0)1 the Cumposition of Waters of Land-Drainage. 153 



of soda plot, which contained as much as 47"85 grains of solid 

 matter per gallon, the amount varied from 14'15 to 2iJ'65 grains 

 in the unmanured and the highly manured plots, whilst the 

 drainage from the latter collected in the month of January con- 

 tained from 47 to 57^ grains. This seems to indicate that the 

 s(jluble matters of the soil are taken up by the wheat-plant in 

 spring, and that the loss which the soil sustains in fertilizing 

 matter by drainage is not nearly so great during the active growth 

 of a crop as during the autumn and winter months when no crop 

 is on the ground. 



2. The composition of the drainage-waters collected on the 

 21st of April 181)8 was Aery similar to that of the drainage-waters 

 collected on the 21st of May 1867. 



3. Practically speaking potash was not removed to any extent 

 from anv of the experimental plots in the drainage. 



4. In the drainage-water from the plots manured with sulphate 

 of soda, more soda was found than in the drainage from the 

 plots upon which no sulphate of soda had been employed as a 

 manure. 



5. More magnesia likewise occurred in the water from the plot 

 manured with sulphate of magnesia than from other parts of the 

 same field. 



G. All the drainage-waters from the plots to which super- 

 phosphate and chlorides had been applied contained more phos- 

 phoric acid and more chlorine than the unmanured plots ; the 

 amount of phosphoric acid which passed into the drains, how- 

 ever, was but small in every case. 



7. In some of the samples the amount of soluble silica was 

 considerable. For instance, the drainage from Plot (> contained 

 5*10 grains of soluble silica, and that from Plot 13, 5 '20 grains. 

 Both samples were perfectly clear, they having been filtered 

 through fine filtering-paper before they were submitted to 

 analysis. Moreover, in both samples the proportion of soluble 

 silica was determined twice with closely agreeing results, as I 

 feared some mistake might have occurred in the first determi- 

 nation. There is, therefore, no doubt about the correctness oi' 

 the silica determination, but I am at a loss to understand why 

 some of the drainage-waters contained but little soluble silica 

 and others, comparatively speaking, a large amount. 



8. With the exception of the drainage from Plot 0, all the 

 samples contained comparatively little nitric acid. 



On the whole, however, there was more nitric acid present in 

 the April collections than in the samples collected in May 1867. 



9. Particular attention is directed to the large proportion of 

 nitric acid in the drainage from Plot 9, which was manured in 

 spring with 550 lbs. of nitrate of soda. 



