Nearly 200 varieties of wheat were subjected in the fall of 1930 at Corvallis, Oregon, to the usual inoculation trials to determine their resistance or susceptibility to wheat bunt, Tilletia tritici (Bjerk.) Winter and Tilletia levis Kuhn. The inoculum used on this seed, however, consisted of a mixture of equal parts of the 10 physiologic forms of bunt reported by the writer. The results were so striking and positive that they are presented at this time.

1. It appeared unnecessary with spring wheat to correct the rate of seeding in rod-row trials on the basis of differences in average kernel weight. 2. At University Farm, St. Paul, Minn., the hand method of sowing rod rows appeared to be slightly superior to the drill method, although the differences obtained were very small and perhaps due only to chance. The drill method of seeding was considered satisfactory and is being used at Minnesota. 3. The trials made in 1926 and 1927 at University Farm, St. Paul., indicate that a heavy rate of seeding in rod-row trials was more satisfactory than a lighter rate. The comparison of yields in rod-row and in 1/40-acre plats was conducted in only a single season and with a group of highly selected varieties. The results obtained gave further evidence in favor of the heavier rate of seeding.

A cross studied at the Utah Station several years ago between the wheat varieties Sevier and Federation indicated two factors for awns with a suggestion of linkage and a peculiar spike density inheritance. To test these segregations the cross was repeated. Spike density behavior was again found to be similarly peculiar in that wide transgressive segregation was found, some progenies having spikes more dense than either parent and other progenies with spikes more lax than either parent. Neither parent is a T. compactum and yet about 25% of the progenies are distinctly compactum. The parental spike densities were recovered in only a few progenies, fewer for the more dense parent, Sevier, than for the more lax parent, Federation. In awn behavior there were found four true-breeding awn classes, one like each parent and two intermediates. The suggestion of linkage was again found as it had been in the original cross studied in 1925, which data were corroborated almost exactly by the repeated cross. One parent Federation was awnless and the other, Sevier, was fully awned. Two intermediate awn classes, designated as awns 2 and 3, were obtained. According to the early study, these were thought to represent different intermediate types each genetically one factor removed from either parent, Federation lacking both and Sevier containing both factors. If these conclusions are correct, a cross of the two intermediate forms should re-assemble the factors and restore the parents. When this cross was made and the F2 plants and F3 progenies studied, the grandparental types (awnless and fully awned) were restored along with the intermediates. The only difference noted in crossing awnless X fully awned and in crossing the two intermediates was that there was no suggestion of linkage in the second case. It was clear-cut that awn classes 2 and 3 were re-assembled into awn class I (awnless) and awn class 4 (fully awned). A back cross of the awnless parent (Federation) on each of the intermediate awn classes 2 and 3 should reveal if there was really a one-factor difference. When the two backcrosses were studied, there was found to be a single-factor difference between the awnless parent and each of the intermediates, as is shown by the fact that P = .37 for one backcross and P = .46 for the other. It was also clear that awn class 2 was different from awn class 3 as only progenies true-breeding for awn class I and 2 were found in the backcross of Federation on awn class 2. On the other hand, progenies true-breeding for only awn classes I and 3 were found in the backcross of Federation on awn class 3. These two backcrosses establish the correctness of the original awn classes set up in 1925 when the original cross was studied. Fully true-breeding progenies in all four awn classes also bear out the original classification. Incidentally, the data presented in these crosses and backcrosses furnish some presumptive evidence that it is the awnless wheat which lacks the factors, i.e., is recessive. However, in a strictly intermediate dominance, it is really difficult, and not highly essential to breeding work, to prove whether the awnless condition is recessive or dominant. One of the two backcrosses permitted a test to see if there was a one-factor difference between the spike density of the lax parent and a dense-spiked segregate. There was found to be a one-factor difference, and as X2 = .39, P is extremely high.