380 
ON WATER ANALYSIS. 
Permanganate. Combustion. 
Sample 5. . . '016 . . '068 
„ 6. . . -016 . . -042 
,, 7 . . . -001 . . -012 
„ 8. . . -022 . • -076 
The next operation in water-testing consists of the estimation of the nitrates 
and nitrites. These very important constituents cannot be determined in water 
by any of the processes ordinarily in use, and in studying the question the 
author was led to the employment of a reaction, which is the foundation of a 
process many years ago described by Mr. Walter Crum. This convenient me¬ 
thod, which has been adopted with very satisfactory results, is based upon the 
fact that the nitrogen may be completely expelled in the form of nitric oxide, 
and that this gas may be very accurately measured. The fluid obtained by 
concentrating half a litre of water to a very small bulk serves the purpose. It 
is necessary, however, before proceeding to the analysis, to get rid of the chlo¬ 
rides, and this is effected by adding a solution of sulphate of silver in sufficient 
quantity, and filtering off 1 the precipitated chloride of silver. 
For the decomposition of the nitrates and nitrites and the collection of the 
evolved gas, a tube is employed similar to that represented in the drawing. It 
consists of a tube about eight inches long, constricted about an inch 
from the top into a narrow neck into which a little stopcock is care¬ 
fully ground ; it must be well annealed, or it will crack with the heat 
evolved during the chemical action which takes place in the opera¬ 
tion. This is filled with mercury to the stopper and inverted into a 
mercurial-trough ; the residue of the water is then poured into the 
little cup at the top, and the stopcock turned so as just to admit the 
liquid without any air. The beaker which previously held the con¬ 
centrated liquid is then carefully rinsed in with successive small 
quantities of water, and lastly, a small quantity of strong sulphuric 
acid is admitted in the same way, taking care that no air is suffered 
to enter. The reaction does not commence immediately; but, on 
closing the lower extremity of the tube with the thumb, and gently 
agitating the whole, taking the precaution to keep the stoppered end 
upwards and not to bring the acid liquid into contact with the skin, 
a strong pressure is presently felt from within, and this is gradually 
relieved by letting out a little mercury at the bottom. After a short 
time it will be found that an accumulation of gas has occurred, and 
when the action is over, which happens in about two hours, this is 
transferred to the reservoir a of the gas-apparatus already described, 
wdience it is conveyed to the graduated limb of the tube and the 
volume is noted. 
In evaporating w T aters which are undergoing analysis another point 
has presented itself. The exceeding difficulty of excluding ammo- 
niacal fumes, dust, and various other sources of contamination, sug¬ 
gested the possibility and desirability, if possible, of conducting this 
operation in vacuo. The plan is easy of execution. 
The waters are placed in the small glass vessels before adverted to, which are 
made to stand on a tin plate under a large receiver fitting on to a metal trough 
connected with the air-pump. Exhaustion is then effected and evaporation 
goes on pretty rapidly under the influence of heat, which is radiated from above 
down upon the surface of the liquid in the several basins. In order to accom¬ 
plish this, the receiver, which encloses the samples of water, is covered by a 
metal pan inverted over it, and at the side is placed a small Bunsen flame, 
•which must be properly regulated. Under these circumstances it might per¬ 
haps be expected that the water would boil, but, inconsequence of the evapora- 
