1893. BIOLOGICAL THEORIES. 211 



function which I have ascribed to tentaculocysts. In position, how- 

 ever, these bodies correspond exactly with the rudiments' of the 

 tentaculocysts in a scyphistoma. 



The function of the tentaculocyst-rudiment, supposing it to have 

 only one function of importance, is to develop into a tentaculocyst. 

 The development in such forms as I have examined (and have failed 

 to identify !) is direct, and exhibits nothing which justifies the 

 assumption that they have any specially " larval " function, or 

 function not connected with the production of the adult organ : and 

 though we may be unable to say at what precise stage they come to 

 be capable of performing the " tentaculocyst-function," I think we 

 are justified in believing that the efficient performance of that function 

 after the organ is fully developed may reasonably be regarded as the 

 chief factor which has determined the evolution of these organs under 

 the guidance of Natural Selection. 



Considering the resemblance in form, and, to a certain extent, in 

 structure, and the identity in position of these rudiments on one hand, 

 and of the marginal bodies of Haliclystus on the other, a community 

 of origin is hardly doubtful ; in other words, the influences (or some 

 of them) which have determined, or helped to determine, the presence 

 and the structure of either may reasonably be supposed to have 

 played a part in the case of the other also. 



According to the usual view, the Lucernarians are the unmodified, 

 or little-modified, descendants of the common ancestor of Scypho- 

 medusae in general. This view further suggests that Lucernavia is the 

 " most primitive," and that Haliclystus is " higher " or more modified ; 

 and that all free-swimming medusae (Scyphomedusae only, I mean, of 

 course) are descendants of forms which, like Haliclystus, exhibited no 

 " alternation of generations " — no strobilation, while remoter 

 ancestors still were very similar indeed to Lucernavia. Such a view 

 is an almost necessary consequence of the Recapitulation Theory. 



For convenience, I will now use the names Lucernaria and Hali- 

 clystus a little loosely, so as to include all hypothetical ancestors pos- 

 sessing the distinctive characters of these genera. 



The orthodox view just set forth brings us face to face with a 

 problem which presents, I believe, insurmountable difficulties, i.e., how 

 to account for the evolution in Lucernaria of marginal bodies so as to 

 give rise to Haliclystus. 



We must, in attacking the problem, assume either — (i) that the 



^ I use the term " rudiment " in its true sense, and as equivalent to the German 

 " Anlage." Use may justify a change in the meaning of a word, but the persistent 

 misuse of the terms "rudiment" and "rudimentary," when "vestige" and 

 "vestigial" are meant does not appear to me to justify the change in this case, 

 especially as we have, so far as I know, no other words to put in their place ; for the 

 German word " Anlage" is so utterly " un-English " that its adoption in English (or 

 even American ! ) writings appears to be open to serious objection. I deem it best, 

 therefore, to use "rudiment" for what it really means, and to ignore the prevalent 

 misuse of the term, in the hope that it will soon die out. 



P2 



