72 
nitric acid, or nitrous acid, by the action of chromic acid or 
sulphuric acid. A similar fact was discovered by Kjeldahl 
(Zeits. Anal. Chem., XXII. 366) where he describes the con- 
version into ammonia of nitrogenous matter, by boiling it 
with strong sulphuric acid, phosphorous pentoxide, and 
powdered manganate of potassium. 
Marker tested this method thoroughly (Zeits. Anal. Chem., 
XXIII. 553 — 557) against the well-known method of 
Yarrentrap and Will, and found the results by Kjeldahl’s 
method sufficiently correct. 
To determine the organic nitrogen in the water I take 
250 c.c. or more of the solution obtained by the previous 
organic carbon process (the solution made up to 1 litre), 
place it in the boiling-flask,’' pour down the funnel tube a 
perfectly- ammonia-free caustic soda solution in excess, and 
attach the exit-tube of the “ boiling-flask ” to the Liebig’s 
condenser and flask as used in the organic carbon determi- 
nation, placing, however, in this case, about 50 c.c. of am- 
monia-free water and a few drops of pure hydrochloric acid 
into the “receiving-flask.” It is also better to take the 
necessary precautions to prevent the sucking of the water 
in the “receiving-flask” back into the “boiling-flask,” I 
boil the contents of the flask for about 30 minutes (keeping 
the condenser cool), then I make up the condensed water in 
the “ receiving-flask ” to 1 litre, take out 100 c.c. and de- 
termine the amount of ammonia present in it in the usual 
way with Kessler’s reagent, and calculate how much nitro- 
gen it corresponds to. 
I do not lose any nitrogen by this method, because all 
the ammonia evolved from the “boiling-flask” is passed into 
cold acidulated water, whereas by the old “ ammonia- 
method,” the violent bumping in the retort often drives 
steam and no doubt ammonia right through the long con- 
densers used in that process, consequently there must be a 
loss of ammonia, 
