236 PRINCIPLES OF HYDROID ZOOGRAPHY. 



understood by regarding it as the product of two factors — one of them finding its expression in 

 the trophosome and the other in the gonosome ; and whether the gonosome remains permanently 

 attached to the trophosome, or becomes in whole or in part free, attaining thereby an independent 

 existence, it is equally necessary that it should take its place in our diagnoses of families, 

 genera, and species. 



We accordingly now find that our leading descriptive treatises recognise the necessity of 

 combining in their diagnoses, not only those parts which are destined for the nutrition of the 

 colony, but those also which are destined for the sexual perpetuation of the species, whether 

 these last- mentioned parts be in the form of fixed sacs or iu that of free locomotive planoblasts.' 



If all trophosomcs which presented no difference sufficiently great to justify generic 

 separation had their gonosomes just as closely resembling one another, while, on the other hand, 

 all closely resembling gonosomes had closely resembling trophosomes, the classification of the 

 Hydroida would be freed from one of its great difficulties. 



This, however, is far from being the case, and the study of the Hvdroida renders us 

 acquainted with two sets of phenomena, which signally break down the uniformity assumed 

 in the above suppositions. These are — 1. The association of similar trophosomes icith dissiinilar 

 (gonosomes; and 2, the association of dissimilar trophosomes with similar gonosomes. The difficulty 

 which these phenomena throw in the way of a natm-al classification of the Hydroida may be 

 compared to that which the mineralogist meets with when he finds isomorphism and dimorphism 

 interfering with the uniformity of his mineralogical system. 



But the great ditficulty, after all, in the application of the method here advocated is found in 

 the fact that the planoblast at the time of its liberation is still in an immature state, and may be 

 destined to undergo important changes before arriving at its adult condition. In such cases, 

 unless we have succeeded in following it to its ultimate form, our determination of its type must 

 be regarded as only approximate. Analogy, however, will greatly aid us in this determination, 

 by pointing out what are the parts most liable to change, and what the direction in which this 

 change is likely to take place. 



From these considerations we learn that the number and form of marginal and oral tentacles 

 in the recently liberated medusa must be accepted with great caution, as aftbrding valid systematic 

 characters, these organs being especially liable to an increase in number, and often to an alteration 

 in form as the medusa advances towards maturity, while the form of the umbrella is also subject 

 to a certain amount of change. 



In some cases, however, especially as regards the number of the marginal tentacles, we may 

 fairly assume the condition presented by the young medusa as representing its permanent 

 characters, as, for example, in the case of the single long tentacle in the planoblast of Corymorpha, 

 where we find, by going back to the early stages of the development of this planoblast, that the 

 peculiar asymmetrical form which in a later stage finds its expression so decidedly in the great 

 development of a single tentacle is quite apparent before any trace of a tentacle can be detected. 



^ This principle has been recognised in the following publications : 



Agassiz, ' Coiitr. Nat. Hist.' U. S. ; M'Crady, ' Gymnopbtbalmata of Cbarleston Harbour ;' 

 Victor Carus in ' Carus u. Gerstaecker Handbucb der Zoologie,' vol. ii ; Allman, " On the Construction 

 and Limitation of Genera among the Hydroida,'' in ' An. Kat. Hist.' for May, 18Gi ; Alex. Agassiz, 

 ' lUiistr. Catal. of North American Acalcpbie;' Hincks, 'Brit. Hydr. Zoophytes.' 



