The Graspedosomatidas of North America. 69 



and other characters, it seems that the present species is more re- 

 lated to Trichopetalum than to any other genus described in this 

 paper. It is possible that a closer stud}^ will show that it is the 

 type of a new genus. 



In using Dr. Packard's description, the fact that he was com- 

 paring with species of Polj^desmus should not be forgotten, and 

 the " usual prominent lateral ridges" appear to be much larger 

 than those of any American craspedosomatid known to us. Dr. 

 Packard also counted the anal valves as a segment, which is not 

 customary. 



That there is but one row of bristles on each side of the median 

 line, that there are but two olfactory cones, and that these are 

 Jiattened^ are such anomalous characters that we hesitate to credit 

 their actual existence. 



Two years after the original description was published. Dr. 

 Packard printed a note he had received from Dr. Anton Stuxberg, 

 of Gothenburg, Sweden, relative to the " Generic position of Poly- 

 desmus ocellatus.''^ The species is referred to the European genus 

 Craspedosoma on account of the general fact that this genus has 

 eyes and thirty segments. Dr. Stuxberg having had no oppor- 

 tunity of examining specimens. 



A change in the specific name is also proposed, on the follow- 

 ing ground : 



" In consequence of the necessity of ranging 3'our species in 

 another genus, all the members of which are provided with eyes, 

 the specific name ocellatus should be removed and another intro- 

 duced instead of it. I propose the name Craspedosoma pack- 

 ardii.^'' 



Distinctive specific names are of course desirable, but a general 

 application of Dr. Stuxberg's principle would necessitate too 

 many changes to make possible its admission as a rule of nomen- 

 clature, hence the original name is here retained. 



Dr. Stuxberg also implies his belief that Dr. Packard's speci- 

 mens were immature, being credited with but 28 segments (29 if 

 the anal valves are counted), but we are convinced that 28 is the 

 normal number for Trichopetalum. Dr. Stuxberg's statement 

 that immature specimens of Craspedosoma have 29 and 21 seg- 

 ments, does not agree with other European writers, according to 

 whom the stages are 26, 28, 30. 



