. 13 



Increase of seed cotton per acre when kainit was 

 added : 



To unfertilized plot. .". —27 lbs. 



To cotton seed meal plot — 1 lbs. 



To acid phosphate plot 124 lbs. 



To seed cotton meal and acid phosphate plot. . . — 50 lbs. 



Average increase with kainit, - r - - - 12 lbs. 



Mr. Cunningham's experiment, like both of the tests- 

 made by Mr. Purifoy, on the same class of land, white 

 prairie, indicates that phosphate was most needed. The 

 largest yield was made with a mixture of cotton seed 

 meal and phosphate. Kainit did not increase the yield, 

 though it did seem to somewhat restrain the rust on 

 Plots 9 and 10. 



It should be noted that white prairie soil was not 

 very responsive to commercial fertilizers and that none 

 of these paid a very large profit. 



Although phosphate was undoubtedly useful in each 

 of these experiments, its effects were far less notable 

 than the favorable influence that is exerted by adding 

 suitable vegetable matter to this class of soils. We can- 

 not yet recommend the use of phosphate on these soils, 

 believing that the same money invested in the seed of 

 melilotus or of other renovating plant would be more 

 profitably spent. 



EXPERIMENTS MADE BY J. S. DUNCAN ON G. W. FREEMAN^S- 



FARM, 1^ MILES SOUTHWEST OF MAPLE GROVE, 



CHEROKEE COUNTY. 



In 1899 the test was made on gray sandy upland, with 

 red subsoil ; in 1900 6n light alluvial second bottom of 

 a dark gray color, with red subsoil. Both fields had 

 been cleared for more than a quarter of a century. The 



