FORMULA. 



(a) For certain red lands I 2.3 



(b) For certain sandy lands I 2.0 



(c) For low longleaf pine lands .... 1.9 



(d) For "rusting" soils \ 3.0 



10. On llic lime soils of the Cciitral I'raiiic Ki'j;iou coiu- 

 iiiercial fertilizers are not j»enerally used. Prairie soils are 

 often in poor mechanical condition and need ve«etable mat- 

 ter and drainage more than commercial fertilizers. The 

 poorer soils often need both cotton seed meal and jdiosphate. 



IXCOXCLUSIVH Exi'KRl .\1 KNTS. 



These sometimes afford suggestions or hints \vhi( li may 

 serve to strengthen the lonclusions derived from ilic more 

 positive experiments previously recorded. 



At Town Creek, one-ciuarter of a mile southeast of the 

 town, ^Ir. A. A. Owens made the test on what he describes 

 as white sandy land with yellowish subsoil. There was no 

 rust, but drought. This experiment was undertaken by Mr. 

 R. R. Reed, who turned the fertilizers over to Mr. Owens. 

 The test is inconclusive for the reason of the tearing of one 

 of the fertilizer sacks, probably that for plot 0. 



The Xewtonville experiment in Fayette county was made 

 by G. W. Gravlee, but was vitiated by late germination and 

 irregular stands. 



The experiment at Hanover, Toosa county, was made by 

 J. M. Logan on gray gravelly or rocky land with red clay 

 subsoil. The results suggest that kainit was not needed. 



The test one mile east of Fredonia, Chambers county, was 

 made by E. W. Smart on dark ujdand with red subsoil. In- 

 equalities in stand, due, he rejtorts. to disease of the young 



