REVISION OF BORARIA AND GYALOSTETHUS — HOFFMAN 311 



that of the prozonite or is but slightly so, and there are no lobations 

 between the posterior coxae. Instead, the sternum is produced into 

 acute, often curved, subcoxal spines that include the ventral coxal 

 condyle. This form occurs most characteristically in Nannaria. 



(d) Podosterna: In a few xystodesmid genera (and in genera 

 of other famihes) the metasternum is elevated sharply and promi- 

 nently, the anterior face of this elevation sometimes being perpendicular 

 to the surface of the prozonite. The legs thus appear to be set on the 

 lateral sides ofjventral tumidities. There is no tendency for the forma- 

 tion of subcoxal spines or transverse grooves. This type of sternum 

 occurs in Pachydesmus and Dicellarius in almost indistinguishable 

 form, and this similarity is reinforced by numerous other structural 

 concordances among members of these two genera. 



(e) Excavate: In this type, the sterna become proportionately 

 very broad and the surface between the posterior legs is depressed 

 deeply and conspicuously. There is no formation of subcoxal spines. 

 The posterior edge of the metasternum takes the form of a transverse 

 ridge or rim bounding the depression. I know this sternal form only 

 in Gyalostethus ; presumably it is a rare type of speciaUzation. 



I judge that the sterna are less liable to change than the gonopods, 

 and, therefore, provide useful criteria for the determination of affin- 

 ities as well as a means of identifying female specimens. 



In the three species here referred to Boraria, the sterna are unmodi- 

 fied while those of Gyalostethus monticolens are totally different in 

 formation. This in itself would be sufficient prima facie evidence for 

 generic segregation of G. monticolens. 



Coxal spines. — Previously the presence or absence of distal spines 

 on the basalmost podomere has been considered to be of generic 

 importance. I have found this character to be a mutable one, how- 

 ever; in some genera (Dixioria, Boraria, Bhysodesmus) the coxal spines 

 may be present or absent. There is no trace of them in B. stricta and 

 B. media, but B. deturkiana has prominent spines, and, partly for this 

 reason, I segregated the species into a separate genus some years ago. 

 Apparently this character varies among different groups and may be 

 consistent in some, mutable in others. 



Tergal sculpture. — The species treated here are mostly smooth 

 dorsally, B. stricta is an exception; the metatergites are provided with 

 several transverse rows of distinct, setiferous tubercules that become 

 larger and more prominent on the posterior body segments. Most 

 xystodesmids have transverse rows of microsetae; normally, however, 

 these are set on extremely small tubercules that can be seen only when 

 the tergum is dry and illuminated from a particular angle. The rows 



