Tlie Graspedosomatidse of North America. 7 



Order Anocheta. 

 Suborder Spieoboloidea. 

 Family Spirobolid^, — genera Acanthiulus, Rliinocricus, Spirobolellus 

 Spirobolus, Thyroproctus, Trigouiulus. 



Distribution : Tropics of both hemispheres. 



Subclass ARCHIPOLYPODA. 



Family EuphobeeiDjE,— witb numerous genera. 



This group is an assemblage of widely different fossil forms, 

 some of which will probably proA^e to be true Chilognatha; thus 

 the genus Amnil^^spes bears striking resemblance to certain gen- 

 era of Polydesmoidea. The species of Euphoberia and Acan- 

 therpestes, however, are types apparently very distinct from the 

 Chilognatha. The six dorsal setse are located as in Craspedoso- 

 matidae, while the setse and unpaired seminal duct are both pres- 

 ent in Stemmatoiulus, the representative of an order equal in an- 

 tiquity to the Archipolypoda, so that the derivation of the 

 Chilognatha from the Archipol.ypoda is not indicated by present 

 evidence, though a common origin is most probable. 



The orders and families of Chilognatha are in need of careful 

 description which will make the labor of characterizing species 

 and genera less difficult. In all the suborders, except the Poly- 

 desmoidea, the assignment of the genera to the different families 

 is made with some confidence. The classification of the Poh'des- 

 moidea is in a backward state, owing to the fact that the older 

 writers neglected manj^ of the more important characters and the 

 arrangement of some of the genera may prove to be more or less 

 artificial and provisional. 



The character which during our studj^ of the Craspedosoma- 

 tidse we supposed to be unique among recent Diplopoda — the 

 dorsal setse — is shared by Stemmatoiulus, which has even the 

 curious setiferous papillae of the last segment, though there are 

 four in Stemmatoiulus, and onl^' two in Craspedosomatidae. The 

 co-existence in Stemmatoiulus of pores and setae show that the 

 presence or absence of either of these structures is not of itself 

 evidence of close affinit}^ or wide diversity. The supposition 

 that the spines are in some way a substitute for the repugnatorial 

 apparatus is also shown to have no foundation. 



The presence of external seminal ducts in lulidae and their ab- 

 sence in Craspedosomatidae is at once an important and an easily 



