lo8 John Henry Comstock 



GGG. The Hawk-moths can be recognized by the well 

 known form of their wings, and the presence of what has 

 been termed an intercostal vein (Fig. 24). If I am correct in 

 my interpretation of the homology of this vein, (see p. 78), 

 the family can be characterized as those specialized Frenulum- 

 conservers in which the base of costa of the hind wings is 

 preserved and is remote from the costal border of the wing, 

 and in which subcosta of the hind wings is consolidated 

 with radius for a distance and then separates from radius and 

 joins costa. This arrangement of the veins is quite different 

 from that which exists in Zygcena, (Fig. 19), where also 

 costa of the hind wings is preserved. It more closely re- 

 sembles that of the silk worm, Sericarial (Fig. 32.) 



Family Sphingid^. 



GGGG. I place the Zygsetiids last in this group of isolated 

 families because it is among them that we find the most 

 highly specialized representatives of the frenulum-conserving 

 Lepidoptera, {Cosniosoma, Syntomis, Fig. 11). On the other 

 hand certain genera, Triprocris (Fig. 10) and Pyrojnorpha 

 present a remarkably generalized condition of wing structure. 

 The range of variation is greater than I have observed in any 

 other family or superfamily. In the more specialized forms 

 a greater degree of cephalization of the powers of flight has 

 taken place than occurs elsewhere in the order. And with 

 this cephalization there seems to be correlated a lengthening 

 of the fore wings and a narrowing of the basal part of the 

 area lying between radius and cubitus of these wings. This 

 narrowing of this area appears even in our most generalized 

 forms, in which the discal cell of the fore w-ings can be well 

 described as petiolate. Another characteristic of the Zygae- 

 nina is the extent of the coalescence of the subcosta and 

 radius of the hind wings. A somewhat similar coalescence 

 occurs in certain genera of the Arctiidse ; but it takes place 

 earlier (z. e., in more generalized forms) in the Zyggenina, 

 and is carried farther than in the Arctiidae. The Zygseuids 

 form a superfamily. The relationship of the families compos- 

 ing this superfamily have not been worked out. The Ameri- 

 can genera, so far as they are known to me, differ markedly from 



