444 MR FRANK E. BEDDARD ON THE STRUCTURE OF 



generally, a chitinous structure known as the rhabdomere. Bullae* has not 

 figured or described more than a clear point at the summit of each cell of the 

 retinula in Cymothoa, which, however, he regards as the equivalent of the 

 rhabdomere (Sehstabchen). In all the forms which I investigated the rhab- 

 domere of each retinula cell was very largely developed, rather more so in some 

 cases than in others. It always presented a remarkably striated appearance 

 (see figs. 5, 6), clear lines alternating with dark lines ; the appearance, in fact, 

 suggested that it was composed of a number of separately formed rods or 

 plates, between which a certain amount of pigment had been deposited ; its 

 structure is unlike that of most other Isopods and Crustacea generally, where 

 each rhabdomere is a simple plate applied to the edge of the retinula cell. The 

 rhabdomere of Hyperia galba\ seems, however, to resemble it in certain points. 



At the upper extremity the seven rhabdomeres are in contact, although even 

 here they retain for the most part their independence, i.e., they do not become 

 fused. This is illustrated in fig. 15, which represents a section through the 

 upper part of the retinula and rhabdom. Lower down the several rhabdomeres 

 diverge from each other; each is somewhat conical in form, and projects down- 

 wards from its secreting retinula cell (see figs. 5, 12). The rhabdomeres do not, 

 however, project freely into an interspace left between the retinula cells. 

 They are closely enveloped by two large round cells, which fit into the inter- 

 space between the retinula cells. 



These two cells are each furnished (figs. 5, 8, 9, h) with a distinct granular 

 nucleus, in the centre of which is a spherical nucleolus ; the boundary between 

 the two cells was in no case very distinct. They are quite homogeneous and 

 transparent ; the outer surface of the cells was occasionally spotted with 

 pigment granules, which may or may not be deposited on the outermost layer 

 of the cells themselves. In transverse (fig. 7) and longitudinal (fig. 5) sections 

 these cells are seen completely to envelop the rhabdom, which is plunged into 

 their interior. In the species illustrated in figs. 5, 7, 8, 9, these cells were very 

 conspicuous in teased preparations both depigmented and undepigmented ; in 

 the species illustrated in fig. 14 they were not conspicuous in such preparations, 

 owing to the larger size of the rhabdom ; in sections of both, however, they 

 were perfectly obvious. 



In my account of the eye of Serolis\ I have described and figured a pair of 

 precisely similar cells, showing an exactly corresponding relation to the rhab- 

 dom of the cells of the retinula. 



In this structural feature the Cymothoidse show therefore a close resem- 

 blance to the Serolidae, and both differ in a corresponding degree from all other 



* Loc. tit., p. 513, pi. 46, fig 13. 



f Carriers, loc. tit., p. 161, fig. 124, rh. 



| Loc. tit., p. 20 et seq. pi. ix. 



