26 MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 



la the full-fed larva of C. sepidcralis there are no definite stripes, the oblique ones are not 

 present. 



In C. regalis, especialh', the oblique, black stripes are continued by the large conspicuous 

 black si)iracles, and also by the black stripe on the outside of the niid-abdoniinal legs. Thus the 

 effect of the stripes themselves is heightened and e.xtended by these supplementary markings 

 situated in line with but below the oblique stripes. 



In stage IV the oblique stripes are more diffuse, broader, and reach so as to almost include 

 the spiracles, which are as large as in stage V and last, and from each spiracle a black stripe 

 descends obliquely to just behind each infraspiracular spine, and then blends with the black 

 stripe on the mid-abdominal legs, though there is no corresponding stripe on the other legless 

 abdominal segments. It is also to be observed that the blackish oblique stripes are confined to 

 a single segment, not passing on to the next one, and an oblique antero-posterior stripe extend- 

 ing from the front edge of the segment down to the spiracle is the reverse of what obtains in 

 the majority of the Sphingidre-. 



In stage V, C. regalts (in a dried, blown example) is rather less blackish than in stage IV, 

 and under each blackish stripe is a pale shade. The spines are also much smaller, so that the 

 larva is less black; the light shade is a little wider than the black band itself. 



Another set of black markings is the great dorsal black patches on the second and third 

 thoracic segments of stages III to V, and the narrow cross-band between the third thoracic and 

 first abdominal segments; their significance may be cleared up by observations on the living 

 insect. 



VI. DICHROMATISM OR COLOR VARIATION IN THE LARVA. 



The best-known case of dichromatism in larvte is that of Thyn-iiK (ihJ)otii^ described by 

 Riley, in which there is a dark brown and a green form. Lieutenant-Colonel Fawcett describes 

 a dark form of the larva of Protoparce mauritu, of Natal, while AcJierontia atropos is said by 

 Trinien to have a dark form at Cape Town. Light and dark color varieties have been artificially 

 produced by Prof. E. B. Poulton. Mr. Meldola had previously (Proc. Zool. Soc, London, 

 p. 155, 1873) called attention to the fact that the younger larvfe of Geometra papllionaria are 

 brown, "and remain brown during hibernation, when the leaves are bare, while many of them 

 become green when older, after the leaves have expanded in spring." Mr. Poulton also 

 discovered that " the younger larva? possess the power of adjusting the shade of their brown 

 color to that of the twigs of their food plant." Mr. ]Meldola calls this phenomenon "seasonal 

 adaptation," and besides the species mentioned he cites Acidalia degenentria, and Gnophos ohscii- 

 rata, adding, "and many more could be named."" 



In Eacles imj.)er!alis there are two color forms, i. e., a normal green and a brown form. 

 They do not, however, seem to be phytophagic varieties, as both occur on the white pine. 



While the caterpillars in the early stages are usuallj' a light yellowish or claj^ brown, in the 

 fourth and last stages they vary in being either pale green or reddish brown. As these cater- 

 pillars are not common, it is not easy to state the proportion of brown to green individuals. 

 From what I have seen, I should suppose that the green in the two final stages were the normal 

 or more common, and the brown wei'e more of the nature of aberrations. 



Mr. Bridgham has, more commonly, found the pale green form on the white- pine at 

 Providence, though the brown form occurred on the same tree. 



I will describe the color forms in my possession: 



The green form — Stage. I Y. — The bodj' of a blown example is pale green; the hairs very 

 long, white, and rather thick; the head is pale ocherous, paler than in the reddish brown blown 

 specimen, and of the same hue as the thoracic legs. The thoracic horns and other spines are 



"Meldola also states that the larvte of Emnvlc^ia unifaaridta feeil on the seetls of a species of Bartsia when the 

 ca]>sules are in various stages of growtli, "and those caterpillars found on the green capsules were green, whilst tliose 

 on the brown capsules were of a corresi)onding colour." (Trans, of Weismann'a "Studies in the theory of descent," 

 I., p. .S07, 1882.) 



