ON WATER ANALYSIS. 
379 
the tube leading to the reservoir, a. There is a little glass stopcock in the 
interval. The rest of the arrangement consists of r, a wide glass tube, held by 
a cork at the bottom, and filled with water ; this latter preserves the enclosed 
gas-tubes constantly at the temperature of the apartment. A thermometer, t , 
is immersed in the water. The whole apparatus is filled previous to use with 
mercury, by opening the stopcock in the capillary tube at the top and pouring 
it in at the open end, o, of the wide limb; when this is effected, the stopcock 
is again closed, and the gases collected from the combustion tube introduced 
into a . By now opening the tap, s.t., at bottom, the mercury can be drawn 
off to the requirtd level, a proceeding which causes the gas in a, when the little 
stopcock is opened, to pass over into the graduated tube, g. Here a reading 
of the volume is taken, at the same time noting the temperature and barometric 
pressure. The gas is then driven back into the reservoir, a, and the carbonic 
acid absorbed by throwing up a strong solution of caustic potash ; another 
reading of the volume is taken, and then a bubble of oxygen introduced for the 
determination of the quantity of nitric oxide present. The excess of oxygen is 
afterwards absorbed by pyrogallic acid. The usual corrections have to be made 
for barometric pressure, temperature, and also moisture, with which the gases 
are all saturated. The data thus obtained furnish all that is necessary for the 
calculation of the carbon and nitrogen in a water residue ; and the delicacy of 
the methods will be appreciated when it is understood that the nitrogen can 
be determined to within of a milligram, and of carbon a difference so small 
as rroioo °f a milligram can be detected with certainty. 
The following examples will serve to illustrate the kind of results usually 
obtained: and when the difficulty and uncertainty attending the weighing of 
such small quantities of material, as were operated upon, are considered, they 
appear to be very satisfactory. 
Sugar. Found. Calcd. 
Carbon . . . *01386 . . . *01482 grams. 
Carbon . . . *00440 . . . *00480 ,, 
Carbon . . . *00530 . . . *00516 ,, 
Sugar Mixed with Chloride of Ammonium. 
Found. Calcd. 
Carbon . . . *00434 . . . *00484 grams. 
Nitrogen . . . *00254 . . . *00246 ,, 
The process for the determination of the nitrogen a short time ago suggested 
by Messrs. Wauklyn, Chapman, and Smith,* was carefully tested in the course 
of these experiments; but it was almost invariably found to yield a smaller 
quantity of nitrogen than was obtained by the combustion method just de¬ 
scribed ; and, indeed, Mr. Wanklyn himself states that the whole of the nitrogen 
cannot be expelled except by carrying the distillation to dryness, and, in fact, 
submitting the residue to a kind of Yarrentrap process. 
The following is a comparison of a few of the numbers obtained by the per¬ 
manganate process and the combustion process, and it will be seen that there 
are, in several instances, great discrepancies :— 
Nitrogen in 100,000 Parts of Water. 
Permanganate. 
Combustion. 
Sample 1. . . *006 . 
. *010 
„ 2. . . *006 . 
. *011 
„ 3. . . *006 . 
. *010 
„ 4. . . *002 . 
. *010 
* See page 335. 
