2. Gene Action. Among tissues capable of anthocyanin production there are marked differences in response; cells of certain types produce anthocyanin readily with any R-allele above the R^g level, while cells of other types may produce anthocyanin only in the presence of the strongest alleles. For example, among epidermal cells of the leaf, there are distinctive differences in the reaction of the long, narrow cells over the veins, the long and short surface cells, the stomatal cells, the hairs and the specialized cells at the base of the hairs, and the paired silicaceous and suberized cells. Anthocyanin is formed much more readily in the epidermis than in the underlying mesophyll cells, but in the chlorophyll-lacking sectors of japonica plants it is produced abundantly in mesophyll cells also. The same is true of certain white and virescent types, and in normal green plants the mesophyll cells of the auricle (which lack chlorophyll) are well colored by even relatively weak alleles. With strong B alleles, green mesophyll cells containing anthocyanin are more frequently found.

The alleles of R and B thus provide a series of reagents, so to speak, for the study of tissue differentiation. Thirty years ago Keeble, Atkins, and others showed certain interesting relations between anthocyanin patterns and the occurrence of oxidase systems detectable by the use of histochemical test-substances. Mr. Fogel has undertaken a study of this kind with maize, which is however still in a preliminary stage.

The study of competitive action of certain A alleles (News Letter 1943, page 21) is being continued in collaboration with John R. Laughnan. The dominant action of a^p upon plant color is manifested with all of the visibly weakened A alleles tested (A^w, a^lt, A^br, A^rb). The alleles A^br and A^rb (both obtained by Rhoades, out of a by Dt) are purple plant types distinguished from A by their reduced effect upon pericarp color. When these are compared with A in sib plants (in backcross progenies marked by et), they show slight but distinct reduction in anthocyanin pigmentation of the plant as well.

The dominant effect of a^p upon plant color is shown also, to a slight extent, by certain A's which appear to have full plant color and pericarp color effect. The different A's used were extracted, after parallel backcrossing to a C R, from various stocks, chiefly the Indian strains used as foundation material for the R and B studies. With some A's the difference between A/a^p and A/a sibs is clear enough to permit reasonably accurate prediction of the genotype at the flowering stage, and this identification may be made somewhat more accurately by testing the extracted pigment. The difference is due to the presence of varying quantities of yellow pigment in addition to the purple. With other A's and with A^b no difference is found. The a^p reaction thus serves as a sensitizer for the recognition of differences between the A alleles, and indicates the occurrence of considerable additional allelic variability at this locus. Conversely, the extent of the effect varies among different pale alleles obtained by mutation from Ab (News Letter, 1943, page 21), when these are tested against a common A. All of the pale alleles showing the dominant plant color effect have dominant brown pericarp action; the two pale aleurone alleles with recessive brown pericarp (A^w and A^lt) give negative results in parallel tests.

Mr. Laughnan is making a chemical and spectrographic study of the pigments involved in the action of the A alleles, and is developing methods for the quantitative study of the mixed pigment phenotypes.