SPATANGINA. 99 



of their unequal value but I think further knowledge will increase rather than 

 decrease the number. 



The weight given to the various characters is the point at which the "per- 

 sonal equation" enters most largely into the composition of the following key 

 to families. I consider the fundamental make-up of the test as the matter of 

 first importance, and except for convenience, should place the meridoplacous 

 or amphiplacous character of the interambulacra, as the primary feature in 

 dividing the genera into groups. For convenience at least, it is better to first 

 remove the genera in which interambulacrum 5 is not differentiated at all to 

 form a sternum. Next the position and covering of the peristome should be 

 considered, and the development of a floscelle or of a distinct labium on its 

 posterior margin. The modifications of the ambulacra to form petals and 

 their accompanying depression seems of next importance but here the possi- 

 bility that absence of any depression and incompleteness of the petaloid form 

 may not necessarily be the primitive condition in a given group is puzzling. 

 We have no proof to show that it is not a derived condition, a reversion, in 

 some genera, as I strongly suspect it is. Next there are the fascioles, and here 

 again, until more data on growth changes are available, we are quite in the 

 dark as to what are primitive and what are specialized conditions. I feel inclined 

 to think that the subanal fasciole is a very old one, perhaps the oldest, and 

 its presence or absence is therefore of more significance than the presence or 

 absence of any other. Next I rank the peripetalous fasciole but I am in doubt 

 as to whether this precedes the marginal or not. We know the history of the 

 lateroanal, and there is no question that the presence of an internal fasciole, 

 or of anal branches to the subanal fasciole, is a highly specialized feature. The 

 real difficulty then as regards the fascioles is the original relation between margi- 

 nal, peripetalous, and subanal. 



In the use of the following key, it will soon appear that no one character 

 can be relied on for separating the amphisternous families. It ultimately 

 becomes a question of judgment, one might almost say opinion, as to whether 

 ambulacra are petaloid or not, whether they are depressed or not, and to which 

 of these two features, the more weight should be given. The recognition of 

 the Palaeopneustidae results in the arbitrary separation of Palaeopneustes from 

 Linopneustes, and of Homolampas from Lovenia and in both cases violence 

 is done to the truth. Yet the Palaeopneustidae is a natural family and we 

 shall some day understand its history well enough to define it without being 

 guilty of the inconsistencies present in the classification here used. If the 



