— 85 — amount varying with different types of soil. Russell * suggests that this stability is caused by the formation of a compound which is not a true ammonium salt.

To ascertain what percentage of an added ammonium salt is retained by the farm soils despite the leaching effect of the irrigation water, the following experiment was devised. Samples of soil were left in contact with a solution of NH 01 (1 cc. =1 CO. N.) for forty-eight hours. A measured portion of the supernatant solution was then taken, and the nitrogen content measured by means of a strong solution of hypobromite of soda and a Hempel's Gas Burette.

The samples were eight in number, six of which were the same as have been used in previous investigations. The history of the plots was as follows : Sa,Tnple.