J m mediately preceded by two corn, crops. From many val- 

 uable notes recorded by the experimenter the following ex- 

 tract is taken as explanatory of the results on plots 7, 9 and 

 10: "Aug. 4. — No. 10 not fired or matured so much as No. 

 9 or No. 7. . . . Sept. 20. — All plots containing phosphate 

 have eufifered from drought, but plots with phosphate and 

 kai'nit more than others. There was no rust, but considera- 

 ble loss from the shedding of form«. Leaves were shed 

 about the middle of September, which wae due not to ma- 

 turity, but to drought and heat." 



The number of plants on each eighth-acre plot wae as 

 follows: 1131 on plot 1, 1151 on plot 2, 1037 on plot 3, 1042 

 on plot 4, 1143 on plot 5, 1126 on plot 6, 1105 on plot 7, 1013 

 on plot 8, 988 on plot 9, and 931 on plot 10. , 



The actual^yields, independent of the number of plants 

 per plot, constitute the basis for the following table. In 

 studying these results to learn whether the yields were 

 greatly affected by variations in the stand, a calculation 

 was made of the theoretical yields on the basis of a per- 

 fectly uniform stand. An analysis of these ^'corrected 

 yields" pointed to the same general conclusions as those 

 drawn from the actual yields. That is to say, the average 

 increase due to acid phosphate was 329 pounds on the basis 

 of actual yields and 332 pounds on the basis of yields cor- 

 rected to allow for variations in the stand. Likewise the 

 average increase on. four plots attributable to cottonseed 

 meal was 168 pounds by actual yields and 177 pounds by 

 ■"corrected" yields. For kainit the average increase on four 

 plots was 39 pounds, reckoned on actual yields, and 81 

 pounds on a basis of a uniform stand. 



