Eyes of Molhiscs and Arthropods. 689 



tion, since such processes are often omitted. In Pecten^ where it is 

 pvetty certain that the eyes wcre formed by the modification of in- 

 vaginated cups. outogeny gives uo indication of such a i)rocess. If it 

 is once agreed that the compound eye is a modified ocellus , theu the 

 safest eourse would be to study the development of the simplest and 

 most primitive ocelli. It is ueedless to say that this has not, as yet, been 

 done. Basing our supposition upon the facts obtained in the Mollusca, 

 concerning the development of ganglionic cells in general, and of nerve 

 branches, or nervous ceutres of sense organs in particular, we are 

 eompelled to apply the same principles to the Arthropod eye, and 

 although we ha ve no such embryological evidence, the same method of 

 nerve endings i. e. intercellular nerves, — in the most primitive con- 

 dition, reaching to the outer extremities of the cells, — furnishes very 

 good reason for supposing that the ontogenetic and phylogenetic devel- 

 opment of nerve cells must be essentially the same in the two groups 

 under consideration. If, then. this is true, we can no more admitthat, 

 in Arthropods, the sense cells and their nerve ends are outgrowths from 

 the brain, than in the Mollusca, unless we suppose a case analogous 

 with that found in the Vertebrates, where a part of the brain, originally 

 external, has become evaginated to form a retina with inverted rods. 

 This supposition, however, will not apply to the Arthropod eye, even 

 leaving other difficulties out of consideration , because the rods are not 

 inverted. 



The observations of Bobretzki are founded upon the supposi- 

 tion that the crystalline cone cells and the retinulae form two distinct 

 layers, and he was therefore led to mistake the corneal hypodermis — 

 not knowing of the existence of such a layer. — for the crystalline cone 

 cells. I have myself had the opportunity of making sections of some 

 young lobsters' eyes in a comparatively late stage of development, and 

 found that the corneal hypodermis was then much more highly devel- 

 oped than in the later stages. This fact would also explain why 

 Bobretzki has asserted that the outer ends of the crystalline cones ave 

 surrounded by four cells, believed by him to be identical with the 

 nuclei of SExAiper Balfour, Vol. II, p. 397). If the development 

 is the same here as in the Mollusca, the foUowiug must be the 

 processes undergone: il) a tliickening of the cephalic hypodermis, 

 giving rise by cell proliferation to two layers, an inner one, the brain, 

 and an outer one forming the permanent hypodermis; that part of the 

 brain arising from the seat of the future eye gave rise to the optic 

 ganglion, which isnever entirely separatedfrom the seat ofits 



Mitteilungen a. d. Zoolog. Station zu Neapel. Bd. VI. 46 



