Tlie Graspedosomatidse of North America. 21 



of species have been described from the Baltic amber. They are 

 at present probably among the least numerous as far as individ- 

 uals of each species are concerned. North America is perhaps 

 to be looked upon as the headquarters of the famil}^ on account 

 of the comparatively^ large number of generic types, though the 

 number of new forms yet to be discovered in Europe is probably 

 very great. 



If the Graspedosomatidse jCre the highest, in certain senses at 

 least, and at once the most primitive of recent Diplopoda, it fol- 

 lows that the other families are to be looked upon as comparatively 

 degraded, even if more complex, as in the addition of the repug- 

 natorial apparatus. 



Generic Classification. 



The more recent European writers have seemed inclined to in- 

 clude all the previously described American genera, except 

 Pseudotremia, under the European genera Craspedosoma and 

 Chordeuma. Pseudotremia was kept distinct mainly on account 

 of the misapprehension that it had no dorsal setae. From defer- 

 ence to the usually more thorough methods of the European in- 

 vestigators we began our study with somewhat of an expectation 

 of the probable correctness of their view, which we still think 

 was more or less justified b}'^ the incompleteness of American de- 

 scriptions. We were soon convinced, however, that among the 

 American forms are natural and compact groups of species merit- 

 ing generic recognition. Between the species of such groups the 

 characters of the antennse, eyes, body-segments, gnathochilarium, 

 legs and genitalia agree in indicating close aflanities. Especially 

 valuable, by reason of their constancy and ease of expression, are 

 the characters of the ninth* legs of the males, and differences in 

 these have always been reinforced by those drawn from other 

 parts of the body. That these secondary sexual characters have 

 not had their importance generally recognized in classification 

 is no reason why they should not now be made available, and 



* These have been referred to by Pac>ard and others as the eighth legs, while 

 Latzel considers them a j)art of the genitalia and does not count them. We 

 believe it will be found more convenient to think of the eighth legs as re- 

 placed by the genitalia, and thus keep the numbers of legs and segments uni- 

 form in the sexes. 



