INTRODUCTION. Ixiii 



larity in the polypites, originate mcdusiform gonozooids 

 \vhicli are not merely alike, but identical at the time of 

 detachment*. In these examples the trophosomes are 

 dissimilar, whilst the gonosomes agree. But the cases are 

 much more numerous in Avhich the alimentary zooids 

 exhibit the closest relationship, while the gonozooids pre- 

 sent differences that would be commonly accounted generic. 

 Amongst the Corynidce, three genera (Coryne, Sijncoryne, 

 and Zancka] are undistinguishable one from the other, so 

 far as the trophosome is concerned. In the large and 

 beautiful family Campanulariidae, all the generic groups, 

 with a single exception, are founded on characters sup- 

 plied by the gonosome alone; and many similar cases 

 might be cited. 



Perfect agreement in the alimentary characters does not 

 of necessity imply agreement in the sexual characters ; 

 whilst, on the other hand, the trophosomes may be 

 strikingly unlike, and the gonozooids identical. 



Our system of classification must be harmonized with 

 these perplexing facts ; and it is hardly a paradox to say 

 that in some respects it may appear less natural if strictly 

 conformed to the order of nature. 



It may be remarked, in passing, that in some genera 

 the differences between the species are chiefly exhibited 

 in the trophosome, and the gonozooids are almost, or alto- 

 gether, identical. This is remarkably the case in the 



* A. Agassiz (in bis ' Cat. of North American Acalcphae, 1865) expresses 

 his belief that at a more advanced stage these gonozooids would exhibit diffe- 

 rencesand will not allow that " medusae generically identical " are " deve- 

 loped from Hydroids generically distinct." Since the publication of his 

 work, however, observations have placed it beyond doubt that many cases 

 occur in which the gonosomes are identical, while the trophosomes present 

 differences that must be accounted generic. 



