STUDIES ON OAT BREEDING. 185 



In regard to the relation of straw to grain the average for all 

 eleven varieties was 1.41 pound of straw to a pound of grain. 

 The Senator gave as much as 1.7 pounds of straw to a pound of 

 grain. On the other hand the Imported Scotch gave only 1.21 

 pounds and the Kherson 1.23 pounds. 



VARIATION IN THE WEIGHT PER MEASURED BUSHEL. 



In tables 3 to 6 there is given the weight in pounds of a meas- 

 ured bushel of each of the varieties that have been continued in 

 the test. These weights were obtained by using an ordinary 

 two quart standard grain tester. Further the determinations 

 were all made by the same method and by the same person (C. 

 W. Barber). It has recently been shown by one of us^° that 

 variations in the method of taking bushel weights may affect 

 the result very materially. The method used in these varieties 

 was to pour the grain into the tester, shake the tester five times 

 and then stroke with a zig-zag movement. This method gives 

 a slightly higher average weight (ca. 1.6 pounds per bushel) 

 than the usual method but it has the great advantage of being 

 subject to much less random fluctuation and hence a smaller 

 probable error. (Cf. Barber loc. cit.). 



It has further been shown that tested on 100 samples of 

 Lincoln oats this method gave a standard deviation of 0.3245 ±: 

 .0110 pounds. On the basis of this result we would not expect 

 to get deviations of more than one pound from the mean in 

 either direction in any sample drawn from the same lot of seed 

 and measured by this method. Consequently differences in the 

 bushel weight of one pound or more indicate clearly a differ- 

 ence in the character of the oats. 



In all cases the oats have been run through the fanning mill 

 once before the bushel weights were determined. Of course 

 it is not always possible to run different lots of seed, especially 

 in different years, through the fanning mill in exactly the same 

 way. However, it appears that only in exceptional cases would 

 the difference in fanning make as much as one pound difference 

 in the bushel weight of a given sample. 



"Barber-, C. W. Note on the Accuracy of Bushel Weight Determina- 

 tions. Me. Agr. Exp. Sta. Ann. Rept. 1914, pp. 69-75. (Bui. 226). 



