Ascending from this we enter the middle court, in the centre of which stands what may be considered as the temple itself. It measures 200 ft. by 213 ft., and is crowned by five towers or spires, one on each angle, and one, taller than the others, in the centre, rising to a height of 180 ft. The central tower has four cells, like that at Sadri, one facing each way. The general appearance of these towers may be gathered from the elevation ([Woodcut No. 371]), and from Woodcut No. 375. They are very Indian in character and outline, but, when looked closely into, are unlike anything known in that country. The building which resembles the inner temple most, so far as at present known, is that at Sadri ([Woodcut No. 133]). Its dimensions are nearly the same, 200 ft. by 225 ft.; like this, it has five spires similarly disposed, and four open courts; and at Sadri, as here, there are a certain number of snake-images, which suggest a connexion between the two. But there the similarity ceases. The extraordinary amount of richness and exuberance of detail in the Cambodian temple far surpasses that of the Indian example; and the courts at Nakhon Wat are not courts but water-tanks. How far the lower courts were also capable of being flooded is not clear, nor whether the whole area, 1100 yards square, in which the temple stands, was not also capable of being turned into a lake.[643] Judging from the analogy of the Kashmiri temples, it would seem probable that this may have been the case. If it were, it is difficult to conceive a more fairy-like scene than this temple would have presented, rising from the lake which reflected its forms in the calm stillness of a tropical sunset.
One of the most curious circumstances connected with the architecture of this temple is, that all its pillars are as essentially of the Roman Doric order, as those of Kashmir are of the Grecian Doric.
375. General View of Temple of Nakhon Wat. (From a Photograph by Mr. J. Thomson.)
Even if this is disputed, one thing at least is certain, that no such pillars occur anywhere in India. At Nakhon Wat there is not a single bracket-capital nor an Indian base. The pillars nowhere change into octagons or polygons of sixteen or thirty-two sides,[644] and all the entablatures are as unlike Indian forms as can well be conceived. At Nakhon Wat, also, there are intersecting vaults and ingenious roofing-contrivances of all sorts, but no dome, and no hint that the architects were aware of the existence of such a form. On the contrary, take such a pillar as that shown in Woodcut No. 376: the proportion of diameter to height; the entasis; the proportion between the upper and lower diameter; the capital with its abacus; the base with its plinth; the architrave, &c., are so like the Roman order that it is difficult to conceive the likeness being accidental.
But whoever gave the design for these pillars—and, according to M. Mouhot, there are 1532 of them in this single building—we have abundant evidence to show that the people for whom it was erected were of pure Turanian blood. Without insisting on other facts, there are in every part of the building groups of female figures in alto-relievo. They are sometimes in niches or in pairs, as in the Woodcut No. 377, attached to pilasters, or in groups of four or more. There are a hundred or more in various parts of the building, and all have the thick lips and the flat noses of true Tartars, their eyes forming an angle with one another like those of the Egyptians, or any other of the true building-races of the world. Unfortunately, no statues of men are so attached, though there are several free-standing figures which tell the same tale. The bas-reliefs do not help in the inquiry, as the artist has taken pains to distinguish carefully the ethnographic peculiarities of all the nations represented, and, till the inscriptions are read, and we know who are intended for Indians or who for Chinese or Cambodians, we cannot use the evidence they supply.
376. Pillar of Porch, Nakhon Wat.