PHOTOPHORES OF DECAPOD CRUSTACEA 349 



from those in Systellaspis debilis. The curious gradation in size and abrupt truncation 

 characteristic of S. debilis is not apparent. The uniformly dense cytoplasm of the 

 photogenic cells (Fig. 21, cyt.phot.c.) is concentrated distally in the now familiar manner, 

 and again the greater part of the proximal portion of the cells appears to be completely 

 without contents (Fig. 21, phot.c.cl.a.). The smoothly rounded narrow proximal ends 

 of the photogenic cells rest on a plate-like granular zone (Fig. 21, gr.z.) which has the 

 same appearance and staining reactions as in Hoplophonis. On its inner surface this 

 granular zone shows a slight conical elevation corresponding with the more pronounced 

 cone of Hoplophonis, but no indication of any segmentation corresponding with the 

 number of photogenic cells is seen. 



The granular zones of Hoplophonis and Systellaspis affiiiis agree not only in being 

 conical, but in having the apex of the cone basally directed, whereas in S. debilis it is 

 distally directed, pointing towards the photogenic cells. Furthermore, in S. debilis not 

 all the photogenic cells make contact with the granular zone, the most peripheral ones 

 apparently lying against the fibrous reflector (Kemp, igiob, pi. lii, fig. i ; pi. liv, fig. i). 



Innervation of the photophore is carried out by a bundle of nerve fibres (Fig. 21, 

 phot.nv.) which approaches the granular zone directly, without the intervention of a 

 striated zone as in Hoplophonis. Peripheral elements of this bundle, presumably non- 

 nervous, do not impinge on the granular zone but diverge to form a funnel-like fibrous 

 capsule, enclosing the organ, which probably acts as a reflector. Numerous small, 

 dense, and rounded nuclei are scattered throughout the fibres of this reflector, as they 

 are in Hoplophorus and Systellaspis debilis, and probably represent chromatophores. 

 Superficially and peripherally in the same plane as the nuclei of the photogenic cells 

 occur greatly elongate nuclei of uncertain significance. They are clearly distinguishable 

 from the neighbouring photogenic nuclei on account of their delicate vacuolation. 

 Similar nuclei are present in the pleopod photophores oi Hoplophonis (p. 332), and also, 

 as far as can be seen from Kemp's observations {i()iob, pi. lii, fig. i), which are con- 

 firmed by my own examination, in those of Systellaspis debilis. For convenience I refer 

 to them as nuclei of reflector cells without denying the possibility of their subserving 

 some other function, such as the secretion of an active principle of luminescence. 



It would appear improbable that the above-mentioned differences can exercise any 

 significant influence on the precise mode of functioning of the photophores, but in the 

 absence of any extended observations on their luminescence or of any detailed in- 

 vestigation into their physiology it would be premature to assume this. In structure the 

 pleopod photophores of S. affinis occupy a position somewhat intermediate between 

 those of Hoplophonis novae- zealandiae and Systellaspis debilis. In S. affinis the trun- 

 cation and well-marked gradation in size of the photogenic nuclei seen in S. debilis is 

 absent, but the full complexity of innervation and musculature of Hoplophonis novae- 

 zealandiae is not attained. 



6-2 



