358 DISCOVERY REPORTS 



Microscopic study of the photophore shows that this substance is deposited in a very 

 regular manner. It occurs in the form of radially segmented spherical masses, which ap- 

 pear in the sections as rosette-shaped structures (Fig. 26 and Plate XXVI, fig. 6, phot. n. 2) 

 most of which show no central cavity. On the inner side of the organ, however, most of 

 the spherical masses do show a cavity (Fig. 26 and Plate XXVI, fig. 6, phot.u. 2), and a 

 close study of this part of the organ suggests that it is a growing zone exhibiting, in 

 passing towards the outer surface of the organ, successive phases in the differentiation 

 of the segmented spheres of granular material. A similar condition was suspected in 

 the organs behind (p. 337) and actually on (p. 334) the fifth thoracic limb of Hoplopfwrus. 

 The organ under discussion will be described on the assumption that such intermediate 

 phases are truly represented, although without a study of developmental stages of the 

 animal it is difficult to form a really definite conclusion. 



It appears that the precursors of the spherical granular masses are represented by 

 certain large nuclei on the inner side of the organ (Fig. 26 and Plate XXVI , fig. 3 , phot.n . i ) 

 which ciifi^er from their neighbours in being as much as three or four times larger and in 

 showing a less dense chromatin disposition. I refer to these nuclei as photogenic 

 nuclei without wishing to imply that I am aware of their precise role in the mode of 

 functioning of the organ, but because there appear to be no other structures in the 

 organ capable of interpretation as photogenic structures. In no case has it been possible 

 to detect the cell membrane of any corresponding cell with certainty. 



The nuclei appear to increase in size until they are of about the same diameter as 

 the spherical granular masses, and at the same time their chromatin forms a uniform 

 and delicate reticulum (Fig. 27 a, b). Intense degeneration of the nucleus, resulting in 

 its complete disappearance, then follows (Fig. 27 c, d.n.), being often foreshadowed by 

 the appearance of an ill-defined central cavity. The degeneration of the nucleus appears 

 to leave the nuclear membrane intact (Fig. 27 c, d, e, f, l.m.), but whether this is 

 truly the nuclear membrane, or in reality a cell membrane, as would be suspected from 

 a comparison of the organ with that of the photogenic streak and organs on the fifth 

 thoracic limb of Hop/ophorus, I am unable to say. Whatever the nature of the membrane, 

 it forms the boundary of a space empty but for the inclusion of certain structures 

 shortly to be described (Fig. 26, phot.u.cl.a.). This space strongly recalls the clear areas 

 encountered in the photogenic cells of HopJophonis and in the pleopod photophores of 

 Systellaspis affinis itself. A noteworthy feature in the degeneration of the nuclei is that 

 their last remnants are always to be found within the limiting membrane on the side 

 remote from the external surface of the photophore (Fig. 26, e.g. phot.n. 2). 



Small vacuoles (Fig. 27 c, vac), usually apparently empty, but sometimes appearing 

 to contain a coagulated fluid, are scattered within the limiting membrane, and may be 

 sometimes observed in the nucleus at an early stage of degeneration (Figs. 26, 27, vac). 

 The first appearance of the granular substance, which forms the main mass of the organ, 

 has been noted as minute deposits, segmentally disposed, on the inner side of the 

 limiting membrane opposite the degenerating nucleus (Fig. 27 c, gr.d.). 



With continued reduction of the nucleus (Fig. 27 d) these deposits increase in size. 



