28 



plants up to the time when the root tubercles begin to 

 eyej'cise their function of supplying nitrogen. 



The fertilizer test detailed in a preceding paragraph 

 shows tliait with a complete fertilizer the yield of pt as 

 was 3.8 bushels per acre greater than where only phos- 

 phate and potash Avere used together. 



This increase seems to be attributable to the use of 

 80 pounds of nitrate of soda. 



The majority of experiments agree with the one 

 where cotton seed meal was used in indicting that 

 nitrogen is not a profitable constituent of the fertilizer 

 for cowpeas. 



Forms of Phosphate for Cowpeas. 



A test was made in 1896 of acid phosphate, 

 crude Florida soft phosphate, and a moistened 

 mixture of these two, which mixture should have 

 produced reverted phosphate. The crop was a 

 failure, probably because of injuries to the roots 

 by nematode worms, and there were only slight differ- 

 ences in the yields of seed on the plots differently fer- 

 tilized. This was^on very poor white sandy soil. 



In 1898, co-operative tests of acid phosphate in com- 

 parison A^lth equal Aveights of Florida soft phosphate 

 (crude) were made for this Station by Mr. A. A. Mc- 

 Gregor, on a loam soil with clay sub-soil, at Town 

 Creek, Ala., and by Mr. J. P. Slaton, on sandy soil be- 

 tween Notasulga and Tuskegee. Apparently the soil 

 at Town Creek was rich in lime, the other poor in 

 lime. 



Unfortunately there Avas a failure to pick the peas in 

 both the tests, but the notes made by both of the ex- 

 perimenters have no doubt as to the superiority of acid 

 phosphate over insoluble phosphate as a fertilizer for 



