' 



< hamberlin: geopiiiloidea of the southeastern states. 413 



characteristic structure of the head and prehensorial feet and the 

 conspicuous pits at cephalic edge of the anterior sterna with the corre- 

 sponding processes of the caudal margins are features by which the 

 species may readily be recognized when taken in connection with the 

 shape of the last ventral plate and the nature of the coxopleural pits. 

 Mr. Bollman established G. virginiensis chiefly on the presence of these 

 sternal pits in his type; but he must not have had specimens which he 

 recognized as G. mordax, for the type of this species, which I have 

 recently examined, together with all the other specimens studied, 

 have these pits plainly showing, though in some few they are less con- 

 spicuous and may be overlooked when the sterna are closely articu- 

 lated. Mr. Bollman also suggests a difference in the proportions of 

 the "coxa" of the prehensorial feet. As to mordax, Meinert says: 

 " sternum ... sesqui latius quam longius, coxa fere duplo longius 

 (20: 11)." Mr. Bollman appears uniformly in constructions corre- 

 sponding to this in Meinert's various descriptions to have taken the 

 longius as being in agreement with coxa; but plainly the comparison 

 is between length of sternum and that of coxa, not between length 

 and breadth of the latter. A cotype of G. salemensis which I have 

 examined is also a large specimen of mordax in which the sternal pits 

 are conspicuously developed though they appear to have been over- 

 looked by Mr. Bollman. 



Of the coxopleural pores one is typically somewhat larger than the 

 others and more or less isolated on the more caudal portion of the 

 coxopleurae, the others being cephalad of it. In medium sized indi- 

 viduals the pores are usually scattered over the surface of the coxo- 

 pleurae; but with increasing age there is a distinct tendency for the 

 pores to become crowded along the edge of the ventral plate in mostly 

 two rows and similarly along the dorsal plate, leaving the greater por- 

 tion of the coxopleural surface free from pores excepting for the single 

 isolated one. Often the ventral pores, excepting this single one, 

 become shifted entirely beneath the edge of the ventral plate and the 

 dorsal ones similarly beneath the dorsal plate. At the same time 

 there is a tendency for some pores to close up and disappear in later 

 moults, so that there results a progressive reduction in the number 

 of pores. Brolemann's G. louisianae is clearly based upon an old 

 individual of mordax in which the pores are thus somewhat reduced 

 in number and covered by the ventral plate excepting for the single one. 

 In a few individuals, quite old as judged by the appearance of the head, 

 which have been studied there appears a tendency for the isolated 

 pore to become reduced and then closed. In the form described by 



