Synopsis Reduction of wheat populations below 20 plants per square foot through systematic spacing of the individual plants produced progressively lower yields at each level. The adverse effects of reduced stand were reflected primarily in decreased head number per unit area and test weight. Head number per square foot was lowered by thinning to any level, but test weight did not decrease until the population was reduced below 10 plants per square foot.

O NE of the important steps in profitable sorghum production is that of securing stands. Stands may be altered by a number of factors a few of which are due to the peculiar characteristics of the seed. Other hazards to stands are a poor seedbed, low grade seed, planting too early, the use of improper planting plates, and torrential rains which may cover or wash away the young seedlings. Some of these hazards may be more easily controlled if the character of the seed is given consideration. The discrepancy between laboratory and field germination of seed of the better known sorghum varieties, as well as the influence of size of kernel on the number of plants per acre, are discussed in this paper. LABORATORY AND FIELD GERMINATION COMPAREDCertain cultivated crops often show a much wider discrepancy between laboratory and field germination than is generally realized by many farmers. This discrepancy is probably greater for sorghum than for most crops. Experiments conducted at Amarillo, Tex., in I9I4 and I9IS showed (5) 3 that five varieties of sorghum planted on April I in wet, cold soil resulted in field germination ranging from I I ·4 to 33.o%. When the same varieties were planted on June I in moist soil when higher temperatures prevailed, the field germination ranged from 51.3 to 61.9%. Seed of good viability was used in the test, the laboratory germination being from 90 to 98.5%. It was concluded from the Amarillo data that "it is best to expect in field seedings of kafir, milo, and sorgo not more than so per cent of the laboratory germination and of feterita not over 40 per cent."Engledow and Ramiah (I) in England found that under reasonably good planting conditions from 6o to So% of the wheat sown may be expected to grow into plants to be harvested. Parasitic action, adverse weather, bad tilth, and bad seed (physiologically imperfect) were regarded as factors which may bring about very substantial reduction in the number of plants. These workers also found some varieties of wheat to be inherently poor and slow in germination as compared with other varieties.Five varieties of wheat planted under optimum field conditions at Hays in I935 germinated 83.o%. Under equally favorable planting conditions in June sorghums have averaged 6o%.

T HE importance of natural selection among plants has long been recognized, but the rate at which such changes may occur in cultivated crops has only recently received attention. Harlan and Martini 3 determined the rate of natural selection in a mixture of I I easily recognized varieties of barley which was studied at IO stations for a period of 4 to 12 years. They found that at all stations there was a rapid elimination of the less-adapted sorts. At most places the variety that would eventually dominate the population was quickly evident. The leading variety varied with the location of the station. A variety dominant at one station in some cases was eliminated at another. Few varieties survived at all stations. Some varieties increased for a time and then decreased.This paper is a report on the rate at which changes in varietal ratios occurred in successive crops after two varieties of winter wheat were mixed in the proportion of equal numbers of kernels. One mixture included Kanred and Harvest Queen, the other Kanred and Curren.One portion of each mixture was sown at Manhattan, Karis., and another at Hays, Kans., in the fall of I93I. Seed from the previous crop was used to sow each succeeding crop at the respective station so that the changes in the varietal ratios of the population were c~mulative for each station. The experiment was continued for 7 years at Hays and 9 years at Mahnattan. Since Harvest Queen and Currell are awnless, there was no difficulty in distinguishing these varieties from Kanred. The rate at which natural selection occurred was determined on the basis of separations of the plants which were pulled from sample areas. The number of heads and number of kernels of each variety were determined in the tests at Manhattan and the number of heads and weight of grain at Hays. CHANGES IN VARIETAL PLANT RATIOSThe cumulative changes in the percentage of plants that survived in the varietal population when Kanred was grown in competitive mixtures with Harvest Queen and with Currell are shown for both Manhattan and Hays in Table I.

No abstract