Page:Brinkley - Japan - Volume 2.djvu/108

 as a proof of Grecian affinities. But the inference seems to have been hastily drawn; for whereas there are innumerable proofs that the principle of entasis was fully understood by the Japanese, and that they used it intelligently as a device to correct the hollow appearance which the sides of high pillars or long horizontal beams would present if perfectly straight, the so-called entasis of the Hōryu-ji columns is exaggerated to such a degree that they have distinctly bellied outlines. They do not, in fact, show entasis at all, but are intentionally convex. It is possible, of course, that the idea of entasis may have been derived by the Japanese from Greece viâ India, but the practical application of it is seen in the work of later architects, not in the Hōryu-ji columns, and there is no solid reason to suppose that the Japanese borrowed the principle at all and did not discover it by the exercise of their own remarkably accurate observation.

Nothing hitherto written on the subject of Japanese sacred architecture can be compared, in point of accuracy of observation and technical knowledge, with the accounts embodied in essays contributed by Mr. J. Conder to the Royal Institute of British Architects. As these essays are not accessible to the general reader, the following extracts may be quoted here:—

The popular temples of Japan have generally one open enclosure with a grand two-storeyed gateway, continually left open to the public. A water-basin and