THE RADIATING ORGANS OF THE DEEP SEA FISHES. 201 



walls similar to those observed by Leydig and others in certain cells of the 

 ocellar radiating organs described above. If on the otl.er hand intercellular 

 spaces are really present — and the appearance is in favor of this assumption 

 — the question arises whether they are present in these cell plates in life or 

 merely artifacts produced post mortem by transverse cell shrinkage. In the 

 lower end of each of these cells a large oval granular nucleus is observed. 

 These nuclei form a single layer in the basal part of the plate. In the 

 marginal zone nuclei are also observed higher up, and the same is the case 

 in the central groove, where a few nuclei are likewise seen at a higher level. 

 It is doubtful, however, whether these latter belong to the cylinder cells. It 

 seems rather that they are the nuclei of slender spindle cells here interposed 

 between the cylinder cells. The protoplasm of the cylinder cells is pretty 

 granular and stainable with acid-fuchsin and haematoxylin. These stains 

 are taken up more by the terminal than by the central parts of the cells. 

 The cells of the central groove show a marked aflinity for picric acid ; other 

 stains are hardly at all absorbed by them. 



Transverse sections (Plate 11, Figs. 56, 57) show that the intercellular 

 spaces between the cylinder cells above mentioned do not extend quite to 

 the upper surface of the cell plate. Here the upper ends of the cells, in 

 most cases considerably curved, are in direct contact with each other and 

 form a superficial pellicle apparently quite continuous and in sections some- 

 what resembling a cuticle. 



I have previously described ('87, pp. 309-313, Plate 73, Figs. 45-48, 54-64) 

 the structure of the radiating discs in another species of Halosaurus, II. 

 macrochir. In this species a pair of large glandular radiating organs also 

 occurs under the gill covers. These have a different structure. The 

 smaller spindle-shaped discs on the sides of the body, however, resemble the 

 radiating discs on the head of II, radiatus pretty closely. These are in H. 

 macrochir attached to the outer side of the scales over the great lateral slime 

 canal. Below the centre of each disc the scale is perforated by an oblique 

 tube through which a nerve and blood vessels ascend to the base of the disc. 

 In the disc itself a lower layer, chiefly occupied by an exceedingly close 

 reticulation of capillary blood vessels, and an upper composed of slender 

 cells arising vertically, standing side by side, and forming a high epithelium 

 can be distinguished. The individual elements of the latter are spindle- 

 shaped or irregularly cylindrical. In transverse sections through the disc 

 these cells appear separated by clear intervals wider than in the cylinder 



