36o DISCOVERY REPORTS 



Finally, the last traces of the nucleus and the vacuoles disappear, and the segmental 

 granular deposits are laid down around the complete periphery of the limiting mem- 

 brane (Fig. 27 e), and continue to increase in size centripetally. Fig. 27/ shows the 

 last trace of the cavity within the limiting membrane before its entire elimination by the 

 expanding granular segments. 



With reference to Fig. 26, I have used the term "photogenic nucleus" to indicate 

 one of the large nuclei referred to, whether complete or degenerating, and the term 

 "photogenic unit" for the limiting membrane with its granular deposits. There is no 

 morphological indication, however, as to which of these structures is concerned in 

 light production. The difficulty in the way of regarding the large nuclei and their phases 

 of degeneration as being photogenic is, of course, that they are separated from the 

 surface of the limb by the heavy granular deposit, which although it would form an 

 efficient reflector is not adapted to allow of the passage of light. 



Apparent phases in the development of the granular spheres, similar to those 

 described above, are found in addition around the periphery of photophores of this 

 type, and lend support to the belief that we are here considering a zone of active 

 growth in the developing photophore. 



The significance of the nuclear changes described above, and the nature of the 

 granular substance, will be considered later in this paper (p. 374). 



The photophore is bounded internally by loose and scattered strands of connective 

 tissue (Fig. 26, c.t.) with small deeply staining nuclei (Fig. 26, n.c.t.), and is bathed by 

 the body fluid within the haemocoele of the limb (Fig. 26, hoem.). No special vascular 

 supply has been observed. On its outer side the organ is cut off from direct contact 

 with the chitin of the limb (Fig. 26, chit.) by the chitogenous epithelium (Fig. 26, 

 n.chit.ep.) which passes uninterruptedly over it. This would suggest that the organ is 

 not of ectodermal origin. No local lens-like thickening of the integument has been 

 noted. 



It is of some interest that the white granular substance which is laid down in a regular 

 manner o,n the inner surface of the limiting membranes already mentioned is also found 

 irregularly deposited to a far less extent among the connective tissue fibres at the peri- 

 phery of the organ. Here it is extranuclear, and perhaps even extracellular, suggesting 

 that the limiting membranes around the photogenic nuclei are at least not nuclear 

 membranes. At any rate it is clear that the secretion of this substance is not the especial 

 prerogative of the photogenic nuclei. 



{e) The carapace organs 

 It has already been mentioned (p. 346) that Systellaspis affinis possesses carapace 

 photophores in the form of a long posterior streak and a number of more anterior 

 shorter streaks situated near the lower margin of the branchiostegite. These photophores 

 have been examined, and it was expected that they would prove to have a similar 

 structure to those in the same position in the three species of Hoplophonis examined. 

 Actually they have the same form as the organs on the thoracic limb just described. They 



