366 Bulletin 313 



the plats during the season. They stated without reservation that the 

 quaUty of hay produced by the new sorts, owing to the retention of a 

 fresh green quaHty and leafiness, would be superior to that on the check 

 plats of ordinary timothy. This apparent superior quality is due in the 

 main to the greater rust-resistant character of the new sorts and to the 

 fact that the majority of them were selected as " green leaf " types, that 

 is, possessing the ability to retain their leaves and stems green until the 

 heads have become fully mature. This character is doubtless due in 

 large measure to the inherent vigor of the types and to their resistance 

 to rust. 



At the time of harvesting the broadcast plats, the difference in green- 

 ness between the selected types and the ordinary timothy on the check 

 plats was very noticeable. A much larger proportion of the basal and 

 lower leaves was dried up and dead on the check plats, which was clearly 

 not due to greater earliness of the ordinary timothy. 



The various new types in this test differed considerably in season of 

 maturing, some being early, others late, and others mid-season. While 

 there was much variation as to time of maturing among the plants on 

 any plat and, in general, the differences were clearly shown, the plats of 

 the selections ripened much more uniformly than did the ordinary timothy 

 on the check plats. In view of the great importance of cutting timothy 

 at a certain stage in its development, which will be discussed in an ap- 

 pendix to this paper, it would seem to be important to have the crop matured 

 as uniformly as possible in order to secure the best yield and the best 

 quality of hay. Up to the present time no attention has been given to 

 this very important factor in growing timothy hay, as we have had no 

 different timothy varieties to compare. 



The comparative yields as given for two years in Table 3 may profit- 

 ably be more carefully analyzed. It is not to be expected that all the 

 types selected as promising will prove equally good. Indeed, we should 

 rather expect that some will fall down and prove worthless. The most 

 surprising thing in this experiment is the uniform superiority of all the 

 new types tested. It is interesting, however, to note the variation in 

 the yield of the new types. In the column showing gain in yield for 

 19 10, it will be noticed that three varieties gave an increase in yield of 

 over 1,500 pounds per acre, and these three varieties were among the 

 heaviest yielders in 191 1. While the average gain in yield per acre in 

 1911 was only slightly more than i| tons per acre, being 3,062 pounds, 

 four varieties gave over 2 tons per acre. In this year the six best-yielding 

 sorts in comparison with their corresponding check plats gave an average 

 increase of 4,040 pounds per acre. Naturally, for new varieties or races 

 for general propagation we should select the very highest yielders among 



