There is still a fifth Ratha, belonging to this group, which, though small, is one of the most interesting of the whole. It bears the names of Sahadeva and Nakula, the twins, but in order to avoid confusion it may be well to confine its designation to the first name only, as neither have any real bearing on either its history or use. It stands a little out of the line of the other four, to the westward, and like them it is very unfinished, especially on the east side. Its dimensions are 18 feet in length north and south, by 11 feet across, and the height is about 16 feet. Its front faces the north, where there, is a small projecting portico supported by two pillars, within which is a small cell, now and perhaps always untenanted. The opposite end is, externally at least, apsidal, and so probably if on a larger scale, its interior would have been; as it is, it is too small, being only 3 feet in depth by 4 feet 6 inches in width, to be utilised for any altar or image, and the square form is certainly more convenient in so small an apartment.

The great interest of this Ratha lies in the fact that it represents, on a small scale, the exterior of one of those Chaitya caves, which form so important a feature in all the western groups, but all of which are interiors only, and not one is so completely excavated as to enable us to judge of what the external appearance may have been, of the constructed Chaityas from which they were copied. There is one temple at Aihole dedicated to Śiva which does show the external aisle and apsidal termination, and is probably of about the same age as this Ratha.1 Unfortunately it has been used as a fortification, and its upper storey and roof removed, so that it is of little more use to us now than an interior would be for judging of what the effect of the exterior may have been above the first story. From the evidence of this Ratha it seems almost certain that in the larger examples there was a range of small cells in the roof of the aisles, which would naturally be much wider in constructed examples than in caves where there was no possibility of introducing light except through openings in the façade. We may also gather from the Aihole example and other indications that an external verandah surrounded the whole, and if this were so the cells would have been placed over the verandah, and the roof of the aisles used as an ambulatory.

One other peculiarity remains to be noticed. As will be observed from the woodcut representing the exterior, the interior of the roof is represented as semicircular, though the exterior is naturally pointed, or at least with a ridge to throw off the rain. This is the case with all the Chaitya caves in the west of India, and probably was the case with all sacred buildings. On the other hand the evidence of the Ganeśa temple and of Bhîma's Ratha here, as well as that of the Behar caves, would go far to prove that in all secular or quasi secular buildings, the form of the roof was that of a wooden framework of pointed form both externally and internally.

One of the most curious illustrations connected with this little Ratha is to be found a very long way off, in the recently excavated monasteries at Takht in Bahi and Jamalgiri in the Yusufzai country, not far from Peshawar. In both these monasteries the principal court is surrounded by a number of small cells, very similar to this Rath. In that at Jamalgiri the court is circular 45 feet in diameter and is surrounded by 16 cells ranging from 6 feet 2 inches across to 11 feet 8 inches; four of them according to General Cunningham's plan and restoration are more than 11 feet across and 20 feet in height, and consequently larger than this Rath.2 The restoration of their façades is fortunately easy, not only from the numerous fragments found on the spot, but because of the great number of sculptured representations of them which exist there, used as frames for sculpture. One of these, with its sculptures, is shown in the woodcut below, and represents in all essential particulars just such a façade as this. The lower part quite open to the interior; the middle storey, in this instance, with lean-to roofs instead of cells, and above this an overhanging roof terminating upwards in an ogee form.3

Each of these 16 cells at Jamalgiri, according to General Cunningham, originally contained a figure of Buddha seated in the usual cross-legged conventional attitude. This Rath may have contained a Linga, if that emblem was introduced into the south as early as 700 A.D., or more probably a figure of Śiva in some of his manifestations, but which, not being cut in the rock, has disappeared.

The age of the Jamalgiri monasteries has not yet been settled, they are certainly earlier than the Raths at Mahâvallipur, but their distance in time cannot be very great. The Buddhism there developed is very similar to that found in the later caves at Ajaṇṭâ and elsewhere, ranging from the fifth to the seventh century of our era, which cannot consequently be long subsequent to the date of these Peshawar monasteries , which cannot be very far removed from that of the Mahâvallipur Raths.

