i897. MORPHOLOGY IN ZOOLOGICAL SCIENCE 29 



probability, reduction in number of a series of homologous organs per- 

 forming the same kind of function, since there is little doubt that the 

 repetition of members is a phenomenon of fundamentally the same 

 nature as vegetative reproduction, and that it is, primarily at any 

 rate, correlated with growth in size. Thus, we can understand why 

 the nauplius larva should lose the long series of post-oral parapodia- 

 like appendages which we have reason to believe the annelid ancestor 

 of arthropods possessed, and should retain only those first three 

 modified ones, the modification of which was probably one of the 

 chief changes by which an annelid was converted into an arthropod. 

 A most interesting confirmation of this view is obtained by comparing 

 the nauplii of the various crustacean groups. The most annelid- 

 like order is that of the phyllopod Branchipoda, and here the nauplius 

 still retains traces of the post-oral segmentation. In the other groups, 

 notably the cirripedes and the ostracods, the nauplius has lost all 

 traces of this segmentation, and is obviously secondarily modified,, 

 since in each case it shows precociously some features of the adult. 



The general result of this way of regarding ontogeny will be that 

 we shall regard no larva as purely secondary in all its features, and 

 that where we compare two animals more or less allied, we shall 

 attach the greatest importance to the ontogeny of that one which has 

 the longest larval history. It is, so far as I know, a rule without any 

 exception that when we compare corresponding stages, what is obscure 

 and difficult to interpret in the embryo, becomes clear and instructive 

 in the larva. The Echinodermata have perhaps the longest larval 

 history in the animal kingdom ; they begin free life as blastulas, and 

 in no group is the process of the formation of the primary germinal 

 layers so diagrammatically clear. 



Finally, we have to consider the greatest difficulty of all those 

 involved in morphological reasoning, namely, that which the recog- 

 nition of the principle of homoplasy brings with it. The proposition 

 that similar organs might arise in different animals under the stress 

 of similar conditions, seemed at first to offer merely a convenient 

 solution of some difficult problems, so long as the organs in question 

 belonged to members of widely different families, such as the tracheae 

 of insects and arachnids for example. It is, however, becoming every 

 day clearer, that such parallel evolution has taken place again and 

 again within narrow circles of affinity. Let any morphologist 

 examine the structure of a large group ; let him by careful comparison 

 of the various species with one another, arrive at an idea of the 

 primitive form from which all were derived, and then let him attempt 

 to indicate the lines along which the different forms have been evolved. 

 He will find himself driven to the conclusion that important structural 

 features on which he would be naturally inclined to found his system 

 of classification, have been developed twice or thrice. Numberless 

 instances of this might be given. Calcareous sponges used to be 

 divided into three well-marked classes, according to the character of 



