37 



with this cereal is larger than with any other. The extremes in weight 

 per 100 grains were, serial No. 3200, the lightest, from Nebraska, 1.582 

 grams, and serial No. 3277, from Oregon, the heaviest, 3.786 grams. 

 Cleanliness varied from 99.8 per cent, to 70.0 per cent., but of course 

 had nothing to do with locality. The heaviest weight per bushel was 

 found in specimens from Colorado, serial No. 3020, and Dakota, serial 

 No. 3036, weighing 48.8 and 48.6 pounds, The lightest were from Ala- 

 bama, serial No. 3002, and from Florida, serial No. 3043, 24.7 and 26.9 

 pounds, respectively. 



In milling oats the relation of kernel to waste is about one-half. Our 

 results show that the relation of kernel to husk averages for the whole 

 country 7 to 3, those from the Western States being a little less husky, 

 and those from the South considerably more so. It is, however, the 

 inflated nature of the husk in the Southern grain and the fact that the 

 glumes or outer husk is often adherent that affects the weight per bushel 

 more than the slightly larger proportion. 



The extremes found were 79.28 per cent, of kernel in a specimen from 

 Washington Territory, serial No. 3435, and 55.37 per cent, in one from 

 Dakota, serial No. 3391. Washington and Oregon sustain their repu- 

 tation for fine looking grain, while the small proportion of kernel in 

 the Dakota specimen is due entirely to cutting before it was quite 

 ripened. It is hardly a fair example, other specimens from the Terri- 

 tory reaching over 70.0 per cent, of kernel. 



In weight per bushel the warm climate of the South so affects the form 

 of the grain as to lower its average 2. 7 pounds. This is hardly as large 

 as would be expected, and leads to the conclusion that the climate has a 

 greater effect than some other characteristics. One of these is yield, 

 which, from an average of 30 bushels in the North, falls to about 10 in 

 the South, and, as has been said, the color and shape of the grain is 

 much changed. 



CHEMICAL COMPOSITION. 



In examining the physical relations of the specimens they were sep- 

 arated into kernel and husk, and for several reasons they were sepa- 

 rately submitted to analysis with, as it appears, results furnishing much 

 more information than would otherwise have been obtained. Corn, 

 wheat, and rye were analyzed in the clean kernel, and comparison is 

 more readily made between them and oats. The variations, independ- 

 ent of the proportion of husk, are easily arrived at, and since the 

 analysis of the hull and its relative amount are given it is easy to cal- 

 culate for any specimen its composition as it would be used for feed. 



The data obtained are here presented, together with averages derived 

 therefrom : 



