364 Bulletin 313 



seed demonstrated that there was little crossing, as the plats of the dif- 

 ferent sorts clearly exhibited their various characters, such as lateness 

 and earliness, color and form of head, and the like. The plats were i rod 

 wide and 4 rods long, thus containing one fortieth of an acre, excepting 

 in the case of four varieties, of which a shortage of seed necessitated using 

 smaller (^ rod wide) plats. These plats were sown about the middle of 

 September, 1909, the same comparative rate of seeding, 16 pounds per 

 acre, being used in each case. Every fourth plat was used as a check 

 and planted with seed of ordinary timothy. The seed for the check plats 

 was purchased from a local seedsman and was the best timothy seed that 

 he supplied to his trade. As uniform a piece of soil as could be obtained 

 was selected for these tests and was lightly manured before the seeding 

 at the rate of 10 tons of stable manure per acre. 



From a study of Table 3 (page 365) it will be seen that in 19 10 three of 

 the new sorts, 1,841, 1,849, a-i^d 1,853, produced less than the yields of 

 their estimated checks, but that in the other cases the yields were much 

 in advance of the checks. In this year all the 1 7 new sorts gave an aver- 

 age yield of 7,451 pounds per acre, while all the check plats gave an 

 average yield of 6,600 pounds per acre. The new varieties, even including 

 the three low yielders, gave an average increase of 851 pounds per acre. 



In 191 1, which was in general a less favorable season for timothy than 

 was 19 10, all the new sorts showed substantial gains over the checks. In 

 this year the average yield of all the 17 new sorts was 7,153 pounds per acre, 

 while the check plats gave an average yield of only 4,091 pounds per 

 acre. The new varieties in this year thus gave an average increase per 

 acre of 3,062 pounds, or over one and one half tons per acre. The reason 

 for the decrease in the check plats in the second season is clear to one 

 who has followed this work. Ordinary timothy rusts badly and, owing 

 to lack of vigor, is comparatively short-lived. A large part of the decrease 

 is due to these causes, but it is also due in some measure to the poorer 

 season. On the other hand, the new varieties were selected for vigor, rust 

 resistance, and ability to stool, and they would naturally increase in size 

 for three or four years and would gradually cover the ground m.ore thickly. 

 Clearly, these varieties were cut down by the poor season, or they would 

 have given a better yield the second year than the first. Throughout 

 this experiment, which was made as uniform as possible for ever>^ plat, 

 the new varieties have clearly shown their superiority over the ordinary 

 timothy in greater height and in thicker development on the ground. 

 (See Plate IX and Fig. 94.) 



Furthermore, it is important to note that the hay produced by the new 

 varieties was apparently of superior quality to that from the check plats. 

 Several hay dealers who were expert judges of the quality of hay visited 



