142 



REPORT OF THE COMMISSIONER OP AGRICULTURE. 



is made in the seed used for the following: crop. Eeversion is much 

 more easily accomi)]ished than improvement. 



In Colorado, a storm in 1883 prostrated the plant at a time when the 

 injury interfered with assimilation of its nitrogenous constituents, al- 

 though afterwards the grain filled itself out fairly well with starch. 

 This grain, used as seed for the ensuing year, gave a crop which still 

 showed lack of nitrogen, and although a certain recovery had been 

 loade, the depreciation in its chemical composition was very evident. 

 In the same way, in Ohio, a storm prevented the grain from being tilled 

 out with starch, leaving it small and shriveled, and relatively rich in 

 nitrogen. These cases show the necessity of obtaining new seed when 

 any injury happens to the crop of the preceding year. 



An examination of the chemical composition of Avheat grown Avith the 

 greatest care on the same land for a number of years seems to show that 

 many varieties depreciate from their original valuable qualities. It is 

 an old idea, and many farmers practically change their wheat now and 

 then, but the change in these cases is oftener necessitated by careless- 

 ness in cnltivation than by natural causes. In Colorado quite a number 

 of the varieties, cultivated with great care by Professor Blount, have 

 steadily depreciated after improving during their first year in the new 

 locality. Others have held their own, and some have improved. The 

 lesult'of our analyses has, however, shown that care alone is not suffi- 

 cient to improve a wheat. It must be at least fi\irly well suited to the 

 locality in which it grows. 



The data of the first three years of the experiments have been already 

 published; those of last season's work show that, while the results of 

 the storm of the previous year were not entirely obliterated in 1884, 

 nevertheless some wheats v.ere produced which were remarkable for 

 their size and weight. One, for example, weighed 6.620 grams per 100 

 grains, and the general increase has led to an investigation of the re- 

 lation of size and other properties to that of weight per bushel in 

 wheat from various parts of the country and the relations of extremes 

 to each other. The latter are not at all coincident, as may be seen from 

 the following figures and numbers denoting different varieties: 



Extremes among Colorado wheats of 1884. 



T^eltl per acre — 



Weielit per bushel 



Weight per one hundred grains 

 Albuminoids 



nijrhest. 



Serial 

 number. 



Lowest. 



21. 3 bn. 

 62. 2 lbs. 



3. 160 grams. 



9. 45 per ct. 



Serial 

 number. 



3575 

 3534 

 352S 

 3524 



The weight per bushel is apparently dependent on several causes. 

 High weight is almost, if not always, an evidence of high quality, but 

 not always an evidence of a large and plump grain. The hard red 

 spring wheat of the NorthwcwSt, which is smaller in size and not as plump 

 as many other varieties, is heavy in its weight per bushel, while the 

 largo full wheats of Oregon, which are very starchy, are light. The 

 following data show some of the variations : 



