336 DISCOVERY REPORTS 



cells (Fig. 14, phot.c. 3), if the term " cell" may now be properly used, no nuclei are to 

 be seen, and the granular material occupies the limits of the cell membrane completely. 

 It presents the appearance of having faintly defined fibrils which run more or less 

 perpendicularly to the outer surface of the organ. From a comparison of this organ 

 with that next to be described, however, I believe that these are not fibrils, but are an 

 indication that the granular substance is segmented into a number of blocks. Disposed 

 within the "cells" of thjs layer is a central body (Fig. 14 and Plate XXV, fig. 3, c.b.) 

 of entirely obscure origin. Nothing which could be described as representing possible 

 stages in its differentiation have been observed. It is well defined, apparently granular, 

 and stains more deeply than the surrounding substance. It has, however, none of the 

 appearance or staining reactions of a nucleus, and it closely recalls the granular zone of 

 the pleopod photophores (p. 331) and the basal cap of the photogenic cells of the carapace 

 photophores (p. 340). 



Reasons for supposing that stages in progressive differentiation are exhibited by the 

 group of cells described, and also that the granular substance they contain is not 

 cytoplasm, are afforded by the existence of parallel phenomena in the photophore 

 behind the last thoracic limb of H. grimaldii (see later p. 337), and in those on the fifth 

 thoracic limb of Systellaspis affi/iis (see later p. 356). In both these instances the 

 granular material is radially and segmentally arranged, and in S. affinis it is beyond 

 doubt not cytoplasm. 



The group of cells described above are almost certainly the photogenic cells, and in 

 life are probably yellow in colour. Dr Kemp in his notes describes the organ as golden 

 yellow when viewed from beneath, and otherwise surrounded by deep red pigment. The 

 red pigment, acting as a screen or reflector behind the luminous cells, is almost certainly 

 carried by the tissue represented by the cup-shaped group of nuclei (Fig. 14 and 

 Plate XXV, fig. 3, n.r.c.) on the inner side of the photogenic cells. Although I have 

 designated all these nuclei as belonging to the reflector, it is possible that some belong 

 to very early and undifferentiated photogenic cells or to other tissues such as sheathing 

 or supporting fibres of the nerve bundle which innervates the photophore. This nerve 

 bundle (Fig. 14 and Plate XXV, fig. 3, phot.fiv.) leaves the main trunk of the nerve 

 running down the limb (Plate XXV, fig. 3, Liiv.) as a branch at the distal end of the 

 merus, and its neurofibrils are seen among the nuclei of the reflector layer. The precise 

 mode of communication of the neurofibrils with the photogenic cells could not be 

 discovered with the methods it was possible to employ. 



The whole photophore is separated from the integument by the highly vacuolated 

 zone seen in Fig. 14. No trace of differentiation of the chitin in the form of a lens 

 exists in connexion with the organ, and no muscles are associated with it. But although 

 simpler in structure than the pleopod photophores, it presents considerable difiiculties 

 in its cytological interpretation. These are mentioned later (p. 374) in this paper. 



