353 
I wish to say here, in regard to Warase’s recently published 
views on the nature of the compound eye, Johns Hopkins Circulars, 
1889—90, that the large cell in Belostoma is like the ganglion 
cell described by him in the lateral eyes of Limulus, but in my 
opinion the facts here presented show that there is no homology be- 
tween them, and that in Limulus this cell cannot have the morpho- 
logical importance which he attributes to it. Again, the retinula cells re- 
semble in general appearance the gigantic cells at the bottom of the 
larval ocelli of Acilius, but development shows that in Vespa 
there is no invagination of the ommatidia and no bending of the re- 
tinula cells, so that the rhabdomeres and the segments of the true 
erystalline cone must be formed along the sides of their respective 
cells and not on their outer ends. It is impossible, I believe, to har- 
monize these facts with WATASE’S views, the essential features of which 
I had already independently formulated and was ready to maintain, 
when ‘these observations rendered them untenable. 
I was formerly in doubt as to the origin of the corneagen of the 
frontal ocelli (Eyes of Vespa, p. 212), but I am now convinced, after 
the examination of new material, that in Vespa, Aphis and For- 
mica it arises by a process of delamination, similar to that which 
gives rise to the corneagen of the convex eye; consequently it is very 
different from that of the larval ocelli of Acilius. This fact 
strengthens the idea suggested in my paper upon Eyes of Acilius, 
p. 167, that the frontal ocelli were closely related to the convex eyes. 
Moreover the development of the corneagen indicates that the retina 
of the frontal ocelli, like the ommateum, is formed by the union of 
independent organs, each one surrounded by indifferent cells. As I 
have already shown that the larval ocelli of Acilius are also formed 
by the union of sense buds !), it becomes highly probable that all the 
higher sense organs of Arthropods are formed by the aggregation of 
sense buds, and, for reasons to be presented elsewhere, it is also prob- 
able that the same condition prevails in the eyes and nose of Verte- 
brates. Moreover the cephalic sense buds of fishes may be compared 
with the sense hairs, pits, and perforated spines in the cephalothorax 
of Arthropods. For example in Vespa and Aphis the circle of cells 
surrounding each ommatidium forms, at first, a long chitenous tube, 
open at the top; if the rudimentary hairs of the cone cells actually 
1) This fact has an important bearing on Warast’s speculations, but 
he has entirely ignored it, as well as my other observations on the po- 
sition and character of retinal rods, and the nerve fibrillae in them. 
