38 Proceedings of the Biological Society of Washington. 



funnels instead of a single pair of each is an important difference. M . 

 lacustris more frequently has a single spermatheca, and when two are 

 present they are in different somites; while each of the two known speci- 

 mens of M. asymmetricus has two spermathecae in one somite. The 

 musculature of the sperm reservoir is quite different in the two species, 

 that of M . lacustris being unique, while that of M. asymmetricus is more 

 like that in allied genera. The relative diameters of the reservoir and its 

 lumen are quite different in the two species. In M . lacustris the diameter 

 of the lumen is but one-third of that of the reservoir, midway of its length; 

 while in M . asymmetricus the diameter of the lumen is about two-thirds 

 of that of the reservoir. 



The assumption of Michaelsen (1908:163) that M. asymmetricus is a 

 "reduction form" (meaning that it has been derived from an ancestral 

 form with two pairs of male gonads and of sperm ducts) receives increased 

 support, now that we know a closely related form with two pairs of sperma- 

 ries and of sperm ducts. 



The genus Mesoporodrilus was established for the species M. asym- 

 metricus which alone among lumbriculids had median spermiducal and 

 spermathecal pores, together with asymmetry of the related organs. Later, 

 another genus, Premnodrilus F. Smith (1900), was established for a single 

 North American species, P. palustris. In several characters, notably the 

 spermiducal organs, these two genera are closely allied to the older genus 

 Eclipidrilus Eisen and Michaelsen (1901:149) has seen fit to unite the 

 three species in the last named genus. The discovery of a second species 

 of Mesoporodrilus and also of a second distinct species of Premnodrilus, 

 soon to be described, has led the writer to revive these generic names. 

 Notwithstanding their close relationship to each other and to Eclipidrilus, 

 each of these two genera has at least two perfectly distinct species which 

 are much more closely allied to each other than they are to the species 

 of the other genus. Furthermore, these two genera have been separated 

 long enough to permit the development of a "reduction form" in each. 

 As the oligochaete fauna of other parts of the country becomes better 

 known, it seems highly probable that still other species may be found 

 belonging to one or the other. Whether the two groups are at present 

 given generic or subgeneric rank is a matter of no great consequence. 



Lumbriculus inconstans (F. Smith). 



The collections contain more specimens of this species than of M. la- 

 custris, but the latter would be more likely to be selected for study because 

 of their greater size. Scarcely any of the specimens were sexually mature. 

 But two were noticed that are at a stage showing spermathecae and these 

 illustrate the extreme variability among specimens of Lumbriculus. 

 Neither of them has precisely the same number and arrangement of the 

 various reproductive organs that is known in any other specimen of the 

 genus. One specimen, from those labelled as cotypes of Lumbricus la- 

 custris, has paired atria in 10; spermaries and spermiducal funnels paired 

 in 9 and 10; ovaries and oviducal funnels paired in 11, 12 and 13; and 



