408 JOHN H. GEROULD 



upon the presence of light, for many green caterpillars feed and 

 thrive well though enclosed in a box or jar from which light is 

 excluded. This species, however, feeds only by daylight. 



FACTORS AFFECTING THE COLOR OF BLOOD AND COCOON IN THE 



SILKWORM 



Yellow blood in silkworms is closely correlated with the spin- 

 ning of yellow silk, white blood with white silk. Ude ('19), 

 however, has discovered a strain of yellow-blood stock that spins 

 white silk, though their silk glands are yellow. 



A homozygous mutant, i. e., yellow blood, white silk X white 

 blood, white silk, if the latter is a genuine double recessive, gives 

 only the double dominant yellow blood, yellow silk in Fi, which, 

 inbred, produced 9 yellow blood, yellow silk, 3 yellow blood, 

 white silk, 4 white blood, white silk, showing that two pairs of 

 factors interact, viz., C, turning blood yellow, but by itself not 

 affecting silk, and Y, which affects visibly neither blood nor silk 

 unless combined with C, when it makes the silk yellow. More- 

 over, not all individuals of white blood, white silk strains are 

 double jecessives. Some are either homozygous or heterozygous 

 for the dominant yellow silk factor, Y, though they are of course 

 homozygous recessives for the blood color, viz., cc. 



If the yellow blood of the silkworm is due to xanthophyl from 

 mulberry leaves, upon which point Dubois ('09) expresses some 

 doubt, there would seem to be involved in this case two recessive 

 decolorizing enzymes, one of general nature, c, turning the blood 

 white, and the other, y, acting specifically through the nuclei of 

 the silk glands to decolorize their product, the glands themselves, 

 as Ude points out, not being affected. Either of the recessive 

 decolorizers must be present in homozygous or duplex quantity 

 in order to produce a visible result, i.e., CcYy would be yellow 

 blood, yellow silk; Ccyy, yellow blood, white silk; ccYY or ccYy, 

 white blood, white silk. This assumption necessitates no change 

 in Ude's formulas and gives a working hypothesis that may be 

 of value regarding the factors. 



