MEMOIES OF THE NATIO:^rAL ACADEMY OF SCIENCES. 
41 
3. The thoracic dorsal tubercles deep orauge; their homologues on the abdominal segments 
amber yellow. 
4. The tubercles at the end of Stage 11 and in Stage III spotted on the sides with black. 
5. In Stage III the dorsal tubercles of second and third thoracic segments showy coral red. 
The subdorsal andinfraspiracular tubercles tipped with pale bluej in Stage II the same tubercles 
are almost entirely pale blue. 
6. The head becomes green in Stage IV, with a black spot on the side, 
7. The larva is most gaudily colored and conspicuous in the last two stages, while in 8. 
eynihia there are not such marked differences between the different stages, though the last is the 
most variegated, owing to the beautiful turquoise-blue trappings. 
In Callosamia proinethea the freshly hatched caterpillar is most remarkably banded, and all its 
marks and tubercles are in striking contrast with the fully grown larva. The differences may be 
epitomized as follows: 
RECAPITULATION OF THE MORE SALIENT ONTOGENETIC FEATURES OP CALLOSAMIA. 
A. CONGENITAL FEATURES. 
1. Hatched with heavy black transverse bands on a yellow body, and the head black, banded 
with yellow; the bristles moderately long; thus the larva is already a rather conspicuous object. 
2. The dorsal thoracic tubercles already differentiated in size and color from those on abdom- 
inal segments 1 to 7. The differences between the freshly hatched larva and the last stage very 
marked; more so than in Platysamia or Sarnia. 
B. EVOLUTION OF LATER ADAPTATIONAL FEATURES. 
1. In Stage II the body becomes paler, and thus the black bands more conspicuous. The 
second and third thoracic dorsal tubercles and those on abdominal segments 1 to S are now all 
yellowish and of the same size. 
2. Disappearance in Stage III of the tons verse black bands. The abdominal tubercles all 
become blackish, 
3. In Stage IV the head becomes yellow, being less conspicuDusly marked, and the dorsal 
abdominal tubercles are about half as long and large as those on the second and third thoracic 
segments. 
4. The body becomes in the last stage much smoother than Jbefore, the dorsal prothoracic and 
abdominal tubercles being much shorter than in Stage lY. .This reduction of size and inconspicu- 
ousness of the dorsal abdominal tubercles is carried out to excess in 0. angiUifera, where they 
become obsolete, and the larva is simply a large green caterpillar with inconspicuous .markings, 
and simply iirotected by its green color, like the majority of lepidopterous larvie, not being so 
strikingly marked as in the fully fed Sarnia eynihia. 
It is not improbable that the reduction and atropliy of the dorsal tubercles in question is also 
accomiiauied by a great reduction, if not total abolition, of the poison glands at the base of these 
spines. However, having lost the power of resisting or avoiding attack by this means, it, by the 
action of the law of correlation, also loses its bright markings or danger signals, and having 
become harmless to its enemies it is preserved from.extiuction by xiassively relying on its smooth, 
glaucous-green body to escaxie the observation of its natural enemies. 
A tendency to the same end is seen in the larva of Samia eynihia^ which is X)aler, less gaily 
ornamented with bright markings, and also is much less heavily intercalated than the caterpillar 
of Flatysamia cecropia. 
It is evident that of the two species of Callosamia, G. promethea is the more primitive form and 
C. anguUfera a devivation from it; the former is what systematists call a “higher” species and 
C. angiiUfera a “lower,” but many “lower” species are simply a set of those individuals which 
have undergone some degree of modification or degeneration, and are later in point of origin. 
Likewise the Asiatic genus, Samia (S. eynihia being an introduced form), with little doubt, is 
a form which has undergone more or less modification and indeed a slight degree of reduction or 
atroi)hy, and is thus a later form, the genus Platysamia being an earlier type, since it has probably 
been envolved from Saiurnia, which is the most primitive genus of the family. 
