Florida Agricultural Experiment Station would have 50 percent of its frequency in the first class and the remainder tapering off in higher classes. At a somewhat higher position with the mean in the third or fourth class a bimodal distribution of rest period would be produced with one mode in the first class, and the other mode coinciding with the mean seed condition. Any sample whose range was entirely above the germination threshold would produce an undistorted and probably near normal distribution of rest period. On this basis, the classical picture of quantitative inheritance is shown in the frequency distributions of Tables 1 to 6 with approxi- mately the left one-half compressed into the first class which is the class of zero rest period. The right one-half of the picture is shown undistorted in the higher classes. TABLE 6.-SUMMARY OF MEANS OF DAYS FROM PLANTING TO EMERGENCE IN REST PERIOD TESTS ON LESS DORMANT AND MORE DORMANT, PURE STRAINS OF PEANUTS AND THEIR HYBRIDS. Less More Mean Cross dormant dormant of Fi F Fa F- S parent parent parents I 3X* x 221 25 (112-210)** (68-118) 110 7 x 21 19 112 65 45 3 x 21 24 112 68 73 7 x 14 19 154 87 37 1 x 14 10 154 82 42 47 13 x 14 50 154 102 63 3 x 8 24 210 117 79 13 x 8 50 210 130 84 *Two sister plants from strain No. 3 were used as parents for the ten Fi seeds and were labeled 3X. Origins of the different pedigrees are given on page 6. **No test was obtained on strain No. 221 due to constantly poor quality of its seeds but it is undoubtedly a more dormant type. This explanation agrees very well with the general aspects of rest period behavior as has been indicated. Its most signifi- cant and most doubtful feature is the extension of the range of seed condition approximately as far below as above germina- tion threshold and zero rest period. Rest period is a multigenic character as evidenced by a distribution of mean values over all portions of a wide range. The present theory requires also that the range of genotype for rest period extend far below