696 William Patten 



clegree, to the greater number of ommatidia in the eye of the fly. but 

 it is probably due in a much greater degree to the difficulty of invent- 

 ing satisfactory methods of experimentation. 



It seems as though Grenacher had sought refuge under the 

 shadow of Müller's genius for the propagation of some theory of 

 vision that would go with bis own observations. I must confess that I 

 cannot see any evidence of extraordinary brilliancy in Müller's cele- 

 brated Theory of Mosaie Vision; bis genius has, however, been 

 ampie to proteet, even up to the present day, his »geistige Lieblings- 

 kind« as Grenacher calls it. 



A necessary deduction from the theory of mosaie vision, and one 

 which Müller himself recognized, has, apparently, been forgotten by 

 Grenacher in his discussion of vision in the compound eye. Müller 

 says, as Grenacher has quoted, »Die Insecten sehen weder nach diop- 

 trischen noch nach katoptrischen Gesetzen, sondern bloß durch eine 

 nähere Bestimmung der Beleuchtung!« I understand, by this 

 rather loose expression, that Insects are only able to distinguish light 

 from darkness, and it cannot be denied that it is a necessary deduction 

 from Müller's theory. But it is certain that many insects can per- 

 ceive objects with great precision, therefore the theory cannot be 

 true. If only a single ray of light from any part of a small object 

 — say ten feet distant — eutered each rhabdom , then the points of 

 origin of these rays of light would be so minute, few, and far apart, 

 that no perception of form would be possible. On the other band, if 

 we bring to our aid the crystalline cone and corneal facets, as has 

 been done by Exner, we would be no better off. since a widely di- 

 vergent cone of light entering the corneal facet would , by repeated 

 reflection, be concentrated as a confused mixture of light at the 

 apex of the cone, giving absolutely uo effect of form, while only com- 

 paratively great changes in the amount of light would be perceptible. 

 Moreover a most important point has been neglected by the advocates 

 of Müller's theory, and more espccially by Exner, in attempting to 

 explain the function of the crystalline cone. Even if we suppose with 

 him that the latter concentrates at its inner end all the rays of light 

 entering it, it is perfectly evident that the repeated reflection 

 will not make the rays parallel. Now the inner end of the 

 crystalline cone in its most perfect condition never reaches the 

 outer end of the narrow pigmented tube in which the sup- 

 posed rhabdoms are, therefore the rays of light will di- 

 verge from the apex of the cone at nearly the same angle 



