734 EXPERIMEN'T STATION RECORD. [Vol. 38 



tests, the seeds were planted in equal numbers at a noinial rate for the large 

 seeds. 



When two gi-ades each of spring wheat and oats were space-planted to per- 

 mit maximum plant development, the small seed compared with the large 

 produced SO per cent as many culms per plant, 72 per cent as high grain yield, 77 

 per cent as high straw yield, and 77 per cent as great total yield, the small 

 seeds averaging 52 per cent as heavy as the large. In yield tests comparing 

 large and small seeds planted both in equal numbers and equal weights at rates 

 normal for the large seed, (1) the small seed of winter wheat yielded 4 per 

 cent less than the large planted in equal numbers, with equal yields when 

 planted at equal weights; (2) the small seed of oats yielded 11 per cent less 

 than the large when sown in equal numbers, both yielding alike with equal 

 weights of seed; and (3) the small seed of spring wheat yielded 10 per cent 

 less than the large seed sown in equal numbers, and only 1 per cent less when 

 equal weights of seed were used. 



During 12 years of continuous grading of Turkey Red and Big Frame winter 

 wheat (by means of a fanning mill), the heaviest one-fourth seed averaged 0.4 

 bu. more, and the lightest one-fourth 0.5 bu. less than the unselected seed. For 

 the same period the heaviest and lightest one-fourth seed of Kherson oats 

 yielded, respectively, 0.83 bu. and 0.09 bu. more than the ungraded seed. Dur- 

 ing 8 years' continuous use of the fanning mill, the lightest one-fourth seed of 

 American Banner oats has yielded 1.43 bu. more than the heaviest one-fourth. 

 In a 4-year period the ungraded seed was also compared and yielded 1.6 bu. 

 less than the light seed, while tlie heavy seed yielded 3.67 bu. less than the 

 lightest seed. 



Based on a review of 60 experiments by various investigators, regarding the 

 relative yields of grades of small grain seeds, the following principles are in- 

 dicated : (1) When space-planted to permit maximum development of the in- 

 dividual plants, a higher yield per plant is obtained from large than from 

 small seed. (2) When planted in equal numbers at a rate optimum for large 

 seed, a lower yield is obtained from small than from large seed. (3) When 

 planted in equal weights, at a rate optimum for the large seed, all three grades 

 of seed — large, small, and unselected— -yield equally. (4) When distinct grades 

 of light and heavy seeds (or small and large) are obtained from a fanning 

 mill and planted in equal volumes slightly smaller yields are apt to result from 

 the light seed. The difference in favor of large or heavy seed as compared with 

 the original unselected seed is very slight and is deemed to have little prac- 

 tical significance, indicating that the practical use of the fanning mill consists 

 largely in the removal of weed seeds and trash. (5) Competition between 

 plants from large and small seeds sown in a mixture acts to increase the rela- 

 tive yield from the large seeds, suggesting a natural elimination (within a mass 

 variety) of poorly adapted types which produce unduly small or light-weight 

 seed. 



The effect of weeds upon cereal crops, Winifred E. Beenchley {New 

 Phytol., 16 {1911), No. 3-4, pp. 58-76; a&.s. in Physiol. Abs., 2 {1917), No. 6, pp. 

 368, 369; Internat. Inst. Agr. [Rome], Internat. Rev. Sci. and Pract. Agr., 8 

 {1917), No. 7, pp. 984, 985). — The author reports pot culture tests conducted 

 at Rothamsted for the past four years with crops grown in association with 

 weeds to determine whether vegetative competition is the sole factor suppressing 

 the growth of crops or whether weeds excrete poisonous substances from their 

 roots which actively inhibit growth. Wheat, barley, and buckwheat were grown 

 alone and with Alopecurus agrcstis, Brassica alba, Papaver rhoeas, and Spergula 

 arvensis, respectively. The plan of the experiments is outlined in detail, and 



