THE ANCESTRY OF BUTTERFLIES. 237 



thirteen segments, nearly uniform in size, the first of them bearing a 

 chitinous dorsal shield, the first three a pair of horny legs, the sixth to the 

 ninth and the last segments a jiair of fleshy legs, spiracles upon all the 

 segments except the last two, l)ut those of the se(!ond and third in a very 

 rudimentary condition, so as to be physiologically useless. The surface of 

 the body was covered profusely with little papillae, from each of which 

 sprung a minute simple hair. The new-born caterpillar, however, dif- 

 fered perhaps in this respect from the full-grown, in that its body was 

 furnished with short club-shaped hairs arranged in dorsal, laterodorsal, 

 and substigmatal series, their being two appendages to a segment in the 

 lower series, and one to a segment in the others. In short, a form of 

 hypermetamorphosis was already established in a simple condition. 



Soo, too, metamorphosis was complete, and the chrysalis a perfect 

 mummy with ensheathed limbs, its contours smooth, the head rounded, 

 the ventral outline of the abdomen as curved as the dorsal, and the tail 

 somewhat pointed ; it was protected by a slight cocoon, and also secured 

 within, in a definite position, by attachment at the tail and a girt around 

 the middle. 



The butterfly was heavy-bodied and covered with scales and hairs. 

 The head was broad, the antennae did not encroach upon the eyes and were 

 moderately long and clubbed just before the tip ; the club was about three 

 times longer than broad, curved at the base, and tapering but not 

 hooked at the tip ; the tongue was about as long as the body, wath simple 

 inconspicuous papillae near the tip ; the labial palpi were bushy and rather 

 long, the apical joint smaller than the others, and directed forward. 



The front wings were somewhat triangular, twice as long as broad, 

 with rather a pointed apex ; and the hind wungs rounded triangular, of 

 about equal length and breadth. In the neuration of the former the costal 

 and subcostal veins, w'ith the upper branches of the latter, were closely 

 crowded toward the front edge, at the middle of which the costal vein 

 terminated ; the subcostal vein ran to the outer border just beloAv the apex 

 of the wing and had four upper branches and one lower branch, none of 

 them forked, the last upper branch striking the front edge just before the 

 tip of the wing ; the median had four equidistant branches, the last 

 branch uniting by a cross vein a little beyond the middle of the wing with 

 the lowest subcostal branch ; the submedian ran to the inner angle and 

 the internal was soon lost in the membrane. In the neuration of the hind 

 wing the subcostal and median veins with their branches occupied the 

 middle and larger part of the wing; each divided into three branches, all, 

 excepting perhaps the first branch of the subcostal vein, originating be- 

 yond the middle of the open cell ; the costal and subcostal veins were con- 

 nate at the base, suddenly divei-ging when they parted, the former just 

 reachins: the costal margin, but not extendino; bevond it: the submedian 

 and internal were simple and of nearly equal length. 



