THE RADIATING ORGANS OF THE DEEP SEA FISHES. 205 



3. CONCLUSIONS. 



The results of the examination of tlie minute structure of tlie orjrans 

 of the fishes collected by the " Albatross " whicli one can, with more or less 

 certainty, consider as radiating, as described above, illustrate again (heir 

 great diversity. Not only do tlic oeellar, disc-shaped, and tubular organs 

 fundamentally differ, but even within these three main groups very great 

 variations are met with. 



Tiie best known of those organs are the oeellar. A highly developed 

 organ of this kind, such as the compound organ of Chauliodus is — apart 

 from the connective-tissue support, the nerves and blood ves.sels — conipo.sed 

 of a pigment sheath, a reflecting layer, an inner region of large, highly 

 specialized conic cells, and a middle and an outer mass of specialized 

 polyedrical cells of anotlier kind. Apart from the blood vessels, nerves, 

 and connective tissue, which, as occurring in all organs, of course do not 

 concern us here, none of these structures are found in all of the oeellar 

 organs described above as radiating. The pigment sheath is absent in the 

 smallest organs of Chauliodus itself. In these, and in others with a pigment 

 sheath, the reflecting layer is absent. The large conic cells of the inner 

 region are absent in the simple organs and in the highly differentiated 

 compound organs of Ai'gyropelecus. The polyedrical cells of the middle 

 and outer region of the compound organs of Chauliodus which usually 

 occur in simple organs are replaced l)y long spindle-shaped cells in all the 

 compound organs of Lychnopoles and in the anterior compound organs 

 of Stomias. 



An analogous diversity is met with in the radiating discs. Whilst in 

 Halosaurus and Ipnops highly specialized cylindrical or spindle-shaped 

 elements, obviously forming a very essential part of the organ, are met 

 with, such cells seem to be entirely absent in the discs of Bassozetus and 

 Leucicorus. 



In the tubular organs the diversity is not so great, although here also 

 we find the cells forming the walls of the glandular tul)es sometimes high, 

 as in Malthopsis, and sometimes low, as in Pachystomias. 



Even the rule, in general holding good, that the outer covering of these 

 organs is transparent and suited to allow the radiation to pa,ss, is not univcr- 



