-80- 



Those plats showing a larger weight per stalk where 

 only two stalks grew were 1, 8, 11, 17, 18, and 19; those 

 showing about the same or varying comparative weights were 

 2, 3, 7, 10, 14, 20, 21, 24, and 25; those showing a smaller 

 weight where two grew in the hill were 22 and 26. Plat #22 

 the north end of which was markedly the poorest part of the 

 whole groxmd contained five rows with three or more hills 

 with two stalks in a hill. In four of these rows the 

 stalks were heavier where three grew in the hill and in 

 the other row there was hut litttle difference. 20 and 21 

 showed about the same weights for stalks growing under the 

 different conditions, while 18 auad 19 clearly showed an 

 adveuitage, where only two stalks grew in the hill. The 

 comparative uniformity of results in plat 22, together with 

 the gradual change in comparative weights as the plat is 

 approached from the west would seem to indicate a corre- 

 lation. But a glance at 26, where two rows were weighed 

 and where the growth was good shows a much more marked 

 difference in weights in the other direction. Conclusions 

 are, therefore, not warranted. 



The weight per stalk where two grew in a hill averaged 

 3.64 per cent more than where three greijr in a hill. There 

 was considerable variation. A comparatively few hills 

 had but one stalk. These showed a great variation in 



