866 | | The Arthropod Eye. 
This is the most important difference from Sedgwick’s very brief 
account. Certainly Kennel’s figures do not support the view that 
we have here to do with a ‘cerebral eye ;’ but it must be borne 
in mind that he and Sedgwick are studying different species. 
Next comes a preliminary communication on the development 
of the ocelli of Hymenoptera, by Carriére (Zoologischer Anzei- 
ger, IX, pp. 496-500, 1886). Here the optic epidermis becomes 
at first two cells deep, and then these become obliquely invagi- 
nated, both layers retaining their “ normal” position. The outer 
layer forms the lens-generating cells, the inner the retina-form- 
ing ones. The cells which are not invaginated become elon- 
' gate and, together with the invaginated lens-building cells, form 
the corneal lens. These cells never lose their connection with 
the epidermis while those of the retina’do. The cavity of in- 
vagination does not close up but is occupied by the corneal. lens. 
-~ Thus runs Carriére’s account; but it is very difficult to under- 
stand it, as it is not illustrated. 
= Thave now a few observations of my own to record. In Cran- 
gon the eyes arise from invaginated pits, and here as in the spr 
der, Agelena, we have three layers to deal with. These are the 
unmodified epidermal layer and the two walls of the invaginated 
te pouch. The one of these latter which comes to be the more external 
- I have termed the retinogen because from it arises the retinal ele- 
as ‘ments; the other, from analogous reasons, is the gangliogen. So 
_ far the account is closely similar to that of Locy as outlined 
above. I have, however, been able to trace the development of 
versely, each giving origin to five cells which also lie in a radius 
f the eye. Of these the outer forms the retinophora : 
ure above while the others develop into the pigment cells. 
, | of the structures embraced in the bracket zg in the figure are 
, all the parts of the adult eye, which in outline is as follows. The 
ny x 
Se ee ee ae Ay ee ee aes vir 
similar ‘way the cells of the retinogen elongate and divide trans- i 
of the 
