ANATOMY AND CLASSIFICATION OF THE ARENICOLID^.. 529 



Grubii so long that even Ehlers himself doubted it being a 

 distinct form. Yet a moment's glance at the internal anatomy 

 would show that it is by far the most singular species of the 

 genus, in possessing voluminous complex ovaries or testes 

 and a larger number of nephridia than any other member 

 of the family. The confusion between A. Olaparedii and 

 A. marina is another case due largely to the exclusive value 

 set upon external features in discriminating species. 



The second class of error, that of regarding young forms of a 

 given species as distinct from it, is due not only to the lack 

 of comparison between the internal organs of young and old 

 stages, but also to the lack of perception of the important fact, 

 that even such features as setaa and gills, are not precisely 

 similar in form in the post-larva as in the adult. In fact, 

 there are two kinds of variation which systematists have to 

 reckon with, — developmental variation in such organs as setae, 

 of which several different sets appear in an individual's life- 

 time ; and variation in the homologous organs of the indi- 

 viduals of a species. If it can be shown that the first kind 

 of variation in the case of any organ, e. g. setae, is as great 

 as, and similar to, the variation of the second kind, then 

 such an organ ought not to be employed for accurate specific 

 diagnosis. By disregarding this principle, Mesnil (1897) 

 recently published a paper which contained an account of 

 some form intermediate between the ArenicolidsG and Mal- 

 danidse. These intermediate forms were referred to the 

 genus Clymenides of Claparede, on the ground of external 

 characters, sette and the absence of gills. Further, Mesnil 

 endeavoured to pi'ove that the post-larval stages of Areni- 

 cola marina, described fully by Benham (1893) from speci- 

 mens taken at Plymouth, were in reality also members of 

 this genus Clymenides, which he regarded as combining 

 characters found in Arenicolidse and Maldanidse, and as 

 forming a parallel series to the species of Arenicola. Soon 

 after the publication of this paper, however, another appeared 

 by the same author, in which he showed that his so-called 

 species of Clymenides were partly species of Arenicola 



