150 



HATCH EXPERIMENT STATION. 



[Jan. 



an acre. Nitrate of soda at the rate of 150 pounds per acre 

 is applied to plots 2 and 4, while the application on Plot 6 

 is 200 pounds, and on Plot 8 250 pounds. To the remain- 

 ing plots no nitrate is applied. The first crop of hay in 

 this field was housed on July 14. The rate of yield was 

 6,314 pounds per acre. When this experiment has been 

 tried in previous years, it has been found a matter of con- 

 siderable difficulty to spread the relatively small amounts of 

 nitrate of soda used evenly ; and, as a means of obviating 

 this difficulty, the nitrate used on each plot during the past 

 season was mixed with basic slas^ meal. The amount of slao- 

 meal applied was 137^/2 pounds per plot, and the slag meal 

 was applied to the plots receiving no nitrate as well as to 

 the others, and on all in equal amounts. The mixture of 

 slag and nitrate remained dry, and its even application was 

 relatively easy. At the rates used, the nitrate and slag 

 were mixed in proportions varjdng from about one nitrate 

 to three slag to about one nitrate to two slag. Even wnth 

 the higher proportions of nitrate to slag, the mixture re- 

 mained dry and in convenient form for application. The 

 rates of yields on the several plots are shown in the follow- 

 in »■ table : — 



Nitrate of Soda for Roicen. — Yields per Acre (^Pounds') . 



[Basic slag meal at the rate of 1371,^ pounds per plot.] 



Plots. 



Nitrate used (Rates pek Acre). 



Yield. 



Plot 1, . 



Plot 2, . 



Plot 3, . 



Plot 4. . 



Plot 5, . 



Plot 6, . 



Plot 7, . 



Plot 8, . 



No nitrate 



Nitrate of soda, 150 pounds, 

 No nitrate, .... 

 Nitrate of soda, 150 pounds, 

 No nitrate, .... 

 Nitrate of soda, 200 pounds, 



No nitrate 



Nitrate of soda, 250 pounds, 



716 

 1,341 



990 

 1,432 



853 

 1,234 

 1,021 

 1,932 



In the effort to determine whether the application of 

 nitrate is profitable, the yield wherever it has been applied 

 has been compared either with the yield of the nearest plot, 

 or, in cases where it is possible, with the average yield of 



