REPOET ON THE DEEP-SEA FISHES. 297 



There can be no doubt that these structures in the species examined by me are in 

 no way similar to those described by Ussow ; I find, however, a difference between them 

 and the gland-tubes of the internal spherical part, which has not been noticed by 

 Leydig. 



At first sight it appears that the radial lines in the part occupying the cup are much 

 closer together than in the interior of the sperical portion of the phosphorescent organs. 

 Spherical cells, as in the latter part, cannot be distinguished so readily in the cup, and 

 the substance also seems to he more transparent and less granular. The radial stripes are 

 much closer together in Echiostoma than in Astronesthes. 



The structure of this part of the organ is the follo^^^ng : — 



From the disc of large granular cells, situated at the constriction, fibres originate, 

 which extend upwards in a vertical dii'ection towards and nearly to the outer surface, 

 terminating 0'04 mm. below it. 



They extend below to a trumpet-shaped, thickened base, and become more and more 

 slender distally. They are in direct connection with the substance of the basal disc, and 

 interspersed with granular spindle-shaped cells which appear more abundant in the basal 

 than in the distal part of the fibres. These spindle-shaped cells have spherical nuclei 

 identical with the nuclei in the cells of the disc. The bulk of the fibres seem to be 

 composed of nerves ; bloodvessels have not been observed in them. 



The fibres are on an average 025 mm. apart. Their structure is very transparent 

 and they appear closer than they really are in longitudinal sections, because one generally 

 sees several layers of them, but transverse sections show their true distance. 



Attached to each of these fibres are cells, generally long and slender but sometimes 

 also stout and spindle-shaped, in such a way that one layer of them encloses each fibre. 

 These cells appear to be attached to the fibres by one of their ends and they project from 

 the fibres like the hairs on the tail of a squirrel. Sometimes, particularly when the cells 

 are very slender, they extend nearly parallel to the fibres and produce the closely striated 

 appearance of this structure which is observed in some cases (PI. LXXI. fig. 33). Some- 

 times, i^articularly where the cells are stout, they radiate from the fibres in an oblique 

 direction, never exceeding an angle of 45°, and they invariably point outwards. 



These cells are very tender and transparent, their nuclei easily escape observation 

 in specimens not specially preserved, and their contours can only be distinguished Iiy 

 means of very fine sections and good lenses. Under ordinary circumstances this whole 

 mass appears structureless and granular. 



If we now compare this description with the statements of Ussow and Leydig, 

 mentioned above, we find that although the differences are great, stUl the observations 

 are not altogether incompatible. 



Ussow's observation of the organ in the fresh state in Chauliodus shows us that 

 there is a transparent substance with radial lines in this part of it. 



(ZOOL. CHALL. EXP. — PART LVII. — 1887.) Lll 38 



