41] THE HEAD-CAPSVLE OF COLEOPTERA—STICK.XEY 41 



cipital foramen and the submentum, and the gular sutures extend no 

 farther cephalad, whereas in all the other genera of the Adephaga the meta- 

 tentorinae are close to the submentum and complete gular sutures are 

 present. Finally, the tentorium of Peltodytes is not so well developed, 

 and the supratentoria are small and do not reach the ental surface of the 

 head-wall. In all other genera of the Adephaga the tentorium is large 

 and strongly chitinized in all its parts. The Dytiscidae and Gyrinidae 

 are very similar fundamentally, and differ from other adephagous families 

 in the position on the ventral aspect of such structures as the antennariae 

 and the pretentorinae, and by a rudimentary or absent occipital suture. 

 The instability of the epicranial suture is well illustrated by the case of the 

 two closely related genera, Dytiscus and Cybister, the former possessing 

 complete, distinct arms, and a considerable part of the caudal end of the 

 stem, the latter incomplete and rather indistinct arms and no part of the 

 stem. The Dytiscidae and the Gyrinidae are undoubtedly more special- 

 ized than the first four families of the Adephaga. 



The Hydrophiloidea (Figs. 11-14, 159-166, 306-309, 453-456) show 

 by the form and texture of the heads of Hydrous and Hydrophielus that 

 they are quite similar to those of the Dytiscidae and Gyrinidae, being 

 broad and short, and strongly chitinized. The antennariae and preten- 

 torinae of these two species have migrated ventrad, as in the other two 

 families. The metatentorinae, the gula, and the tentorium are also al- 

 most identical in form and condition. These species are evidently very 

 closely related to the Dytiscidae and Gyrinidae. Hydraena is not so simi- 

 lar, the dorsal surface of its head being more like that of the t\T)ical Carab- 

 idae in shape and position of the parts. The gular sutures are short, 

 as in Peltodytes. The head of Hydroscapha, its shape, the course of the 

 epicranial arms, position of pretentorinae and antennariae, is distinctly 

 like that of Omophron. 



The genera of the Silphoidea (Figs. 15-21, 163-169, 310-316, 457-462) 

 show many inconsistencies. The dorsal and ventral surfaces and the endo- 

 skeleton of no two genera are nearly alike. Necrophorus possesses the 

 most generalized dorsal surface, which is very like that of the first four 

 families of the Adephaga; Clambus, perhaps the most speciahzed, being 

 distinctly consolidated. On the other hand, Necrophorus possesses the 

 most speciahzed ventral surface present in this superfamily. The gular 

 sutures are complete, and for more than half their extent meet on the meson. 

 In contrast, there are practically no gular sutures in Connophron, a scyd- 

 maenid, and iMolamba, a corylophid. No greater extremes of this surface 

 can be found within the limits of any other superfamily. The tentorium 

 exhibits just as great extremes in development. That of Leptinus is quite 

 generalized; that of Necrophorus well developed but specialized, owing to 

 the invagination of the gula; that of Connophron with no corpotentorium 



