DEPARTMENT OF MARINE BIOLOGY. 203 



early in the season (before adult butterflies of this species had appeared in 

 New Hampshire) with wild males from a more southern locality (Annandale- 

 on-Hudson, New York). 



The four families of grass-green larvae produced by back-crossing with wild 

 stock came to maturity in June. By good fortune, the largest of these families 

 contained heterozygotes for blue-green, for, when inbred in July-August, 

 3 out of the 9 matings gave 25 per cent of blue-green larvae. In addition to 

 the blue-green, a new recessive mutant, olive-green, occurred in two of these 

 three families, one giving a clear-cut 2:1:1 ratio, viz, 95 grass-green : 46 

 olive: 44 blue-green; another 9 : 3 : 4, i. e., 40 grass-green: 17 olive: 29 blue- 

 green. The third mating producing blue-green gave simply 31 grass-green, 

 10 blue-green, with 1 olive, probably a waif from another culture. 



Ohve-green mutant caterpillars also appeared, in company with normal 

 grass-green and in the simple 1 : 3 ratio, in 4 more of these 9 inbred families. 

 The total numbers in these 4 families were 128 olive, 314 grass-green. 



Ohve-green caterpillars have grass-green hemolymph, in which I am unable 

 to detect any color difference from the normal. The color is hypodermal 

 rather than cuticular, and is probably due to a physico-chemical change in 

 the hemolymph, for the scale-pigments of the under side of the hind wing and 

 of the tip of the fore wing, parts most exposed to the action of the hemoljonph 

 in the pupal wing-bud, take on an orange-yellow hue ("lemon chrome," 21, 

 0-YY Ridgway, in the male; "empire yellow," in the yellow female; buff, 

 or "m.ustard yellow," in the white female). 



Moreover, the eye of the adult butterfly from an olive-green caterpillar is 

 also olive-green, a color corresponding to that of the caterpillar, but a shade 

 lighter. 



The general conclusions are as follows : 



(1) The (hypodermal) eye-color of the normal adult Colias philodice and 

 of each of these two mutants corresponds in each case to the (hypodermal) 

 color of its caterpillar. In both the normal and in the blue-green this color 

 is identical with that of the underlying hemolymph. 



(2) The hemolymph supplies both the compound eye and the larval hypo- 

 dermis with the normal grass-green, or with the blue-green, pigment. 



(3) The pigments of the hemolymph are derived from chlorophyl of 

 the food-plant, and a yellow element, probably derived from xanthophyl, is 

 suppressed in the hemolymph of the blue-green caterpillar (i. e., blue-green- 

 blooded butterfly). 



(4) Absence of yellow pigment in the blue-green hemolymph changes the 

 silk spun by the parasitic larva Uving in it from yellow to pure white, and 

 similarly yellow-free blue-green hemolymph produces a blue-green hypo- 

 dermis both in the skin of the larva and in the compound eye, though it is 

 probable that — 



(5) the primary reaction of the hereditary factor for blue-green acts 

 from the nuclei of the intestinal epithelium directly upon dissolving chloro- 

 phyl during digestion, destroying xanthophyl. 



(6) That the hemol>^nph of the oUve caterpillar, though apparently nor- 

 mal in color, directly determines the hypodermal color of the caterpillar and 

 compound eye is substantiated by the orange effect produced upon the wing- 

 scales on those portions of the under surfaces of the developing wings that are 

 most exposed during development to the physico-chemical action of the hemo- 

 lymph. 



(7) The immediate physical basis of these and other hereditary characters 

 in Colias is not chromosomes but substances dissolved in the hemolymph. 

 Upon this substratum, genes, probably chromosomes or parts of chromo- 



