.'>() Food von Pi ants. 



The yields per acre for the different plots for the years 

 1905 and 11)06 were as follows: 



1905 750 lbs. sei'd cotton. 1, IIG lbs. 1, 272 lbs. 1, 440 lbs. 



IflOG 930 lbs. seed cotton. 900 lbs. 1,284 lbs. 1,776 1})S. 



As above stated, all of these plots were fertilized 

 equally as regards the amount of phosphoric acid and 

 potash, so that the effects of supplying or withholding 

 Nitrate of Soda could be easily noted. 



It will be noted that the increased yields are particu- 

 larly striking in the case of the application of 84 and 12G 

 pounds of Nitrate. On plot 2, in 1906, the yield was 

 practically the same as that on plot 1, but this was due to 

 the fact that a few rows in jjlot 2, owing to the stand on a 

 part of the plot being not so good and possibly on 

 account of some other condition, brought down the aver- 

 age yield per row of that plot. A majority of the rows 

 of that plot, however, undoubtedly gave a better yield 

 than plot No. 1, and it was apparent to the eye that most 

 of this plot was superior to plot No. 1. 



In 1905 it was noted that the cotton grown upon the 

 " No Nitrate " plot rusted quite badly, while plots 3 

 and 4, upon which an abundance of Nitrate had been 

 applied, were almost immune from rust. 



In addition to experiments in which the Nitrate was 

 applied at a single application, tests were made upon 

 some smaller plots to note the effects of the application 

 of the Nitrate in two different applications, the second 

 application being made about sixty days after planting. 

 It was found that there was only a slight difference in 

 the relative yields, but this slight difference was in favor 

 of the two applications. It is doubtful, however, if the 

 increase would have justied the additional cost and labor 

 of the second a])plication. 



Experimental tests upon small lots of the seed cotton 

 produced in 1906, showed that the yield of lint was about 

 :)4.4 per cent, of the weight of the seed cotton, but no 

 data Avas secured with regard to the proportionate yield 

 of lint ill 190;"). Ap])lying these iigures to the excess yield 



