1916] 
DAVIS—FOLIN MICRO-KJELDAHL APPARATUS 409 
solid), then 50 ce. of ammonia-free water carefully added. 
Folin removed the nitrogen by adding saturated NaOH to 
alkalinity, then forcing the ammonia over into standard 
acid with a vigorous air current. While the method is excel- 
lent with the small amount of water used in the Jena tube, 
it does not give good results with the Kjeldahl flask and the 
larger amount of water used there. Distillation is more 
efficient. 
Small condenser tubes were made in the laboratory from 
glass tubing, the outer jacket measuring 402 сш., and Ше 
inner 5 mm. in diameter. The lower end of this latter, where 
it dipped into the collection acid, was fitted with a larger tube 
14 mm. in diameter—this to prevent back-flow of the acid; to 
the upper end of the inner tube was attached a safety trap 
made from a 10-сс. pipette, which, in turn, fitted into the 
Kjeldahl flask by means of a two-hole rubber stopper. 
Through the second hole of this was inserted a small piece of 
glass tubing closed at the upper end with a bit of rubber tub- 
ing and a pinch clamp, thus making it possible to add the 
alkali after the apparatus had been connected up for distilla- 
tion. The distillation is carried on in the usual way. It is com- 
monly necessary to add a pinch of zine dust to the distilling 
mixture to prevent bumping, while a few drops of liquid par- 
affin will keep down a tendency to froth. The ammonia is col- 
lected in N/20 acid and titrated against alkali of the same 
strength. Alizarin red (Alizarin sulfonsáure Natrium, 
Merck) in .1 per cent aqueous solution gave best satisfaction 
as an indicator. 
Folin has chiefly employed the colorimeter for the actual 
determination of nitrogen. The method has its distinct ad- 
vantages, especially if the precautions indieated by Folin are 
observed in Nesslerizing. In the absence of a colorimeter, 
however, and because excellent results were always obtained 
in our work by titration, the latter method has been retained. 
The following tables show how the results obtained with 
the ‘‘micro’’-method approximate very closely those gotten 
with larger amounts of material in the original Kjeldahl. The 
ilustration is that of an ordinary laboratory experiment 
