5. General Conclusions and Summaries. §3 



9. Infected males of I. scorpio which have been completely modified externally, only 

 retaining the copulatory style to identify them as male, are found to have the gonad and its 

 ducts so completely reduced in most cases as not to be visible either by dissection or serial 

 sections. 



10. These perfectly modified males of I. scorpio, if they recover sufficiently to regene- 

 rate gonads, may regenerate a perfect hermaphrodite gland capable of producing mature ova 

 and spermatozoa. 



11. There is no evidence that crabs of any other category, i. e. males or females 

 which have been imperfectly modified externally, ever regenerate hermaphrodite gonads. 



The question which presents itself at the outset is, how does the parasite effect the 

 arrest and atrophy of the gonad? It is clear that in the phenomenon of parasitic castration 

 the atrophy of the gonad is brought about not merely by the parasite attacking and devouring 

 this organ: this is proved by those cases of parasitic castration where the parasite never comes 

 into contact with the gonad at all, e. g. the case of Entoniscus, and especially that of the 

 Bopyridae (8) where the parasite is situated in the gill of its host. There are therefore two 

 alternatives: either the parasite introduces into the host some specific substance that attacks 

 the gonad and causes its atrophy, or else the parasite has a general effect on the metabolism 

 of the host which reacts upon the gonad causing it to be absorbed. There can be little 

 doubt that the latter alternative is the true one; for if the parasitic castration is due to the 

 introduction of a specific substance into the host, it is very curious that parasites of such 

 totally different natures as Cirripedes, Isopods, Trematodes and Gregarines (9 and 23) should 

 all exercise the same influence on their various hosts: but it is intelligible on the second hypo- 

 thesis that all these parasites should influence the metabolism of their hosts in much the same 

 way, and that the altered condition of the metabolism should react upon the gonad. 



To this point we will return after considering another question, namely, in what rela- 

 tion do the secondary and primary sexual characters stand to one another? It is generally 

 assumed from the results of ordinary castration and from observations upon diseased gonads, 

 that the sexual gland itself either by means of an internal secretion or else by some other 

 means, causes the development of the secondary sexual characters proper to it. 



Herbst (16) in his Formative Reize has given an excellent summary and discussion 

 of the available evidence, and he concludes that the presence of a differentiated gonad is not 

 necessary for the first stages in the development of the secondary sexual characters, but that 

 it is necessary for their full development. Herbst clearly brings out the contradictory nature 

 of the evidence on the subject: and we may add a further case which has recently been pub- 

 lished and seems to controvert the idea that there is any necessary connection between the 

 development of the secondary and the primary sexual characters at all. Kellogg (18) effected 

 the complete destruction of the gonad in the larva of the silkworm before any differentiation 

 of the gland had set in, and on rearing these larvae to the imago state found that they 

 developed perfect secondary sexual characters, despite the entire absence of a gonad. 



