34 
On the Rain and Drainage - Waters at Rothamsted. 
on October 23, 1881, in a first running after a cessation of 
drainagre for six months. That a small amount of diffusion of 
soluble salts does take place, whereby the unmanured plots are 
somewhat enriched (diffusion is always from the stronger to the 
weaker solution) is probably true, but it is apparently limited 
to times of rest. So long as water is passing from above down- 
wards, the manured or unmanured condition of the surface soil 
is sharply represented in the drainage-waters ; but when active 
drainage ceases, a very slow general diffusion will take place 
if the soil continues suflficiently wet, the result being that the 
richest plots lose and the poorer plots gain somewhat in soluble 
salts. 
The farmyard-manure plot shows but a small quantity of 
chlorine in its scanty runnings, the amount, 6*1 per million, 
exceeding by very little the quantity in the drainage from the 
unmanured land. 
Turning next to the average amount of chlorine from plots 
receiving chloride of ammonium, we may note that the dif- 
ferences between Plots 6, 7, and 8 are such as we should expect 
to exist in the case of plots receiving respectively 100 lbs., 
200 lbs., and 300 lbs. of a chloride in the manure. If we 
deduct the chlorine found for Plot 5 from that yielded hy 
Plot 6, we obtain 8*8 as the increase of the chlorine due to 
100 lbs. of chloride of ammonium. Adding this figure to the 
chlorine found for Plot 6, we obtain 22*8 as the amount of 
chlorine proper for Plot 7 ; the figure actually found being 23'1. 
Adding again 8'8 to 22*8, we obtain 31"G as the amount of 
chlorine proper for Plot 8, the figure actually found being 30'8. 
The chlorine in the drainage-waters thus corresponds fairh' 
with the chlorine applied in the manures, and we may conse- 
quently assume that Plots 5 to 8 are really comparable in their 
amounts of drainage, though varying a good deal in the amount 
of visible discharge at the pipes. 
When we turn to the other plots receiving chlorides the 
results are not so agreeable. Plot 10, indeed, receiving the 
same amount of chlorides as Plot 7, gives also a very similar 
proportion in its drainage-water ; but Plots 11, 12, 13, 14, 15, 
and 17, though receiving the same quantity of chlorides in their 
manure, yield, all of them, a considerably larger proportion in 
the drainage-water, the quantity of chlorine rising on Plot 13 to 
28-9, and on Plot 15 to 32-5 per million. Plot 13, which 
receives the same amount of chlorides, and at the same time as 
Plot 7, thus yields one-fourth more chlorine in its drainage- 
water. To what is this to be attributed ? On turning: to the 
chlorine determinations of Voelcker and Frankland (Table 
