Page:EB1911 - Volume 14.djvu/455

Rh in wood (see Plate I. fig. 8). The whole of the superstructure of the Sānchi examples is essentially wooden in character, and we are astonished that it should have stood “for twenty centuries nearly uninjured.” These torans reappear to this day in Japan as tori-i and in China as p’ai-lus or p’ai-fangs. The whole of the surfaces, inside and out, are carved with elaborate sculptures of much interest. A cast of the eastern toran from Sānchi is to be seen in the museums at S. Kensington, Edinburgh, Dublin, Paris and Berlin. On the southern one, an inscription appears to indicate that it was erected about 150



The earlier cave temples are of about the same age as the stūpas; some of those in Behar bear inscriptions of Asoka and of his successor in the 2nd century And the earlier cave façades in western India indicate the identity of style and construction in the patterns from which both must have been copied. These Buddhist rock excavations are of two types: the chaitya or chapel caves, with vaulted roofs of considerable height, the earliest with wooden fronts and later with a screen wall left in the rock, but in both forms with a large horse-shoe shaped window over the entrance. The interior usually consisted of a nave, separated from the side aisles by pillars, and containing a chaitya or small stūpa at the inner and circular end. The façades of these chaitya chapels were covered with sculpture—some of them very richly; and to protect them from the weather a screen was contrived and cut in the rock in front of the façade, with large windows in the upper half for the entrance of light. This mode of lighting by a great arch over the entrance has attracted considerable attention, as being admirably adapted for its purpose. As Fergusson remarked, “nothing invented before or since is lighted so perfectly, and the disposition of the parts or interior for an assembly of the faithful. . . is what the Christians nearly reached in after-times but never quite equalled.”

The second type of rock excavations are known as vihāras or monasteries devoted to the residence of monks and ascetics. They usually consisted of a hall surrounded by a number of cells—the earliest with stone beds in them. In the later vihāras there was a shrine in the centre of the back wall containing a large image of the Buddha. In the Orissa caves, near Cuttach, we have a series of excavations that do not conform to these arrangements: they are early, dating as far back as the 2nd century, but they belong to the Jain sect, which dates from the same age as the Buddhist.

On the north-west frontiers of India, about the Swāt and Yūsufzai districts, anciently known as Gandhāra, are found a remarkable class of remains, much ruined, but that must have abounded in sculptures belonging to the Buddhist cult. It is among these we find the first representations of Buddha and of the characters belonging to the Buddhist pantheon. The influence of classical art manifested in these images leaves no doubt that they were modelled after western patterns, carried thither by Greeks or brought from the Levant by Buddhist emissaries. The scenes depicted, however, have frequently an architectural setting in which we find represented façades with pillars fashioned with distinctly Corinthian capitals. These sculptures we can now assign with confidence, from dated epigraphs, to dates from the last years of the century till the 4th century One inscription of  47 is of a king Gondophernēs, who is mentioned in the legend of the apostle Thomas.

In the time of the great Gupta dynasty, from about 320 to 500, the architectural forms developed in variety and richness of decoration. To the columns were given higher square bases than before, and sometimes a sur-base; the capitals, which previously had a vase as the chief member, were developed by a foliaged ornament, springing from the mouth of the vase and falling down upon it from the four corners, and so lending strength to the neck whilst converting the round capital into a square support for the abacus. Often, too, a similar arrangement of foliage was applied to the early bases; and this form quite superseded the Persepolitan pillar, with its bell-shaped capital, which now disappeared from Indian art. The shafts were round or of sixteen or more sides; pilasters were ornamented on the shafts; and the spires of the temple were simple in outline and rose almost vertically at first and curving inwards towards the summit, which was always capped by a large circular fluted disk supporting a vase, whilst the surface of the tower was covered with a peculiar sort of horse-shoe diaper. This style prevailed all over Hindustan, and was continued with modifications varying with age and locality down almost to the Mahommedan conquest.

In Kashmir from the 8th century, if not earlier, till the Mahommedan conquest we find a style of architecture possessing a certain quasi-classical element which has little if any connexion with the art of the rest of India. The best-known example of this Kashmir style is the temple of Mārtand, about 3 m. east —Plan of Temple of Mārtand. from Islāmābād or Anatnāg, the old capital. It stands in a court 220 ft. long by 142 ft. wide surrounded by the ruins of some eighty small cells, with a large entrance porch at the east end. The temple itself was 60 ft. long by 38 ft. wide, with two wings, and consisted of two apartments—a naos and cella. The trefoiled or cusped arch on the doors of the temple and cells is a striking peculiarity of the style, and may have been derived from the section of the Buddhist chaitya. It is used decoratively, however, rather than constructively. The pillars and pilasters of the portico and temple bear a close resemblance to some of the later forms of the Roman Doric, and have usually sixteen shallow flutes on the shafts, with numerous members in the base and capital. A triangular pediment surmounts the doorways, and on gable-ends or projecting faces are representations of double sloping roofs, much in the style of modern Kashmir wooden roofs, of which also many of the temple-roofs in Nepal are exaggerated examples. The Mārtand temple was, in all probability, built in the 8th century, between 725 and 760, and was erected as a temple of the Sun, one of whose names is Mārtand. For, till the 12th century at least, Sun-worship was quite prevalent in the north and west of India. At a remote village called Buniār is a much better preserved