24 ■ On the Rain and Drainage - Waters at Jtotliamsted. 
the drainage from Plot 2, which shows a low proportion, the 
magnesia only ranges from 3 "5 to 5*5 per cent, of the lime, and 
rises or falls with it. Plot 14, which receives nearly three times 
as much magnesia as Plots 5-8, is no exception, the magnesia 
in the water, though the greatest in absolute quantity, being 
still only 5*1 per cent, of the lime. Where sulphate of mag- 
nesium was applied (excepting on Plot 14) the amount of 
magnesia in the drainage alone, that is besides the smaller 
amount in the crops, was approximately the same as in the 
manure, but in the other cases the soil itself contributed not 
much less magnesia to the waters. 
For the purpose of illustrating the annual losses of lime and 
magnesia which the soil suffers by drainage, we will assume that 
the annual drainage in Broadbalk Field amounts to 10 inches 
(2^ million lbs.), and that it has the composition shown by the 
mean analyses of Voelcker and Frankland given in Table XLV. 
The lime and magnesia annually lost by the unmanured Plot 
3&4 will then be 223 lbs. ; by Plot 5, receiving only mineral 
manure, 297 lbs. ; by Plot 9, receiving nitrate of sodium and 
half a dressing of mineral manure, 284 lbs. ; by Plot 10, receiv- 
ing 400 lbs. of ammonium-salts only, 389 lbs. ; by Plot 11, 
receiving 400 lbs. ammonium-salts and superphosphate, 443 lbs. ; 
and by Plots 6, 7, 8, 12, 13, 14, receiving an average of 400 lbs. 
ammonium-salts, with the sulphates of potassium, sodium, and 
magnesium in addition, a mean of 485 lbs. per acre. A reference 
to Table XXXVIII. will show that the quantity of lime and 
magnesia in the usual dressing of mineral manure is 104 lbs., 
the amount lost is thus greatly in excess of that applied to 
the land. The estimates just given are probably rather below 
than above the truth. 
The amount of phosphoric acid found by Dr. Voelcker in 
the drainage-waters was very small. The determinations 
made were lew in number, and the results display little regu- 
larity. It is, in fact, left uncertain whether the amount of 
phosphoric acid is increased by the use of phosphatic manures, 
or whether it is influenced by the addition of ammonium-salts. 
The mean of all the determinations gives 0 93 of phosphoric 
acid per million of water. If we assume, as before, 10 inches as 
the average annual drainage in Broadbalk Field (in recent wet 
years the drainage would be much greater), the annual loss of 
phosphoric acid by drainage would be 2*1 lbs. per acre. 
The determinations of potash are much more numerous than 
those of phosphoric acid ; they also display great irregularity. 
It is quite evident, however, that where potash was applied in 
the manure, the amount is distinctly increased in the drainage- 
water. The drainage-waters from the six plots receiving no 
potash contain an average of 1-G, and the waters from the eight 
