Vol.XXIV, No II 
964 
Journal of Agricultural Research 
The extract is shaken with four to six portions (10 cc. each) of cold 
dilute hydrochloric acid (i to 4). If an emulsion forms during the 
successive extractions with the acid, this can be prevented by adding 
a small amount of 95 per cent alcohol, about i cc. being sufficient for 
each separate extraction. (When it is found necessary to use alcohol, 
it is essential that it be added after the shaking has been completed. 
Further shaking should be avoided as much as possible, as one or two 
inversions of the funnel will break the emulsion. A minimum amount 
of alcohol and shaking are best, since it has been found that when 
alcohol is added along with the acid, followed by vigorous shaking, it 
has some solvent action on extraneous material that precipitates with 
the reagent. Consequently, alcohol should not be used unless necessary, 
and only in the manner described.) The successive extractions are run 
through a funnel into a 100 cc. graduated flask. A small ball of fine 
glass wool may be placed in the neck of the funnel. The extraction 
should be continued until a few drops of the acid solution give no tur¬ 
bidity with tlie reagent. The funnel is carefully washed with water and 
the volume made to 100 cc. An aliquot corresponding to about 0.25 
gm. or more of extract or i to 2 gm. of tobacco, depending on the amount 
of nicotine present, is placed in a beaker and diluted to 400 to 500 cc. 
volume with distilled water. A sufficient quantity of dilute hydrochloric 
acid (i to 4) is then added so as to have at least 3 cc. or more present, 
to each 100 cc. of liquid, including the quantity of acid contained in the 
aliquot. At this point, the solution should test distinctly acid with a 
few drops of methyl orange; if not acid, more should be added. There is 
then added, with constant stirring, i cc. of the silicotungstic acid reagent 
for each 0.01 gm. of nicotine supposed to be present. Stir thoroughly 
for about 3 to 5 minutes or until the precipitate is crystalline and settles 
quickly. To insure an excess of the reagent, a few drops are added to 
the clear supernatant liquid and the absence of any precipitate noted. 
The solution is allowed to stand at least 18 or preferably about 24 hours. 
The precipitate is filtered on an ashless filter (C. S. & S. No. 590 or similar 
quality) and preferably through a Hirsch or Buchner funnel by decanting 
first the clear liquid and testing a few drops of it with a little nicotine 
solution to insure an excess of reagent. By decanting practically all the 
clear solution the filtration is more quickly made. (In case of very small 
precipitates, the addition of a small amount of ashless filter paper pulp 
to the precipitate at the time it is to be filtered will insure a clear filtrate.) 
When the precipitate is transferred to the filter, if the filtrate should be 
turbid, it is filtered again until clear. If, at the beginning, the precipi¬ 
tate is made crystalline by proper stirring, the filtration is made quickly 
and without trouble. The precipitate is thoroughly washed with cold 
dilute hydrochloric acid (i to 1,000). Transfer the precipitate, without 
removing from the paper, to a weighed platinum crucible, dry carefully 
at low heat until the paper is charred, then ignite over a Bunsen burner 
until all carbon is eliminated. Finally, heat over a Teclu or Meker 
burner for not over 10 minutes, or over a moderate blast for about 5 
minutes. It is not advisable to prolong the final heating beyond the 
time mentioned, because of a very small but continued loss at this 
temperature. The weight of the residue multiplied by 0.114 gives the 
weight of nicotine present in the aliquot used. 
In the development of this method, the following experiments were 
made to test important features of the work: 
