GREEK HORIZONTAL CURVES IN MAISON CARREE AT NIMES. 587 



iially lowered in the same direction, and that this was done for per- 

 spective illusion. Maspero is one of the authorities who mentions this. 

 Mentions are also made of this by liawlmson and by Prof. Reginald 

 Stuart Poole. 



Y. 



Altbougli tbese various observations point to a perspective purpose 

 in the Egyptian and Greek curves, I do not wish to appear to antagonize 

 the view that optical refinements were used in Greek architecture to 

 correct optical illusions, for I believe that they may liave been so used; 

 but I wish to point out that the theories which are confined to correc- 

 tion are insufficient to meet all the facts, and that the theories which 

 have considered the creation of optical illusions to have been one 

 purpose of the refinements are now materially strengthened. 



The existence of a temple at Nimes having- curves on the flanks 

 without having them in the entablature of the pediments tends to 

 antagonize the view of Penrose that the correction of a downward 

 optical deflection below the pediment was the first cause of the intro- 

 duction of the curves m Greek architecture. The temple of jS^eptune 

 at Paestum is quoted by Penrose, in support of his view, as having 

 only curves under the pediments, but strange to say, this temple at 

 Paestum has been subsequently announced by Jacob Burckhardt to 

 have convex curves on its flanks in horizontal planes.' This observa- 

 tion is also quoted by Thiersch. Thus I close my paper by pointing- 

 out that we have at Paestum one ancient Greek precedent for the 

 curves in plan at Nimes, and that both point to Egyptian influence. 

 The city of Nimes was settled by a colony of Alexandrian Greeks from 

 JEgypt. It apx^ears therefore probable that the curves in Greece were 

 derived from Egypt and had the same purpose, but that the curves in 

 the Egyptian courts were generally changed to curves in vertical planes 

 by Greek art. This was a more refined expedient for attaining the 

 same end, less conspicuous in buildings using colonnades for exterior 

 porticoes as distinct from buildings using colonnades for the interiors 

 of courts. It is comparatively easy to sight for a bulging- curve on the 

 exterior of a buikling, but more difficult to sight for it in the interior 

 of a court. I was not able, for instance, to sight for the curve at 

 Medinet Habou without going on the loof of the portico, but at Nimes 

 I was able instantly to sight for the bulge on the long sides from the 

 level of the street. These facts, therefore, coincide with the view that 

 the general j^urpose of the curves in Greek art was connected with the 

 wish to have them inconspicuous, and that the curves at Nimes repre- 

 sent either a direct influence from Egypt or the coarser taste of the 

 Roman period. On the other hand, the flank curves of the J^eptune 

 temj)le of Paestum, which is a very early Greek building, will represent 

 the period of direct Egyptian transmission to Greece. 



' Der Cicerone. 