It may probably appear to some that more space have been devoted to these Raths, than is justified either by their relative dimensions or their artistic merits, but the fact is, that it seems almost impossible to overestimate their importance to the history of Buddhist architecture. One of its most remarkable peculiarities is, that though we have some 700 or 800 caves spread over the 1,000 years during which Buddhism flourished in India, we have not, excepting the Topes and their rails, one single structural building, and among the caves not one that has an exterior; without exception the latter are only interiors with one façade through which the light is introduced. No Buddhist cave has even two, much less three, external sides, and not one has an external roof. Under these circumstances, it is an exceptional piece of good luck to find a petrified Buddhist village—on a small scale it must be confessed—and applied to the purposes of another religion, but still representing Buddhist forms just at that age when their religion with its architectural forms were perishing out of the land where it arose. At the same time, no one who has paid any attention to the subject can, I fancy, for one moment doubt that Arjuna’s and Dhamarâja's Rathas are correct models on a small scale of the monasteries or vihâras of the Buddhists, that the Ganeśa temple and Bhîma’s Raths are in like manner models of the Śâlâs or Halls of the Buddhists, that Draupadî's Rath represents a hermitage and Sahadeva's a chapel belonging to the votaries of that religions. The forms of the two last named have fallen into disuse, their purposes being gone, but the other two have been adopted by the Dravidian Hindus, and repeated over and over again throughout the south of India, and continue to be used there to the present day in all the temples of the Brahmans.

In the present state of our knowledge it is to be feared that it is idle to speculate on the mode in which these anomalous phenomena occurred, but it may fairly be inferred from them, that in the seventh century of our era there was no original and appropriate style of Hindu architecture, in the south of India. It seems also most probable that the Pallavas, or whoever carved these Raths, came from some more northern country, where they were familiar with the forms of Buddhist architecture, and that when they resolved to erect temples to their gods, in their new country, they came to the conclusion that they could not do better than adopt the forms with which they were familiar. Having once adopted it in the rock, they seem to have applied it to their structural temples, and gradually dropping those features which were either inappropriate or difficult of execution, by degrees to have developed the Dravidian style of architecture as we find it practised in the south of India from their time to the present day.

If all this is so, it may at first sight seem strange that no trace of this many-storeyed style of architecture is to be found adapted to Hindu purposes in those countries where the style first originated and had long been practised, and was consequently familiar to all classes of the inhabitants. The answer to this difficulty seems, however, not far to seek. In the north of India the Hindus early possessed styles of their own, from whatever source it may have been derived. They had temples with large attached porches, or Mantapas, and cubical cells surmounted by tall curvilinear towers, in which no trace of storeys can be detected. Having thus their own sacred forms, they had no occasion to borrow from a rival and hated sect, forms which they could hardly be expected to admire, and which were inappropriate for their sacred purposes. The result seems consequently to have been that the two styles grew up and developed side by side, but remained perfectly distinct and without showing any tendency to fuse or amalgamate at any period of their existence.

• 1. Burgess’s Report on Belgam and Kaladgi, Plates LI. and LII.
• 2. These particulars are taken from Gen. Cunningham's Reports, vol. v. pp. 23 et seq. 43 el seq., and Plates VIII., IX., and XIV. See also Hist. of Indian Architecture, p. 170 et seq., woodcuts 92 to 95.
• 3. One of the most interesting peculiarities of the Peshavar, or rather Gandhara sculptures, is that it would not be difficult to select from among them several that would form admirable illustrations for a pictorial Bible at the present day. One, for Instance, is certainly intended to represent the nativity. The principle figure, a woman, is laying her child in manger, and that it is intended to be such is proved by a mare with its foal, attended by a man, feeding out of a similar vessel.  Above are represented two horse heads in the position that the ox and the ass are represented in mediæval painting.

  A second represents the boy Christ disputing with the doctors in the Temple. A third, Christ healing a man with a withered limb, either of which if exhibited in the Lateran, and re-labelled, might pass unchallenged as sculptures of the fourth or fifth centuries.

  The scene in the annexed woodcut may, in like manner, be taken to represent the woman taken in adultery. Two men in the background, it will be observed, have stones in their hands ready to throw at her. The similarity in this instance is little mote farfetched than in the others, but still sufficiently near to render a comparison interesting. The study of these most interesting sculptures is now rendered impossible from the closing and dispersion of the India Museum.