1918.] FIELD CROPS. , 733 



designed apparatus, and the loss not recovered in the vegetative growth or the 

 inert seed residue partially accounted for. Considerable tabulated data are 

 presented and fully discussed. The observations may be summarized as 

 follows : 



For all the grades of wheat seed tested in 191S and 1914, the total sprout 

 value of the seed averaged 54.2 per cent and 46.3 per cent of the weight of 

 seed planted, respectively. The total loss of substance not recovered in either 

 the sprout or the inert seed residue averaged 3S.5 per cent for the two years. 

 In all tests conducted during 1913 and 1914 the ratios for the moisture-free 

 weight of unselected seed to the large and small seed averaged, respectively, 

 100 : 127 and 100 : 85, while the ratios for the total sprout value averaged 

 100 : 123 and 100 : 88, respectively, indicating a rather close relationship between 

 the size of seed and its sprout value. As an average for three tests the carbon 

 dioxid liberated from wheat seeds by respiration during fourteen days' growth 

 In the dark in a nonnutritive medium amounted to 39.22 per cent of the 

 original moisture-free weight of the seed, the sprout value of the same seed 

 equaling 47.28 per cent of the original dry matter of tlie seed. 



Very small or shrunken wheat seeds were at a marked disadvantage in 

 comparison with large seeds, when planted at the unusual depth of 5 or 6 inches. 



Separation of the mature crop of wheat, grown at the normal rate of planting, 

 into individual plants was accompanied by an average error of 7.6 per cent, and 

 for this reason the number of individual plants surviving from large and small 

 seeds at harvest was not determined in these experiments. The relative pro- 

 duction of large and small seeds of wheat was determined when planted alone 

 and when grown in competition by alternating the seeds in rows planted at the 

 normal field rate. The small .seeds weighed 66 per cent as much as the large 

 seeds and had a sprout value 68 per cent as large, the germinations of the two 

 grades being practically equal. When planted alone the small seeds produced 

 6 per cent fewer culms, and in competition 18 per cent fewer culms than the 

 large; the yield of grain was 11 per cent smaller when planted alone and 24 

 per cent smaller in competition ; the yield of straw 6 per cent smaller for the 

 small seed alone, and 25 per cent smaller in competition ; and the total plant 

 yield 7 per cent smaller for the small seed planted alone and 25 per cent smaller 

 in competition than for the large seed. 



That competition between alternating plants of two wheat varieties may be 

 very marked was shown by Big Frame winter wheat in 1914, when grown at 

 the normal rates of planting, the yields of grain, straw, total crop, and number 

 of culms being respectively 90, 88, 89, and 80 per cent as large as for Turkey 

 Red. When grown in competition, however, Big Frame yields were respectively 

 only 55, 70, 67, and 68 per cent as large as for Turkey Red. Planted alone in 

 1915, the yields for Big Frame were respectively 82, 105, 99, and 94 per cent 

 as large as for Turkey Red, and in competition were respectively 120, 128, 

 125, and 117 per cent as large. Similar results were obtained for spring wheat. 

 These investigations are thought to indicate that competition may play a very 

 important role in the natural improvement of cereal crops. 



In a 2-year yield test of unselected, large, and small seeds of two winter 

 wheat varieties, the average relative seed weights were 100, 134.6, and 86.9, 

 with corresponding sprout values of 100, 133, and 92.3. The grain yield of the 

 large seed was 2.3 per cent superior to the unselected seed and of the small 

 seed, 3.1 per cent inferior. The 2-year average relative weights of unselected, 

 large, and small seeds of two spring wheat varieties were respectively 100, 

 117.3, and 78.4, while the corresponding relative sprout values were 100, 110.4, 

 and 71.8. The large seed outyielded the unselected for grain 11.8 per cent, 

 while the small seed was 7.7 per cent inferior to the unselected seed. In these 



