41 



stated, was primarily to test rust resistance, the work naturally soon 

 developed into a study of the characteristics in general of the varieties 

 of different natural groups of wheats, and of groups considered geo- 

 graphically, and some most interesting facts were thus obtained which 

 will be of great value in the work of wheat improvement. In fact, 

 many statements made in the foregoing discussion of the " Sources for 

 desiral)le qualities" are based upon the results of these experiments. 

 Varieties were obtained from every wheat country of the woi'ld, 

 aggregating about 1,000 rather distinct sorts in all. The manner of 

 securing these wheats, and the time and labor thus involved, together 

 with the difficulties of nomenclature arising from the confusion of 

 varietal names which prevails generally, have all been discussed in a 

 former bulletin on Cereal Rusts of the United States, and need not be 

 referred to here. The varieties were grown one year in Maryland and 

 most of them one year in Kansas, while about 300 of them were grown 

 two years in Kansas, or three years in all; that is, during 1895, 1896, 

 and 1897, During the same time a number of the varieties, especially 

 from Russia, Siberia, Japan, and Argentina, were tested by other 

 parties in cooperation with the Department, in other localities, viz, in 

 Michigan, Wisconsin, Indiana, Tennessee, and Colorado; and in the 

 case of a few of these, the experiments have since then been repeated 

 in Colorado, Kansas, and Nebraska, 



In conducting these experiments all the principal characteristics of 

 the wheat plant, as shown in its different stages from that of the young 

 plant to harvest time, were studied, though complete notes can not be 

 given for every variety. These features include in a general way 

 (1) the character of the young plant; (2) hardiness, including resistance 

 to rust, drought, and cold; (3) character of the head; (1) character of 

 the grain; and (5) time of maturity. Field experiments alone do not 

 show those qualities of the grain which indicate the value of the variety 

 for different uses, and which are after all more important than any 

 others; though it must be remembered that certain characteristics of 

 the growing plant often indicate quite correctly what these qualities 

 will be. Therefore chemical analyses have been made of a large 

 number of representative varieties, and for many of them the absolute 

 weight and specific gravity of the grain have also ])ecn determined. 



As would be expected, a large number of the varieties either proved 

 to be entirely unsuited to the conditions in this countrj^ or were found 

 to be in other respects undesirable sorts. It was the purpose from 

 the start to discard gi'aduallv all varieties that seemed to be uhjible to 

 adapt themselv(!s to their new environment. During the tirst year oi 

 the experiments at (iarrett Park, Md,, many of them were planted so 

 late, on account of their late arrival in this country, that much allow- 

 ance must be mad(^ for their behavior in comparison with others 

 which were sown in good season. In other respects the trial for that 

 year was very satisfactory, and afforded an excellent opportunity of 

 c()m])iiring the behavior and <|ualities of a large number of varieties 

 under average conditions. 



