12 MASS. EXPERIMENT STATION BULLETIN 346 



These results indicate that with soil of the kind used in this experiment, 

 which has been heavily fertilized for a number of years, a reduction of J>^ to 

 3/8 of the fertilizer can be made by using row, instead of broadcast, application, 

 for a period of 2 to 4 years. After that time, both yield and quality are likely 

 to decline with the reduced quantities. 



This experiment, as have others, showed that pound for pound fertilizer 

 placed near the plant was more effective than fertilizer broadcast. In the 

 practical application of the results, however, one must consider several factors. 

 Is the return on each dollar invested in fertilizer as important as the return 

 per acre of crop? If the cost of fertilizer is high and the price of tobacco low, 

 it may be profitable to reduce the amount of fertilizer used by resorting to 

 row application; whereas, if the reverse relation holds, it would probably be 

 unwise to economize in the quantity of fertilizer. In the last analysis the 

 question is an economic one. In view of results of experiments elsewhere, it 

 is quite probable that a combination of broadcasting and row application of 

 fertilizer might give better results than either method alone. 



General Discussion 



These experiments were conducted through a period when there was a 

 considerable variation in the rainfall during the growing season of tobacco. 

 In Table 8 are given figures for the rainfall in May and the summer months 

 of the years of the experiments. The tobacco plants were usually set in the 

 field about the first of June and harvested about the third week in August. 



Table 8. — Precipitation at Amherst, Massachusetts 



Normal 1927 1928 1929 1930 1931 1932 1933 193^ 1935 

 May 3.63 4.83 3.25 4.17 3.34 7.44 1.67 1.69 3.42 2.17 



Total 12.11 11.78 21.60 5.30 10.79 14.68 9.12 12.56 9.42 9.42 

 (June, July, August) 



Tobacco of lowest yield and poorest grading quality was produced in 1928 

 when the summer months were abnormally wet. In 1929, a year of summer 

 drouth, yield was relatively lower than quality. The best yield was produced 

 in 1931, a year of summer rainfall slightly in excess of the normal; but the 

 best quality was obtained in 1930 when a comparatively dry August followed 

 a nearly normal July, a wet June, and a practically normal May. 



The experiment with the different quantities of nitrogen is hardly com- 

 parable with others which have been reported, because of difference in type of 

 tobacco, soil conditions, climate, etc. For example. Garner and his associates 

 (8) reported experiments conducted in southern Maryland, with different 

 quantities of nitrogen applied to Maryland broadleaf tobacco. The maximum 

 quantity of nitrogen used was 80 pounds per acre, which gave only a small 

 increase over 40 pounds. 



