332 
On the Rain and Drainage - Waters at RotJiamsted. 
have, bj and bye, to consider the drainage-waters obtained from 
manured land. 
A word must next be said on some percolation experiments 
in which nitrate of sodium was applied to the soil ; these 
illustrate afresh the facts just pointed out, and at the same 
time exhibit some special relations of nitrates with soil, which 
we shall do well to bear in mind. The first experiment was 
strictly comparative with that made with chloride of sodium 
just described ; the two experiments were indeed conducted 
side by side, and in precisely the same manner. The quantity 
of pure nitrate of sodium employed was 0'5588 gram (equal to 
519 lbs., or 80 lbs. of nitrogen, per acre), the exact chemical 
equivalent of the chloride of sodium used in the comparative 
experiment. The saturated soil was allowed to stand for a 
week after the application of the nitrate ; successive quantities of 
water were then placed on the surface, and the drainage-water 
obtained collected and analysed, yielding the following results : — 
Table XXIX. — Eesults of Percolation after Nitrate of Sodium 
had been applied to the Soil. 
Water put 
on. 
Grams. 
120 
120 
120 
120 
120 
120 
120 
120 
120 
1080 
Nitrogen as Nitrates and Nitrites 
Drainage in Drainage-Water, 
obtained. 
Grams. 
116-4 
118-7 
97-0 
134-0 
126-3 
120-4 
120-6 
117 2 
118-9 
Per Million. 
none 
none 
none 
none 
9-0 
57-3 
72-6 
20-0 
0-5 
Grams. 
1069-5 
none 
none 
none 
none 
0-00114 
0 • 00690 
0- (10876 
0-00234 
0-00006 
0-01920 
It was observed in making the third extract that the water 
had begun to percolate more slowly than in the experiment in 
progress at the side with chloride of sodium ; this resistance to 
the passage of water increased, so that by the fifth extract the 
percolation of 120 cubic centimetres of water occupied twice the 
time required in the experiment with chloride of sodium. The 
cause of this retardation was quite apparent ; large transverse 
cracks, filled with gas, had formed in the soil, the largest being 
two or three inches from the surface ; no such cracks appeared 
in the chloride of sodium percolator. 
On turning to the analyses of the drainage-water, we see that 
nitrates did not begin to appear until the fifth extract, that thej 
