Eyes of Mollusca and Arthropods. 677 



some traces of their primitive aiTaugement upon restricted areas oftlie 

 Compound basal membrane. But cur studies upon the basal meuibrane 

 of Fenaeiis liave showu just the opposite of what we should expect, 

 provided the supposition was correct. for we have proved that the 

 inner ends of cells belonging to neighboring ommatidia, 

 instead offormingisolatedgroups, inte r min g- le with each 

 otherinsuchawaythat.unlcssonefollüwedthecellsfrora 

 each ommatidium inward, it would be impossible to say 

 whether thev belouged to the same, or different omma- 

 tidia! We have, moreover, seen that in the Spiders and Myriapods 

 each ocellus receives a special nerve brauch. Now if each ommatidium 

 represents an ocellus , it should also receive a sing-le nerve brauch, 

 going to its cells alone. But this is not the case, for I have shown that 

 in Penaeus, although the number of nerve bundles is equal to the num- 

 ber of ommatidia, each nerve bündle breaks up into se- 

 veral smaller ones going to four different ommatidia! 



As far as the third question goes — Does Embryology show any 

 evidence of fusion? we are not in possession of any very important 

 evidence. Neither my own observations, which in this direction have 

 been very scanty, nor those of others, have shown any indications that 

 the compound eye was formed by the fusion of ocelli. 



We have now to ask the same question concerning the progressive 

 development of a single ocellus into the compound eye, and then, by 

 weighing the evidence for. or against each supposition, we may be 

 able to arrive at some conclusion. 



Do we find any evidence in the ancestors ofthe Insecta of trans- 

 itional stages between a simple and a compound eye ? In Myriapods the 

 answer is short: all eyes are alike, and are the simplest form of simple 

 eyes. In Spiders, the anterior ocelli retain their primitive, undifiPerentiat- 

 ed condition. and the percipient Clements form a retineum; but in the 

 posterior eyes the ommatidia have undergone a series of changes to- 

 wfe,rds a higher form. But what are the changes necessary to convert a 

 primitive ocellus into a compound eye? In order to simplify matters, 

 let US first determine what the simplest condition of the compound eye 

 is. The presence of the corneal facets in certaiu higher forms, only, of 

 Insects and Crustacea indicates that they are of late origin ; moreover 

 the presence of a thick corneal hypodermis, and the absence of corneal 

 facets in such animals as Branchipus, the Isopods, Amphipods and many 

 Insects, show this condition to be a primitive one. The retinophorae 

 are usually four in number, but we have seen in Penaeus and Galathea 



