This investigation was mainly directed at the solution of the problem of the multiplicity of eye-colour genes inDrosophila melanogaster. 1. For the purposes of routine quantitative comparison of the red and brown eye pigments of different mutant strains with those of the wild-type, methods are described for the rearing of normal-sized flies and for the extraction of the two pigments and their spectrophotometric analysis. The light-absorption curves are given of these pigments in the wild-type and various mutants, singly and in combination. The two pigments typical of the wild-type are found in all mutants with the exception of the alleles of the white locus which condition the production of a qualitatively changed red pigment. 2. Quantitative estimations of the pigments in the wild-type and the different mutants studied indicate the following effects of mutant genes:wm4 has decreased amounts of both pigments, with the red pigment content varying greatly with temperature changes;ras2 has a decreased amount of red pigment; the genes of the ruby group,rb, cm, g3 andcar, effect a reduction in the content of both pigments, but there is no simple relation between the amounts of this reduction, the four genes showing differential effects on the two pigments; the alleles ofw reduce the amount of red pigment to a great extent and the brown pigment content to a varying extent for the different alleles, but again the effects on the two pigments are differential, there being no simple linear quantitative ratio between the amounts in the different alleles;st suppresses the production of the brown pigment but possibly causes an increase in red pigment content;bw suppresses the production of the red pigment (unless a small amount of qualitatively changed red pigment is formed) and also reduces the brown pigment content; combinations between the genes of the ruby group,inter se and withst andbw, show sub-additive interaction effects and an overlapping in the mode of their action; two alleles ofcar show a simple quantitative relation in the amount of red pigment produced. 3. A scheme is presented to show some of the interrelationships which exist between the various genes which affect eye pigmentation. The probable mode of action of the normal allele of white is at the level where a common substrate is differentiated for the formation of specific substrates for the red and brown chromogens, but where also certain by-products are formed for utilization in the protein carrier and granule system. The normal alleles of scarlet and brown then fit into the scheme at a later level, i.e. that of chromophore or chromoprotein formation. The action of the normal allele of raspberry2 seems to be at the level of cellular differentiation. In connexion with the mode of action of the normal alleles of the genes for ruby, carmine, garnet3 and carnation the concept is developed that many eye-colour genes affect eye pigmentation only indirectly, i.e. eye colour is influenced by them epigenetically, consequent on their main function being the directing of enzyme specificities for the breakdown and resynthesis of proteins during metamorphosis; the products of breakdown are utilized in the eye-pigmentary system which in this activity is partly an excretory system.