32 , 071 the Bain and Drainage - Waters at Rotlianisted. 
year's drainage, immediately before the fresh application of 
manure. The minimum for Plot 15 occurs at the end of 
autumn, and that for the other plots receiving ammonia at the 
end of winter. 
The amount of chlorine present in the drainage-water at the 
period of minimum varies much, according to the amount of 
washing out by rain during the preceding year. The average 
amount of chlorine per million of water for seven plots receiving 
200 lbs. of chloride of ammonium was 10'3 on February 16, 
1879 ; 204, for the mixed runnings of February IG, 17, 
19, 1880; and 8'9 for the mixed runnings of March 5, 6, 7, 
1881. 
The amount of chlorine present in the drainage-waters at the 
different seasons is extremely variable. Thus, on October 1, 
1879, the average amount of chlorine in the drainage-water of 
six plots was 17"4 per million. On October 7, 1880, at 2 P.M., 
the water from the same six plots, running at about the same 
rate, contained an average of 331 per million. On October 
14, 1881, four of the same drains ran, the water giving an 
average of G7'3 of chlorine per million. In these instances the 
strength of the water entirely depends on the amount of washing 
out the soil has previously suffered. 
Where dry weather has occurred after the application of the 
chlorides they may remain for many months above the level of 
the drains, and when a running of the drains occurs, the water 
may be as rich in chlorides as if they had been quite recently 
applied. Illustrations of this will be found in the runnings of 
September 14, 1880, of February 16, 1880 (Plot 15), and of 
August 30, 1881 ; see Table XL VI. 
On the other hand, wet weather may greatly diminish the 
amount of chlorides in the soil, even when the drains do not run. 
Thus, Plots 7, 10, 11, 12, and 13 ran in August 1881, yielding 
waters rich in chlorides ; they ran also in September and in the 
earlier part of October. Plots 8, 14, and 18 did not run before 
October 23, but instead of yielding strong waters as the other 
plots did on their first running, they yield waters which are 
quite comparable with those of the now partially washed out 
Plots 7, 10, 11, 12, and 13. The loss of soluble salts suffered 
by the soil may thus be greatly in excess of the actual pipe- 
drainage. 
There is some indication that when a period of severe 
washing out is followed by a considerable interval of dry 
weather, the drains on again commencing to run will yield 
water stronger than what they yielded on leaving off : compare 
the runnings of August 28, and October 1, 1879, with 
those of February 16, 1880; and the last runnings of Plots 
3&4, 5, 9, 15, 16, 17, 19, in March 1881, with their first 
