1Y6 J. W. Fewkes — Deep-Sea Medusae. 



" Those Medusae," writes Haeckel, " may be regarded with 

 greater probability as permanent and characteristic inhabitants 

 of the deep-sea, which have either adapted themselves by 

 special modifications of organization to such a mode of life, or 

 which give evidence by their primitive structure of a remote 

 phylogenetic origin." He then enumerates those which he 

 places in this category, among which are the two remarkable 

 genera, Atolla and JSauplianta. " It is by no means certain," 

 writes Haeckel, " that all the eighteen medusas described below, 

 (Report on Challenger Medusae) are constant inhabitants of 

 the deep-sea." We have discussed the argument drawn from 

 two of the most characteristic of the Acraspeda, viz : Atolla 

 and JSTauphanta, and can readily subscribe to this statement 

 as far as these are concerned. 



The resemblance of Naiipliantopsis and Atolla to Ephyra 

 is believed to have a morphological significance ; Ephyra is 

 thought to be the ancestral form of the Acraspeda, and these so- 

 called deep-sea medusae still preserve the ancestral form with 

 small modifications, except in size, repetition of organs, and 

 certain other characters. Of the development of Atolla or of 

 the Collaspidce we know nothing, and yet a knowledge of 

 this subject is possibly to reveal the solution of important 

 questions. If the mode of growth should prove to be a di- 

 rect development without a Scyphostoma, it would certainly 

 increase my belief that these medusae somehow resemble the 

 ancestral forms. I have already elsewhere shown that among 

 the hydromedusae with alternation of generations and those 

 with a direct development, the latter method is normal 

 while the former is a secondary modification. Among Acras- 

 peda, also, the direct development of Pelagia is the ancestral 

 method, while the formation of a Scyphostoma is a second- 

 ary modification. We should expect to find in Atolla a di- 

 rect development, if it be an ancestral genus. From its mode 

 of life in the high seas we should also expect the same.* 



Abandoning, for the present, as insufficient, any evidence 

 which might be adduced from the structure of the medusae 



* I believe the Lucernarians are degenerate adult Acraspeda, which, have at- 

 tached themselves to the bottom much in the same way as Gassiopea frondosa 

 and become modified in consequence. While it may be said that they are homol- 

 ogous to the Scyphostoma stage, it is not thought that they are ancestral. They 

 are in reality secondarily modified, for the ancestral method of development is 

 direct, without an attached young, in Acraspeda, as in Craspedota. 



While the primitive structure and relationship of Atolla, Nauphanta, and Nau- 

 phantopsis would seem to ally them closely to Ephyra and stamp them as less 

 modified than such genera as Cyanea, in certain anatomical details, they might be 

 regarded as higher even than the last mentioned. We cannot, consequently, draw 

 from their simple relationship to an embryonic form, the conclusion that they 

 have retained that likeness on account of the simpler conditions of deep-water 

 habitat. Nor is the argument drawn from the supposed abortion of the sense- 

 body conclusive, as far as these medusa? are concerned, although it looks plausible. 



