ANNELID GENUS CAMBARINCOLA — HOFFMAN 293 



The basic similarities in structui'e, sliape, arrangement, and his- 

 tology of the reproductive systems, however, certainly represent the 

 best reflection of affinities. That this statement is true within the 

 confines of genera is strongly supported by the findings outlined in the 

 systematic treatment which follows. The thin, and usually sub- 

 jectively di'awn, line between species groups and genera permits, I 

 think, an extension of the principle into so-called higher categories. 



From the preceding discussions of characters, it will be recalled 

 that within the limits of Cambarincola one finds a fair amount of 

 variation within the reproductive system both in the gross and histo- 

 logical appearance of the individual organs. The casual observer 

 might suspect, perhaps with justification, that by placing major 

 emphasis upon different organs one could arrive at several entu'ely 

 different classifications. This is not only theoretically true, it is a 

 difficulty which has been a source of vexation since this investigation 

 was begun. The solution has been an arbitrary one, influenced in 

 no small measm-e by the more or less unconscious accumulation of 

 small impressions which collectively result in a conscious allocation 

 of species by the totality of their characters. After several abortive 

 classifications had been drawn up and found wanting, I struck upon 

 the one which immediately appeared satisfactory and this, which has 

 met the test of having to accommodate additional and unforeseen 

 species, is the one here used. The organization of the bursa is given 

 pre-eminence, within limits, in the definition of the genus. 



In the preceding section I have discussed something of the vari- 

 ability of this structm'e among various branchiobdellids, and pointed 

 out that a surprising amount of diversification is to be found in such 

 a basically simple arrangement. There is now known to be a number 

 of species which share the fundamental "cambarincoloid" organization 

 of bursa, ejaculatory duct, spermiducal gland, prostate gland, two 

 deferent ducts, and four efferent ducts and funnels, in addition to 

 a terete body form and a generally similar appearance. However, 

 a detailed study of the bursa shows that on the basis of its several 

 modifications, these species can be classified into groups, in which 

 the component species are obviously quite similar and related in small 

 details as well as overall facies. That bursa structure is a character 

 of major importance is attested by the homogeneity of these groups, 

 which are certainly entitled to be called genera by any definition 

 but the most inclusive. It is now altogether likely that "Camba- 

 rincola" in the usage of Ellis will be found to correspond roughly to 

 the bounds of a subfamily[!] in terms of modern classification. 



We have, then, to consider basically a number of species of North 

 American branchiobdellids in which the body is cjdindrical, the 

 nephridiopore single, the spermatheca not divided or branched, 



