574 GREEK HORIZONTAL CURVES IN MAISON CARREE AT NIMES. 



lines of the building are constructed in curves which rise in vertical 

 planes to the center of each side, but these curves do not form parallels. 



Three main facts api)ear throughout all these various i)henomena: 

 First, an unquestionable purjjose and intention, whatever the purpose 

 and intention may have been; second, an avoidance of all exact ratios 

 in iiroportions, of all exact correspondences in the presumably equal 

 objects, sizes, and spaces, and of all mathematically straight, mathe- 

 matically peri)endicular, and mathematically parallel lines; third, an 

 avoidance of all such irregularities as are easily detected, or as are 

 obtrusively consx)icuous to the eye. 



As regards the curves, they are inconspicuous to the eye unless 

 sighted for from some one angle of the building and along the line of 

 the steps, or of the exterior line of the stylobate (the i)latform on which 

 the temple rests). As viewed even from such an angle, they are so deli- 

 cate as not to be obtrusively consi)icuous. As seen from other points 

 of view, especially opposite the center of the ends or sides of the build- 

 ing, they may be detected by close observation, but there is no point of 

 view from which the eye is not naturally disposed to discount the effect 

 as one of perspective. As there are no straight lines, but only delicate 

 curves when straight lines are viewed in perspective, it is natural for 

 the eye to discount the effect of a delicate curve; for this is what the 

 eye constantly does when the actually straight line is curved by nat- 

 ural perspective. As regards the appearance of inclination in the col- 

 umns, we have the testimony of Mr. Penrose that he was months in 

 Athens before he could determine by the eye without plumbing which 

 way a given column leans, and this fact will describe the delicacy of 

 other deviations from the perpendicular. 



As regards the variations in size of presumably equal objects, or of 

 spacing in presumably equal distances, it maybe said that none of them 

 can be definitely asserted to exist on purely ocular testimony, and that 

 the surveyor's work is necessary not only to determine their amount, 

 but even to determine their existence. Here again the difficulty in 

 definite ocular detection depends on the fact that all objects of exactly 

 corresi)onding size vary in api)arent size according to the ])oint of 

 sight. Hence, when an element of delicate irregularity of size or 

 spacing is artificially produced it is impossible for the eye to avoid 

 discounting this irregularity into perspective effect. Let it be noted 

 here that I do not use the words "perspective effect" as necessarily 

 implying an increase in effect of magnitude. If a I'arthenon capital 

 nearer to the eye be smaller than one next to it, and farther away 

 from the eye, the effect in so far would be to diminish apparent distance 

 between the two capitals, but this would still be an illusive effect of 

 persi)ective appearance, because the ordinary effects of perspective 

 woidd prevent the eye from appreciating an exact equality of size if it 

 had existed. Then, again, if a spectator be facing two unequal adja- 

 cent capitals at exactly equal distances from each, in which case they 



