584 GREEK HORIZONTAL CURVES IN MAISON CARREE AT NIMES. 



rose and Peiinetborne. Thiers<3li, however, in the main accents and 

 develops the point of view of Penrose. The views of the hitter as to the 

 theory of the cnrves liave naturally been most familiar to English and 

 American students, and as his measurements are our only authority for 

 the facts, his theories have naturally been generally accepted by his 

 English and American readers. The explanation of Penrose moves 

 from the accepted fact that there is a tendency to optical downward 

 deilection in the straight line of an entablature below the angle of a 

 gable or pediment. It is his theory that these lines of the entablature 

 were accordingly curved upward in order to counteract this defection. 

 As to the curves of the flanks, Penrose regards them as a ccmseqnence 

 incident originally on the methods pursued for the entablatures under 

 the pediments, and then adds: 



"We may attribute the use of this relinement to the. feeling of a 

 greater appearance of strength imparted by it, to the appreciation of 

 beauty inherent in a curved line and to the experience of a want of 

 harmony between the convex stylobates and architraves of the front 

 and the straight lines used in the flanks of the earliest temples. And 

 further, if we may suppose the lirst examples of its application on the 

 flanks to have occurred in situations like those in which the two 

 temples above mentioned (viz: the Parthenon and Olympian Jupiter 

 Temple) are built, the presence of a delicate, but not inai)preciable 

 curve in what may be considered as Nature's great and only horizon- 

 tal line may possibly have combined with other causes to have sug- 

 gested its use."^ 



Although Penrose is distinctly of the view that the hardness and 

 dryness of modern copies of Greek architecture are due to the absence 

 of these refinements, his efltbrt is in each case of the various refinements 

 quoted at the opening of this paj)er to look for an oi)tical correction as 

 distinct from an optical illusion; and yet for the most important curves 

 of all, viz, those of the long sides of the temple, he does not even sug- 

 gest that an optical correction was needed. 



We come finally to the views of Boutmy, Philosophie de TArchitec- 

 ture en Grece, 1870, who returns to and revives the idea of Hofler of a 

 perspective illusion, but still confining his explanation to an effect from 

 one point of view, viz, that opposite to the center of the sides or ends 

 of a temple. 



Now, the importance of the observation of curves in the Maison 

 Carree is that they were not applied to the iiediments at all, but exclu- 

 sively to the sides. The theory of an optical correction is therefore 

 insuflicient, and the theory of a perspective illusion appears to be the 

 only one left us; but this theory has never jireviously been announced 

 as an explanation for the construction of curves in plan convex to the 

 point of vision. It is, however, clear that all curves in plan convex to 

 the line of vision produce an effect of curves in elevation. I am indebted 

 to Prof. William E. Ware, of Columbia College, for the information that 



' The line referred to is that of the sea aloug the horizon. 



