THE SOURCES OF THE NITROGEN OF VEGETATION, ETC. 545 
Preparation of the titrated solutions. 
A weighed quantity of pure, dry carbonate of soda was dissolved in water, and to the 
solution water added to a given volume. As a preliminary step, the strength of some 
dilute sulphuric acid was tested against a given volume of the carbonate-of-soda solution; 
and from the data thus obtained, by further dilution a large quantity of acid was made 
of about the strength desired. The exact value of this acid was then ascertained by 
repeated trials with the standard carbonate-of-soda solution. To accomplish this, a 
given volume of the soda-solution was put into a beaker, a little litmus added, 
and the mixture heated over a spirit-lamp. The acid to be tested was then allowed 
to flow from a burette until a wine-red colour (indicating that the carbonate is converted 
into sulphate and bicarbonate with carbonic acid in solution) was produced. On boil- 
ing, the blue colour is restored ; acid is added until red ; the boiling is repeated, till the 
blue returns; acid again added, and so on, until the solution remains red on the addi- 
tion of the last drop. The point at which the permanent change takes place in the 
first trial being known, the experiment is easily repeated so as to ensure great accuracy. 
Thus, 50 septems of a solution of carbonate of soda, of which 1000 septems contained 
6-652 grammes of the salt, required, for neutralization as above, the following number 
of septems of the dilute acid, in six different trials— 
58°3, 58:2, 58°3, 58°38, 58:2, 58:2; mean 58°25. 
Hence— 
6652.50 N 6652, 50 — 14 
1000 * 58-25 * NaO, CO, 1000 “5825 “52-98 
=0:001508 gramme N. 
The mean of six experiments with a solution of carbonate of soda of another strength 
gave in the same way 0:0015008 gramme N; and we adopted the mean, or 0:001504 
gramme, as the amount corresponding to one septem of the titrated acid. 
It remained to prepare an alkaline solution to test against this standard acid. At first 
a solution of sugar-lime was employed; but this being found to be liable to constant 
change, due doubtless to fermentation, a solution of caustic soda was had recourse to. 
This solution was prepared of such dilution that the extreme error possible in reading 
off a unit of volume on the burette should be much less than would be admissible as 
the maximum error of analysis. The burette was of small enough diameter to allow of 
one-tenth of a septem being read off on it; and the alkali-solution was so dilute that it 
required about three septems of it to neutralize one septem of the titrated acid. Hence 
one septem of the alkali-solution corresponded to only about one-half of a milligramme 
of nitrogen, and the probable error of reading would therefore amount to only about 
one-twentieth of a milligramme. 
In the case of the sugar-lime solution, it was found necessary to test its strength 
against that of the acid every day that it was employed. But the soda-solution, if pro- 
perly prepared, and well preserved, remained for months unchanged; so that, when its 
value was once established against that of the standard acid, it could be expressed 
42 
