222 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY. 
It will be seen, upon inspection of Figure 29 (Plate 5), that the fold- 
ing of the wing membrane in a direction perpendicular to the trend of 
the nervures is very sharp. The wing membrane is, in fact, thrown 
into a very regular series of closely compressed folds (eta. al., Fig. 29), 
a single scale being inserted upon the crest of each fold. It was, no 
doubt, the very sharpness of this folding which induced Landois (’71) 
to believe that the scales were actually inserted into little tubes. 
When the butterfly issues from the chrysalis, these folds in the pupal 
wings flatten out, and it is this flattening which causes the expansion of 
the wings. Figures 13 and 15 (Plate 3) are diagrammatic representa- 
tions of cross and longitudinal sections, respectively, of the pupal wings ; 
and Figures 14 and 16 are similar sections of the mature wing. It is 
evident that the wings after emergence undergo a great stretching and 
flattening. The mechanics of the operation appears to be as follows. 
The hemolymph, or ‘ blood,” within the wings is under considerable 
pressure, and this pressure would naturally tend to enlarge the freshly 
emerged wing into a balloon-shaped bag; but the hypodermal fibres 
(for. Wdrm.) hold the upper and lower walls of the wing membrane 
closely together, and so, instead of becoming a swollen bag, the wing 
becomes a thin, flat one. And thus it is that the little, thick corru- 
gated sac-like wings of the freshly emerged insect become the large, 
thin flat wings of the imago. In Figure 30 (Plate 5) we see a longi- 
tudinal section through a portion of the mature wing of Callosamia_pro- 
methea, killed about two hours after emergence. The chitinous wing 
membrane is represented by cta. al., and the contracted hypodermal 
fibres, which in the pupa had the form of long tapering cells, by fbr. 
Wdrm. Figures 45 and 46 (Plate 6) give the natural size of the pupal 
and imaginal fore wings, respectively, in Danais plexippus. The area of 
the wing of the imago is 8.6 times that of the pupa. Now, as the wing 
of the young pupa has about 60 times the area of the wing in the 
mature larva, it is evident, that in passing from the larval state to 
maturity the area of the wings increases more than 500 times. 
2. The Development of the Pigment within the Scales. 
In this portion of the paper we shall consider only those changes 
which take place within the scales themselves; these will be traced 
from the first appearance of the scales up to the time when the image 
issues. We shall pay especial attention to those phenomena which 
accompany the formation of the pigment. 
