THE ANTENNAE 57 



buds are much more precocious, and that the eyes are charged with pigment at 

 a period when the insect still preserves its larval form. 



"... I believe that this mode of formation of the head occurs in all Hyme- 

 noptera with apodous lame, in this sense ; that a more or less considerable part 

 of the first thoracic segment is always soldered to the head of the larva to con- 

 stitute the head of the perfect insect. The arrangement of the nervous system 

 is naturally in accord with this peculiarity of development, and the cephalic 

 ganglia of the larva to which the ocular blastems later adapt themselves, are 

 found not in the head, but in the succeeding segment (Figs. 39, 40, 41). 



"Relying on these facts, I maintain that the encroachment of the head on 

 the prothorax is a consequence of the preponderance in size of the brain, and 

 indicates the superiority of the Hymenoptera over other insects. ..." 



That the pronotum is derived from the larval prothoracic segment is proved 

 by the fact that the first pair of stigmata becomes what authors call the "pro- 

 thoracic" stigmata of the perfect insect. But Bugnion thinks that the projec- 

 tion which carries it, and which he calls the shoulder (Figs. 41 and 42), belongs 

 to the mesonotum. 



b. Appendages of the head 



The antennae. These are organs of tactile sense, but also bear 

 olfactory, and in some cases auditory organs ; they are usually in- 

 serted between or in front of the eyes, and moved by two small 

 muscles at the base, within the head. In the more generalized 

 insects the antenme are simple, many-jointed appendages, the joints 

 being equal in size and shape. The antennae articulate with the 

 head by a ball and socket joint, the part on which it moves being 

 called the tornlns (Fig. 32, ?). In the more specialized forms it is 

 divided into the scape, the pedicel, and a fwjcUum (or clavola) ; but 

 usually, as in ants, wasps, and bees, there are two parts, the basal 

 three-jointed one being the scape, and the distal one, the usually long 

 filiform flagellum. The antennae, especially the flagellum, vary 

 greatly in form in insects of different families and orders, this varia- 

 tion being the result of adaptation to their peculiar surroundings and 

 habits. The number of antennal joints may be one (Articerus, a 

 clavigerid beetle), or two in Paussus and in Art ran es cn'cus (Fig. 43 1 -), 

 where they are short and club-shaped ; in flies (Museidse, etc.), they 

 are very short and with few joints, and when at rest lying in a cavity 

 adapted for their reception. In the lamellicorn beetles the flagellum 

 is divided into several leaves, and this condition may be approached 

 in the serrate or flabellicorn antennae of other beetles. In Lepi- 

 doptera, and in certain saw-flies and beetles, they are either pecti- 

 nate or bipectinate, being in one case at least, that of the Australian 

 Hepialid (Abantfartes argentens), tripectinate (Fig. 44), and in the 

 dipterous (Tachinid) genus Talarocera the third joint is bipectinate 

 (Fig. 45). In Xenos and in Parnus they may be deeply forked, 



