28 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I42 



mentum or the submentum, according to the number of subdivisions in 

 the labium. The original line of confluence betvv^een the gula and the 

 labium must be between the tentorial pits, as shown by the silphid 

 larva (A). The gula is usually continuous proximally with the post- 

 occiput, since actually it is merely a ventral sclerotic union of the 

 lower ends of the postocciput. The general tendency of the postgenae 

 to come together ventrally on the prognathous head now in some 

 cases reduces the gula to a narrow median strip between the post- 

 genal margins (E, Gu), and may proceed so far as to eliminate the 

 gula. The line of union between the postgenae is commonly termed 

 the "gular suture," though really it is a postgenal suture. DuPorte 

 in a recent paper (i960) gives a good comparative account of the gula. 

 A most unusual gular condition is present in the head of a soldier 

 termite (fig. 18 A). The long gula is here limited by lateral grooves 

 {pt) continuous from the postoccipital sulcus (pos), which super- 

 ficially appear to be parts of the latter as in other insects. In the 

 termite, however, these grooves are the greatly drawn-out tentorial 

 pits, from which is inflected internally the long, tentlike tentorial 

 bridge iC,TB). 



V. THE ANTENNAE 



The antennae are segmented appendages of the head characteristic 

 of the trilobites and of all the mandibulate arthropods except the 

 Protura, but they are absent in the chelicerates. They are freely 

 movable by basal muscles arising in the head, and ordinarily have 

 only a sensory function, though in the nauplius larvae of Crustacea 

 they serve temporarily for swimming, and in the barnacles for at- 

 tachment. The antennae are always of postocular origin in the em- 

 bryo, and receive their innervation from the second, or deutocere- 

 bral, brain centers. Being sensory organs, however, principally tactile 

 and olfactory, they commonly assume a facial position in postem- 

 bryonic stages, where they more effectively serve as feelers or as 

 odor receptors. 



An antennal segment, as a leg segment, must be defined as a sec- 

 tion of the appendage individually musculated by muscles inserted on 

 its base, arising in the segment proximal to it, except that the muscles 

 of the basal segment arise in the head. Segments, however, are often 

 divided into nonmusculated subsegments, which are thus not to be 

 confused with true segments, though they are usually counted as 

 such in enumerating the parts of an appendage. 



Among the hexapods the antennae are of two types of structure. 



