180. At the same time the whole evidence tends to show that they are more modern than the dated caves at Barabar, and that they were consequently excavated subsequently to the year 225 B.C.

We are fortunately relieved from the necessity of dismissing the theory, so strongly insisted upon by General Cunningham, that the Son Bhandar cave is identical with the Sattapanni cave, where the first convocation was held,9 from the fortunate discovery by Mr. Beglar of a group of caves which almost undoubtedly were the seven caves that originally bore that name (Sapta parna, seven leaved).10 On the northern side of the Vaibhâra (Webhára) hill there exists a group of natural caverns, six in number, but there is room for a seventh, and evidence that it did originally exist there. As unfortunately Mr. Beglar is not an adept at plan drawing, his plan and section (pp. 92 and 96) do not make this so clear as might be desired, in fact without his text, his plans would be unintelligible. With their assistance we gather that owing to some abnormal configuration of the rocks there are at this spot a series of fissures varying in width as 4, 6, 8, and 10 feet, and ranging from 6 to 12 feet in depth (p. 96). What their height is is not stated, nor can the fact be ascertained from the drawings, it is not however of much importance11 The real point that interests us most in this instance is, that as in the Jarasandha-ka-Baithak (ante, p. 33) with its 15 cells, we have the earliest known example of a structural Vihâra in India, so here we have the earliest known instance of a rock—we can hardly add—cut Vihâra with 7 cells, and for both of which we have historical or at least traditional evidence, to show that they existed contemporaneously with, if not before, the lifetime of Buddha himself. Like all those, however, which have any claim to an antiquity earlier than the age of Aśoka (B.C. 250), it is a mere group of natural caverns without a chisel mark upon them, or anything to indicate that they were not rather the lairs of wild beasts than the abodes of civilised men.

There are still two other caves or groups of caves at Bajgir, which are of considerable interest from their historical, though certainly not from their artistic value. The first is known as the house or residence of Devadatta, the persistent enemy of Buddha. It is only a natural cavern situated at the foot of the hill in the north-eastern corner of the city at a spot marked M in General Cunningham's map (Vol. III., Plate XLL), but not described by him nor by Mr. Beglar,12 but as it is merely a natural cavern this is of little consequence, except as affording another example of the primitive form of all the earlier caves. In front of it is still to be seen the rock which, according to tradition, Devadatta rolled down from the mountain athwart Buddha's path and wounded a toe of his foot.13

The other group of caves is on the Gridharakuta hill, about 3 miles north-east from the city, is of still greater interest, as it is described minutely by both the Chinese pilgrims as a place much frequented by Buddha and his companion Ânanda. The elder pilgrim describes it in the following terms: “The peaks of this mountain are picturesque and imposing; it is the loftiest of the five mountains that surround the town. Fah Hian having bought flowers, incense, and oil and lamps in the new town, procured the assistance of the aged Bikshus to accompany him to the top of the peak. Having arrived there he offered his flowers and incense, and lit his lamps, so that their combined lustre illuminated the glories of the cave; Fah Hian was deeply moved, even till the tears coursed down his cheeks, and he said, Here it was in bygone days that Buddha dwelt... Fah Hian, not privileged to be born at a time when Buddha lived, can but gaze on the traces of his presence, and the place which he occupied.”14

Neither General Cunningham nor Mr. Broadley ascended the peak high enough to reach these caves; the hill may be 100 to 150 feet in height. It was consequently reserved for Mr. Beglar to make the discovery. He followed the causeway that led to them a few hundred yards further, and hit at once on two about 50 feet apart, which seem to answer to Buddha's meditation cave, and the Ânanda cave as described by the Chinese pilgrims. They are both natural caverns, the larger measuring 12 feet by 10, of irregular shape, but, the irregularities slightly reduced by filling in with brickwork on which are some traces of plaster, and inside there are now found some fragments of sculpture lying about, but evidently of a much more modern date. As Mr. Beglar's map is nearly as unintelligible as his drawings, we are left to conjecture which of the two caves marked upon it are those just referred to, nor how many more exist on the spot. The text says 7, 2+5, but only four are shown, and the other buildings he describes cannot be identified on it.15 Enough, however, is shown and said to make it quite clear that these are the caves referred to by the Chinese pilgrims, and to prove to us that, like all the caves connected by tradition with the name of Buddha, they are mere natural caverns untouched by the chisel, though their irregularities are sometimes smoothed down with brickwork and plaster, and that the latter may, in some instances at least, have been originally adorned with paintings.

