ARTHROPODAL VISION". 53 



accommodation. But this implies that the receptive elements 

 are too far back to be in the focus of parallel rays, which has 

 not been proven . 



It is true that the images of distant objects must be 

 extremely' minute, the foci being shorter, in some cases at 

 least, than the foci of the corneules of compound eyes, and 

 the focal plane is correspondingly close to the lens, but it may 

 be shown, perhaps, that the receptive elements extend to that 

 plane. It has been stated, moreover, that the lens is made up 

 of layers of different curvatures and different indices of 

 refraction, from which F. Dujardin inferred that there would 

 be as man}' successive images behind the lens as there were 

 zones with these differences. Adding experiment to theory, 

 he exhibited at the Academy of Science, Paris, in 1847, a 

 telescope whose objective was composed of several zones, and 

 w^hich, the ocular remaining fixed, gave four distinct images 

 for as many distances in the field of view. He claimed also 

 that he had obtained similar results with actual ocelli. 



The compound eye was at one time supposed to be an 

 assemblage of simple eyes, but that view is not now enter- 

 tained. It is a highly complex organ, of which the facetted 

 cornea is only a superficial layer. The structure and func- 

 tions of the related parts have been carefulh- studied by many 

 able investigators, one of w^hom, Professor Sigmund Exner, 

 of Vienna, concludes that the zoologist who has mastered the 

 compound eye will find the vertebrate eye an easy task. 



At an early date it was found that behind each of the cor- 

 neules, as the facetted divisions of the cornea are termed, 

 there is a transparent body, now called pyramid, crystalline 

 cone or cr>'stalline lens T Plate i, Fig. 3J, and that these pyr- 

 amids or cones were separated from each other by opaque 

 pigment. This arrangement suggested the mosaic theor\- of 

 Johann Miiller, who supposed that the rays of light collected 

 by any corneule would be converged to a point on a retinal 

 sifrface at the inner end or apex of the cone, and he describes 

 the theoretical result as follows : — 



