THE EYE IN CERTAIN CYMOTHOID^!. 451 



unless it be supposed that the crystalline cones of Serolis, &c, are not homo- 

 logous with the crystalline cones of the Decapod, but with the hypodermal 

 layer of the latter, which on this hypothesis must be assumed to be absent in 

 Serolis. This, however, is a purely gratuitous assumption, and does not seem 

 likely. On the other hand, my results do harmonise with those of Grenacher, 

 as regards the distinction between the crystalline cone and the rhabdom. 



As Dr Patten has pointed out, the retina of the Molluscan eye contains 

 clear cells and pigmented cells, and he has shown reasons for believing that 

 the former are mainly connected with the process of vision. The eye of 

 Annelids is also similar in this respect. In Nereis, for example, the retinal 

 layer contains clear cells (termed by Carriere * " secret-zellen ") and pig- 

 mented cells. 



I have no facts to bring forward which would allow of the hypothesis that 

 the hyaline cells of the Serolidae and Cymothoidse are visual cells in the sense 

 of being connected with nerve fibres. 



The main facts in the present paper and the conclusions to which they lead 

 are as follows : — 



The Serolidse and Cymothoidse possess eyes which differ in certain important 

 particulars from the compound eyes of all other Crustaceans, as at present 

 understood. 



The points of difference concern the retinulae. Each retinula consists, in the 

 first place, of four (Serolis) or seven (Cymothoidae) elongated cells resembling 

 those of other Isopoda ; each of these cells secretes a chitinous body, the rhab- 

 domere. In Cymothoa (Bullar) the individual rhabdomeres retain their dis- 

 tinctness. In other Cymothoidse and in the Serolidse the rhabdomeres become 

 fused to form an axially placed rhabdom, which has often a complicated 

 form, and in which a large quantity of pigment is deposited. The Serolidse (not 

 the deep-sea species) and many Cymothoidse possess a pair of large hyaline 

 nucleated cells, surrounded by the other retinula cells. In the axis of these, 

 and enclosed by them (in the Serolidaa), is a delicate fibre, passing back as far 

 as the ommatial membrane, and expanding anteriorly into a conical body, which 

 appears to penetrate into the axis of the rhabdom. 



In young specimens of Serolis schythei the future hyaline cells are small and 

 granular, and enclose the extremity of this axial cone and fibres, which may be 

 partly a product of their activity, though chiefly formed by the other retinula 

 cells. Each retinula therefore consists of two central clear cells (corresponding 

 in number to the cells of each vitrella), surrounded by four or seven pigmented 

 cells. 



The pigmented retinula cells are connected with transversely striate fibres, 

 which pass into the ganglion, and are generally regarded as nerve fibres. The 



* Figured by Carri£:rk, Ioc. cit, p. 31, figs. 26, 26a. 

 VOL. XXXIII. PART II. 3 U 



