» 
AMMONIA AND NITRIC ACID IN RAIN-WATER. 
171 
error might be between *th and |th, or about 16*5 per cent of the whole. 
In spring-water, again, M. Boussingault found genera y ) * . j 
much ammonia as in rivers; in which therefore the smallest errorin the de 
termination in 1 litre only, would amount to about s id o ie " >0 1 •_ 
According to Method S again, assuming that ltMbs. of 1 ntor (-about 
litres) were used, this, taking the average composition ut the Uothumstea 
fains, would contain about 54 milUgr. of ammonia; ami determining as wc 
calculated to do, to 0034 mifligr., the minimum error would thus be About 
r^th, or about 0*> per cent, of the whole. The probability of accuracy 
therefore, so far as the process of liquid testing w roucerm'd. would be 
greatly in favour of the third method, that of successive distillations. 
At the foot of Table III., however, is given the determination oi ammonia in 
‘he rain of April 1851, according to Method % which indicates by the method 
of small single distillation 098 ammonia per million of water, by the Hide ot 
W67, as obtained by the method of successive distillations; figures which 
filicide sufficiently to give confidence at least in the approximative truth ol 
* e results obtained by the two methods. This brings us to a further con- 
wderation of the proportions of ammonia found in the rains collected at 
nothamsted by the different methods, and of the actual amounts contained in 
•he water which fell over a given area in different seasons; and also to a com- 
Prison of our own results with those obtained by other experimenters in 
oth «r localities. 
Ante IV.—Showing the proportion of ammonia per million rain-water, 
and the lbs. of ammonia in the monthly rain per acre, in different locali- 
hosand seasons, determined by different experimenters, and by different 
methods. 
Ammonia per million rain-water. 
lbs. of ammonia per acre. 
Months. 
Paris. 
Leib- 
fraun- 
berg. 
Rothamated. 
Paris. 
Rotbamsted. 
1851. 
1852. 
1853. 
1854. 
1851. 
1853. 
1854. 
February 
March ' 
April, 
.' 
'One.. 
I illy 
Aujn st 
sys 
^ctober 
2«mber 
Mean 
Barral. 
loussin- 
gault. 
Moth. 1. 
Ueth. 2. 
Meth. 3. 
Barral. 
Meth. l.| 
Meth. 2. 
Meth. 3. 
0-779 
0-945 
0-32 
0-20 
1-440 
0-812 
1-448 
1-399 
0-935 
0-945 
077 
Oil 
0624 
0-445 
0-435 
0-817 
1-225 
1-144 
0-939 
0-851 
0-967* 
0-469 
055 
0-55 
107 
095 
0-55 
0-47 
0-88 
096 
0 11 
0-46 
298 
1-494 
0-821 
075 
0-55 
0-57 
3-84 
0-745 
0-785 
0688 
0-808 
1614 
1-00 
.i„, a 
0 33 
1 os 
0-721 
0-51 
0-57 
117 
0-45 
1*34 
! 0-37 
8-38 
1 0-15 
•! 340 
0-744 
1142 
0-979 
0-821 
081 
0-74 
054 
024 
By Method 2, 0 980. 
