166 
REPORT— 1854. 
Table I. 
Method 1.—The Rain twice distilled to }ths with a little caustic alkali, and 
the second distillate evaporated with sulphuric acid to the measure of 
1000 septems. (Year 1853.) 
Months. 
lbs. of rain 
distilled. 
Septems ol 
sulphuric 
acid at 10° 
added to 
second 
distillate. 
Septems of 
alkali at 10 
to neutralize 
100 septems 
of concen¬ 
trated acid 
product. 
Difference 
=septems 
ammonia a 
| 10° in 100 
septems acir 
product. 
Grains of 
ammonia In 
| rain-water 
■ taken. 
Ammonia 
per million 
of 
rain-water. 
March ... 
267-2 
300 | 
17-5 
17-5 
12-5 1 
12-5 J 
2-681 
1-433 
April . 
109-6 
400 j 
35-5 
35-5 
4-5 1 
4-5 f 
0-965 
0-812 
May . 
95-2 
300 | 
25-5 
25-5 
45 1 
4-5 j 
0-965 
1 448 
June . 
175-2 
700 | 
62-0 
620 
80 1 
80 / 
1-716 
1-399 
July . 
2180 
500 j 
43-5 
43-6 
6-5 1 
6-5 f 
1-394 
0 935 
August ... 
167-9 
400 
35-5 
4-5 
I 
0-965 
0-821 
It is seen that the smallest quantity of water submitted to distillation in 
this senes of experiments was about 95 lbs., and that the largest was about 
i 7 Recond d >»ti|late of 4ths of the amount nut into a retort of 
g ass with a little caustic alkali, the quantities of sulphuric acid at 10 g |Ven 
c 1 0 !" , ' m 3 worp “dde'l* and the whole was then evaporated to the measure 
of 10 °0 septems at 60°. One-tenth of this, or 100 septems, after being 
coloured by a given measure of infusion of litmus carefully freed from excess 
of al kah was then tested with n caustic alkali solution, also at 10°. It*» 
invariably round, that, with liquids of this degree of concentration, the deter- 
munition could be made to a single drop of the test alkali. This would g' vc 
a maximum range of error of about T ' n th, or 2 per cent, of the whole am- 
C °" tumed !" water, dependent on the manipulation of the test 
1 , ? us stllla luestion whether there might not be some source 
would “T"'"* with M - iioussingault that the whole of the arninon'S 
vouhl b e obtained in a «,| 1S distillate, or in supposing that there would be JO 
loss either of ammoma or of acid, in the evaporation of the acidulated <h- 
■ e ,n ?P en Ves5e J B - Ry the comparison which will presently be » ia Ll 
however, between the above determinations by Method I, and others in 
specimens of the same waters by Method 2. which in fact is that adopted by 
* “ Uil1 ■**" tl>at there was probably no error due to 
lame dU, aUSl:S *l USt »n fact we conceive rhat the method of 
stilUte tn ] at, -° n * " 1h fil| t' #< ’quent acidulation and concentration ot the 
Sllllate to a riven a..-, , , .fmn? test 
rnndnet.no d;..;n .• ^convenience, and irequeni b 
whkV'is obtained by MettoJ*; 
Which is substantially that of \I. Boussinga.dt. M. Boussingaults pr“;«* 
in submitting to a single distillation, generally not more than 1 
Which is 
consists in 
