586 GREEK HORIZONTAL CURVES IN MAISON CARREE AT NIMES. 



accordingly by beams duriiiii' and since tlie excavations not quite com- 

 pleted in 1891. Measurements taken by me in all three courts at Luxor 

 show curves in all lines of columns at the bases, all convex to the cen- 

 ters of the courts, varying from 1^ to 7 inches. It is clear at Medinet 

 Habou that the lower curves in the lines of tlie basis and in the lines 

 of columns near the bases were comparatively slight and that the curve 

 was obtained in the architrave and cornice (as it was at Nimes) by 

 leaning forward the center columns. This would explain the movement 

 of the masonry which has required the columns at Luxor to be shored 

 up by timbers. 



All earthquakes and other forces tending to disintegrate these build- 

 ings, such as pulling down and destroying the accessible parts of the 

 temple, would tend to exaggerate the lean of the center columns and 

 bring about the threatened downfall now imminent at Luxor. My 

 observations at Edfou point the same way. On all four sides of the 



— ^ y y y"^ y y"^ y"^ y y^ 



I 



^v^v^OOOO^^O^, 



Fig. 3. 



DIAfiKAMS SIIOWINC) AN OPTICAIj EFFECT OF CURVES AND OBLIQUITIES IX LINES WHICH AHE IN FACT' 



STRAIfiHT AND PARALLEL. 



ch. Optisi-he Tniischung" 



I Gebiete iler Architectur. ) 



court I have measured curves in the line of the bases of \h inches on 

 each side of the court. Very heavy curves — of 10 inches in one case — 

 appear in the cornice lines, but the cornices have moved forward and 

 the original lean of the center columns has been exaggerated by acci- 

 dental tipping. The joints of the columns have parted at the rear, and 

 it will require careful examination and survey at Edfou to show how 

 much of the upper curve is due to movement of the masonry and 

 how^ much is due to construction. One main fact remains to be men- 

 tioned for Egyptian temples. Although their curves have so far been 

 utterly ignored and neglected, excepting by Pennethorne and myself,' 

 the existence of other perspective illusions is admitted for Egyptian 

 temples by Egyptological experts. 



It is noted by a number of authors that the temples were generally 

 built with pavements rising toward the sanctuary and with roofs grad- 



' I must make an exceptioa for Prof. Allan Marquand, of Princeton, -who has briefly 

 noticed in the American Journal of Archicology the discovery of Pennethorne at 

 Medinet Habou. 



