Qualitative Comparison of Induced Mutations. The very high frequency of UV mutations, with the much lowered frequency of chromosomal derangements, suggests that these may include types of mutation not included among the Xray mutants, and may be relatively free from the various sorts of pseudo-mutation which occur under Xray treatment as byproducts of induced chromosomal derangement.

The problem is to find criteria which may be applied to distinguish types of "mutation." Possible criteria available in maize include the following;

(1) Gametophyte viability. Many induced mutations are of lowered viability in the gametophyte, particularly as shown by reduced transmission through male germ cells. Differences in viability among mutants are usually regarded as characteristic of the different mutant alleles, the higher viability of standard alleles being considered, the result of natural selection.

This view is contradicted by results with the known spontaneous mutations in maize. A large number of mutants representing various endosperm genes is available, and in these gametophyte viability and male transmission are regularly normal. This suggests that the low viability of induced mutants may be due to the loss of something more than the dominant allele which is assumed to have mutated.

Transmission of the mutant through pollen, in competition with the normal non-mutant pollen grains, provides a very rigorous test of gametophyte viability, which may be applied to mutations at any locus.

(2) Use of genes which mutate normally to an intermediate allele. Spontaneous mutations of R^r, identified by colorless seeds, are regularly mutations to small r^r, as previously reported. Recent studies have shown that R^r mutates also, and with comparable high frequency, to R^g. It does not mutate spontaneously, or at most does so very rarely, to r^g. This may mean that the effect of R^r on anthocyanin coloration of the aleurone and of the plant is due to two separate but very closely linked genes, but whether this is true or not, the fact provides a convenient method for distinguishing between spontaneous mutations at this locus and the type of pseudo-mutation which could result from haplo-viable deficiencies.

A similar situation may apply at certain other loci. Recent trials show that the gene A^b also mutates spontaneously, with a fairly high frequency, to an intermediate allele. The results of an experiment in which the suspected mutations were identified by loss of aleurone color and all were subsequently checked by progeny tests show the following frequencies:

The a^p mutants, when combined with the appropriate complementary genes, have the red-brown plant color and brown pericarp characteristic of the standard a^p, although some of the mutants show a somewhat deeper color in aleurone and plant than the standard. Nine of these mutants have been tested for dominance of the brown pericarp effect. In all of these the effect is dominant as in the standard a^p.

(3) Reverse mutability. The analysis of the action of Dt by Rhoades makes possible the effective application of this criterion in the case of apparent mutations to a.

It is not applicable to the a^p mutations from A^b, since Dt is without effect on a^p. Whether it is applicable to all mutant a's, or to the colorless mutations from all A's, also remains to be seen, since the present stocks of a, on which Dt is effective, trace to not more than two original sources. Reversibility of a mutant a under the influence of Dt is good evidence against deficiency, but failure of a mutant to be reverted by Dt is not convincing evidence against intragenic mutation.

(4) Detailed analysis of phenotypic effect. In the case of the genes affecting anthocyanin pigmentation, mutant phenotypes may be compared quite precisely by the use of methods developed by Karrer, Robinson, Scott-Moncrieff, and others for the identification of the various anthocyanin pigments. A study of the anthocyanin pigments in maize now being made by J. E. McClary indicates that there is a very rich variety of these pigments in maize, including several which do not commonly occur among the flower pigments genetically studied by the English workers.

One of these is the anthocyanin pigment which occurs together with a flavonol in the a^p stock. In the presence of B and Pl, A^b, like A, produces chrysanthemin, but a^p produces an anthocyanin of distinctly different properties. The dark a^p obtained by mutation from A apparently produces the same pigment in larger quantity.