THE GENERAL SHAPES OF ANIMALS. 207 



rooted Hydrozoa, the Polyzoa, and the Ascidioida, the united 

 individuals are not reduced to the condition of segments of a 

 composite individual, and do not display any marked differ- 

 entiations; yet there are some animals in which such 

 subordinations, and consequent heterogeneities, occur. The 

 oceanic Hydrozoa form one group of them ; and we have seen 

 reason to conclude that the Annulosa form another group. 

 It is not worth while, however, to occupy space in detailiug 

 these unlikenesses of homologous segments, and seeking 

 specific explanations of them. Among the oceanic Hydrozoa 

 they are extremely varied; and the habits and derivations of 

 these creatures are so little known, that there are no ade- 

 quate data for interpreting the forms of the parts in terms 

 of their relations to the environment. Conversely, among 

 the Annulosa those differentiations of the homologous seg- 

 ments which accompany their progressing integration, have 

 so much in common, and have general causes which are so 

 obvious, that it is needless to deal with them at any length. 

 They are all explicable as due to the exposure of different 

 parts of the chain of segments to different sets of actions and 

 reactions : the most general contrast being that between the 

 anterior segments and the posterior segments, answering to 

 the most general contrast of conditions to which annulose 

 animals subject their segments; and the more special con- 

 trasts answering to the contrasts of conditions entailed by 

 their more special habits. 



Were an exhaustive treatment of the subject practicable, 

 there should here, also, come a chapter devoted to the in- 

 ternal structures of animals — meaning, more especially, the 

 shapes and arrangements of the viscera. The relations 

 between forms and forces among these inclosed parts are, 

 however, mostly too obscure to allow of interpretation. 

 Protected as the viscera are in great measure from the inci- 

 dence of external forces, we are not likely to find much 

 correspondence between their distribution and the distribu- 

 tion of external forces. In this case the influences, partly 