• 1. I state this deliberately, notwithstanding what is said by General Cunningham in the Ancient Geography of India, p. 401, et seq., though this is not the place to attempt to prove it. Hiuen Thsang, however, places Vaisaka 500 li or 83 miles S.W. from Ṥravasti which can only apply to Lucknow, and Fa Hian's Sa-chi, measured from Canouge or Ṥravasti, equally points to Lucknow as the city where the “tooth-brush tree” grew. Neither of the pilgrims ever approached Ayodhya (Fyzabad), which had been deserted long before Buddha's time. Tf the mounds that exist in the city of Lucknow were as carefully examined, they would probably yield more treasures than even those of Mathura.—J.F.
• 2. See History of Indian Architecture, vol. i., p. 136.
• 3. Hiuen Thsang, vol. i., p. 151.; Ma-twan-lin, J. A. S, B., vol. vi., p. 69.
• 4. Cunningham, Reports, vol. v., Plate XIX.
• 5. The greater part of the information concerning this cave is taken from General Cunningham's Reports, vol. iii. p. 140, Plate XLIIL, but his drawings are on too small a scale and too rough to show all that is wanted. Kittoe also drew and described it, J. A. S. B., September 1847. It is also described by Broadley, Indian Antiquary vol. i, p. 74,
• 6. Reports, vol. i. p. 49.
• 7. Kittoe, J. A. S. B., Sept. 1847, Plate XLTI.
• 8. Indian Antiquary, vol. i. p. 74,
• 9. Cunningham, Arch. Report, vol. iii. pp. 140 to 144.
• 10. Although we may not be able to fix with precision either the purpose for which the Son Bhandar caves were excavated, nor their exact date, it is quite clear they are not the Sattapanni cave, near which, according to all tradition, the first convocation was held immediately after the decease of the founder of the religion. In the first place, a hall, only 34 feet by 17, about the size of an ordinary London drawing-room, is not a place where an assembly of 500 Arhats could assemble, and the verandah, 8 feet wide, would add little to the accommodation for this purpose. It is hardly worth while attempting to refute in any great detail the various arguments brought forward in favour of this hypothesis, for there is no proof except the assertion of modern Ceylonese and Burmese authorities, who knew nothing of the localities, that the convocation was held in a cave at all, and everything shows that this was not the case. The Mahawanso (p. 12) states that it was in a splendid hall like to those of the Devas at the entrance of the Sattapanni cave. Mr. Beal's Translation of Fa Hian (p. 118) makes exactly the same assertion, but with an ambiguity of expression that might be construed into the assertion that it was in and not at the cave that (he convocation was held. But Bemusat's translation, as it is in strict accordance with the more detailed statements of Hiuen Thsang, is at least equally entitled to respect. He says:— " Au nord de la montagne, et dans un endroit ombragė, il y a une maison de pierre nominė Tchheti, c'est le lieu où après le Nirvaṇa de Foě, 500 Arhans recueillirent la collection des livres sacr ės.”^{Foë Ku ë Ki, 272.} Hiuen Thsang makes no mention of a cave, but describes the foundations which he saw of “une grande maison en pierre,” which was built by Ajâtaśatru for the purpose in the middle of a vast forest of bambus. ^{Julien, vol. iii. p. 32.} Even the Burmese authorities, who seem to have taken up the idea of its having been held in a cave, assert that the ground was first encircled with a fence,—which is impossible with a cave,—and within which was built a magnificent hall. ^{Bigandet, Life of Gaudama, p. 354.} The truth seems to be that the modern Buddhists, like the mediaeval Christians in Palestine, thought everything was, or at least ought to have been, done in a cave, but when read with care, there is certainly nothing except in the most modern writings to indicate that this was the case in this instance, and there certainly is no cave in Râjagriha which is fitted or ever could have been made suitable for such a purpose. The convocation was in fact held in one of those great halls of which we have several instances among the western caves. The last woodcut, however, representing one from the rail at Bharhut, 150 years B.C., and one at Kanheri shown in plan, Plate LIV., with the examples to be described hereafter at Mahavâllipur and probably also the Nagarjuni cave at Barabar just described, show us the form of Dharmasalas that were in use among the Buddhists in that age, and were perfectly suited to the purposes of such an assembly. It probably was a building measuring at least 100 feet by 50, like the cave at Kanheri, with a verandah of 10 feet all round. With the knowledge we now have of the architecture of Aśoka's time there would be no difficulty in restoring approximately such a hall, and in a general history it might be well to attempt it, but it has no direct bearing on the history of cave architecture.
• 11. Beglar on Cunningham's Reports, vol. viii. pp. 89 to 99.
• 12. Archaological Report, vol. viii. p. 90.
• 13. Fah Hian, Bears Translation, p. 115; Julien, vol. iii. p. 27.
• 14. Ibid, vol. iii., p. 20.
• 15. The information regarding these caves is not to be found in the body of Mr. Beglar's report, vol. viii., but in a prefatory note, pp. xv to xxi, which makes no reference to the text, which it contradicts in all essential particulars, or to Map XXII., which is equally ignored in the body of the work. In fact, it is very much to be regretted that the manner in which these reports are put together is not creditable to any of those concerned in their production.