156 On the Composition of Waters of Lmtd-Drainage. 



This sample it will be seen contained no less than 15*74 grains 

 of nitric acid per gallon. It will further be noticed that the 

 nitric acid passed into the drains partly in combination with 

 soda, and partly with lime, magnesia, and potash. 



Tlie combined results of the analysis of the water from Plot 9 

 show that there was in a gallon 15"48 grains of nitrate of soda, 

 *yO nitrate of potash, 2*55 grains of nitrate of magnesia, and 

 5-33 grains of nitrate of lime. In percolating through the soil, 

 nitrate of soda, it appears, suffers decomposition, and gives rise 

 to several compounds of nitric acid with bases ; and it passes off 

 not simply as nitrate of soda, for the amount of soda which was 

 found in the drainage was not sufficient to bind the nitric acid, a. 

 portion of which entered in union with other bases, and passed 

 into the drains as nitrate of lime and magnesia. 



We learn from these experiments that in a wet spring the 

 loss of nitrogen by drainage is unquestionably very great when 

 the wheat crop is top-dressed with nitrate of soda ; and that the 

 loss of nitrogen in wet seasons is much more considerable in the 

 case of nitrate of soda than when ammoniacal salts are employed 

 as top-dressings for wheat. 



Fifth Seeies of Deainage-Water Analyses. 



The last series of analyses was made with samples collected 

 December 29, 1869, at Rothamsted, at an enormous flow of 

 the drains. 



In October there was a deficiency of rain, and in November 

 hardly any rain fell. In December, on the other hand, we had 

 a very great excess of rainfall. 



Notwithstanding the heavy and continuous rains which fell in 

 December, and which caused the pipe-drains from all the other 

 drains to run freely, little or no water passed through the drain- 

 pipe from the plot usually manured with farmyard-manure, and 

 consequently no analysis of the drainage from that plot could 

 be made in December 1869. Dr. Gilbert informs me that whilst 

 the pipe-drains from every one of the other plots in the experi- 

 mental wheat-field run freely, perhaps four or five times or more 

 annually, the drain from the dunged plot seldom runs at all more 

 than once a year, and in some seasons not at all. The fact is 

 the vast accumulation of decomposing organic matter on the plot, 

 annually dunged at the rate of 14 tons of farmyard-manure per 

 acre, lightens the soil and promotes the disintegration of the 

 clayey portions, and altogether renders the surface-soil more 

 porous, and capable of retaining much more water. Hence a 

 much greater amount and continuity of rain is required before 

 the porous surface is fully charged with rain, and an excess can 

 reach the drains. 



