152 On the Comj)Osition of Waters of Laiid-Drainaf/e. 



8. The preceding- analyses also affcjrd abundant evidence of 

 the marked difference in the relative power oi soils to absorb 

 potash and soda. 



Without exception all the drainag^e-waters contained but little 

 potash, and in all cases more soda than potash. It will be 

 noticed that the drainage from Plot 9, manured with nitrate of 

 soda, contained 2*94 grains of soda, and that from Plot 12, 

 upon which sulphate of soda was used, 2*41 grains, or very 

 much more than the drainage from the plots upon which 

 soda-salts formed no considerable part of the manuring mixtures 

 used. 



9. The drainage-waters in which phosphoric acid was deter- 

 mined contained only small, but apprecialjle, quantities of phos- 

 phoric acid. No great loss in phosphoric acid thus takes place 

 by drainage if land is highly manured with soluble phosphatic 

 manures, for the soluble phosphate of the manure is rapidly 

 fixed l)y almost everv kind of soil, and rendered insoluble to an 

 extent which prevents the rain passing through the land from 

 removing it to any great extent. 



During the summer months of 18(37 little or no rain passed 

 through the drains. Comparatively little rain fell in October, 

 and unusually little in November. In December there was a 

 full amount of rain. 



Fourth Series of Draixage-Watek Analyses. 



The next collection of drainage-waters (Table IV.) was made 

 on the 21st of April 18(58, when Dr. Gilbert again sent me 

 twelve samples lor examination. Unfortunately the drain from 

 the plot manured annually with farmyard-manure did not run at 

 the time the other samples were collected, nor could a sample be- 

 obtained from that plot in sufficient quantity for an exhaustive 

 examination at any later period of the year. In all the samples- 

 the proportions of potash and soda were determined separately, 

 and also, with one exception, the minute quantities of ])hosphoric 

 acid which the different samples contained. 



The ammonia was also determined quantitatively in several 

 samples of this series ; some sainples, however, contained merely 

 faint traces, and others no ammonia whatever. 



The results of analyses are given in detail in Table I\ ., 

 (pp. 154, 155). 



A glance at the analytical results contained in this Table will 

 bring to view some points of interest. 



1. In the drainage-waters froin the majority of the experi- 

 mental plots, the total amount of solid matter varied much less than 

 in the preceding series. Excepting the drainage from the nitrate 



