^90 William Patten 



origin ; 2) that part of the hypodermis from whicb tlie optic ganglion 

 originateci agaiu becomes thickened and divides into twodistinctlayers, 

 the outer one forming the corneal hypodermis, and the inner, the 

 ommateal layer, consisting of the retinophorae surrouuded by their 

 circles of retinulae. I must insist that, if the description I bave given 

 of the structure of the compound eye is eorrect, thenitisimpossible 

 for the sense cells, either the so-called retinulae of Gre- 

 NACHER with their rhabdoms, or the retinophorae, to bave 

 reached their present position, as outgrowths of the brain. 

 This mistaken conception is undoubtedly due to the close proximity of 

 the ommateal hypodermis to the brain , and to their common point of 

 origin. The mesodermicingrowth, between the inner, hypodermic, and 

 the outer, neural layer did not, in ali probability, give rise, as Bobketzki 

 believes, to the pigmented cells and perforated membrane (the latter 

 has no existence) between the crystalline cone cells and the retinulae, 

 but to the often pigmented mesodermic cells between the basai mem- 

 brane and the optic ganglion. 



Weismann (49j and Carrière (19) have likewise shown that 

 the entir e optic layer arises from a hypodermic thickening, and 

 not from any outgrowth of the brain. 



Vision in the Compound Eye. 



The facts which I hope have been made clear in the preceding 

 pages necessitate a modification in the supposed process of seeiug in 

 the compound eye, as advocated by Grenachek, — the so-called 

 Müller's Theory of Mosaic Vision. 



The facts which bear on this point are the foUowing : 



(1) I have shown that the so-called rhabdom of Grenacher has not, 

 in any case examiued by me, the structure he has ascribed to it. 



(2) The rhabdom does not owe its origin to the pigment cells, or 

 retinulae. 



(3) What Grenacher regards as a highly specialized rhabdom (or 

 the pedicel , that should consequently possess the greatest sensibility as 

 well as greatest number of nerve fibres, is, in fact, entirely devoid of 

 those nerve fibrillae which we have shown to be the light percipient 

 Clements. 



(4; A more accurate knowledge of their structure (the rhabdoms) 

 shows that they can in no way be regarded as homologous with 



