SUGAR BEETS: PREVENTABLE LOSSES IN CULTURE. 7 



Plat E shows a fairly normal correlation between stand and yield, 

 while plat F decreases on account of late and careless thinning, no 

 less than 9.82 per cent of pairs of plants being found. 



Plat G shows a greater yield than plat E, because its soil was very 

 well prepared and manured and the cultural work done by Japanese 

 who were working for themselves. 



Referring to the analysis of observations for the season of 1912, as 

 presented in Table II, the plats in group 1 were operated under 

 almost experimental conditions. The seed was sown under the 

 direction of the writer. Two pairs of plats were studied, one pair in 

 each of two contiguous fields possessing very uniform soil. One 

 farmer worked plats 1 and 2 while another farmer worked plats 3 

 and 4. The rows extended entirely across each field. The two 

 farmers carried on or supervised all the cultural operations. The 

 only appreciable difference in the handling of the plats was that the 

 seed was sown slightly deeper in one plat of each pair than in the 

 other. This resulted in more damping-off in the deeper sown plats, 

 thus producing therein rather poorer stands. In this group the 

 correlation between yield and stand is complete. From this group 

 and plat D of 1911 it would appear that the optimum area per plant 

 is considerably greater in deep, light, well-manured sandy soils than 

 in loams and heavy loams. In group 2 a perfect correlation is shown 

 between yield and stand. At the same time, the effect of inexperience 

 and lack of supervision is evident in plat 9. In group 3 the corre- 

 lation between stand and yield actually is accentuated by the 

 exception shown in plat 10. This discrepancy is entirely accounted 

 for by the adverse conditions to which this plat was subjected. 



CORRELATION OF STAND AND YIELD. 



In Table III all the plats for the three years are arranged according 

 to the percentage of the harvest stand, beginning with the highest. 

 They are divided into three groups, representing three soil types. In 

 this table the percentage of the harvest stand is placed side by side 

 with the yield of each plat. This reveals a striking correlation between 

 the percentage of stand and the yield. The apparent exceptions are 

 accounted for by the special conditions described under the heading 

 "Notes" in Tables II and III. Upon allowing for the adverse or 

 especially favorable conditions mentioned, it may be said that the 

 correlation is complete. 



In Table III the means of the yields of groups 2 and 3 exhibit a 

 ratio almost exactly equal to the relation of their respective stands. 

 Based on the figures given, the mean yield of group 3 would be 17.72 

 if the stand were the same as in group 2. 



