NEOTROPICAL PLATYRHACID MILLIPEDS — HOFFMAN 19 



Proaspis Loomis (1941): Uncertain 



Aymaresmus Chamberlin (1941): Valid 



Dynesmus Chamberlin (1941): ?=Tirodesmus 



Ernostyx Chamberlin (1941): == Psammodesmus 

 Of the foregoing names, some have been stabilized recently by the 

 examination of their type species or closely related ones. A few, 

 such as Nanorrhacus, have been based upon fairly well described 

 species. In the three earlier papers (1953a-b, 1956), I suggested an 

 identification for Platyrhacus that seems to be well founded and estab- 

 lished the synonymy of Ernostyx with the older Psammodesmus. It 

 is now possible to consider three additional genera from the list, and, 

 equallyfimportant, to dweU briefly upon the hitherto untouched 

 subject^of generic classiflcation within the family. 



GoNOPOD Morphology and Suprageneric Groupings 



Since systematic groupings in the Diplopoda are based to such a 

 large extent upon the structure of the gonopods, it would seem 

 obligatory that these appendages be studied and illustrated in detail 

 for each new species. Actually, however, some workers have proposed 

 dozens of new forms without any illustrations whatever, or have pro- 

 vided only the most schematic and unsatisfactory drawings. Only 

 within recent years have a few investigators become concerned with 

 careful study of gonopod structure and the establishment of serial 

 homologies throughout various generic groupings. 



Although the American platyrhacids, so far as is known, are struc- 

 turally rather similar in comparison to their more diverse counterparts 

 in the Indo-Australian fauna, it is possible to detect morphological condi- 

 tions in the form of the gonopods that undoubtedly reflect evolutionary 

 divergencies on a suprageneric level. Of the genera treated in this 

 paper, Platyrhacus and Nyssodesmus appear to be closely related 

 and may be considered as belonging to a group in which the sem- 

 inal groove proceeds along the coxal side of the telopodite, whereas 

 the solenomerite originates from the adcoxal surface, the latter being 

 reflected somewhat proximad by a slight torsion of the tibiotarsal 

 region. The groove is thus directed transversely across the face 

 of the telopodite in order to gain entry to the base of the soleno- 

 merite. 



The gonopod in Psammodesmus departs radically from this arrange- 

 ment in that the solenomerite is derived instead from the coxal face 

 of the telopodite and thus in direct line with the normal course of the 

 seminal groove (contrast the gonopods in figs. 1, a and e). 



In the forms known to me there is nothing to suggest which of these 

 two conditions may be the more specialized. It is possible that two 

 primitive terminal processes became independently adapted to carry 

 the seminal groove, and that the distal torsion of the telopodite in 



