Nortli Americait Liintbriciduhe. 471 



part of the intestinal plexus ; and a posterior pair, connecting 

 the dorsal and ventral vessels. In several somites nearest to 

 the posterior end, both pairs of perigastric vessels have coeca 

 and connect the dorsal vessel with the ventral part of the 

 intestinal plexus. 



Thus far, our knowledge of the Liiiiihricnlidte of North 

 America has been limited to species collected in very restricted 

 and widely separated regions. Two species of Sutroa and 

 one of Eclipidrilus from California have been described by 

 Eisen ('81, '88, '92, and '95); one species each of Mcsoporo- 

 drilKn and Thinodrilus from Illinois, by the writer ('95 and 

 '96); and this paper contains the description of a species 

 from Florida, for which still another genus name is proposed, 

 namely, PremnodriliiH.^ Of these, Thinodrilus is much more 

 nearly allied to Lumhricidus and certain other European forms 

 than to its North American associates that have thus far 

 become known, while Siitroa seems in certain particulars 

 intermediate between the European genus Bhijiichelmis 

 (Yejdovsky, '76) and the peculiar group of North American 

 Lumbrlcididd- which includes Eclipidrilus, McsoporodrilKS, 

 and Preinnodrdns. The species included in these three 

 genera are much more nearly related to each other, so far as 

 the structure of their reproductive organs is concerned, than 

 is any one of them to species of other genera of the family, 

 and yet the differences beeween them seem to the writer too 

 great to be considered as merely specific. Thus, at present, 

 six species of North American Lnmhrindid«' are known, and 

 they have been placed in five different genera. 



Such a condition of things, in which we have in a com- 

 paratively small group of animals a number of genera nearly 

 or quite as great as the number of species, may be due to 

 one or more of several causes^ and in this case it may be 

 owing to the fact that at present our knowledge of the Luvi- 

 hricidid(e is insufficient to make it possible to determine which 



' Leidy's desfri{)tions of species presumably belonging to the l-nmbricnUdiv are 

 inadeciuate and must be disregarded. 



