AN IMPROVED method FOR THE DETERMINATION OF 
NICOTINE IN TOBACCO AND TOBACCO EXTRACTS ‘ 
By O. M. Shedd ^ 
Chemist, Kentucky Agricultural Experiment Station 
HISTORICAL REVIEW 
The two methods which are now more generally used in this country 
for the estimation of nicotine in tobacco and tobacco extracts are those of 
Kissling and the silicotungstic acid method as perfected by Chapin.^ 
They are the only ones for this determination which have been adopted 
as official by the Association of Official Agricultural Chemists.^ It is of 
interest to note in this connection that while Chapin’s procedure for 
extracts was adopted by the association, a material change appears in 
the association method. This change is in the amount and manner of 
adding the alkali preparatory to the distillation; it is opportune, even if 
made unwittingly, because it tends to increase the accuracy of the method 
when applied to tobacco. In this connection it is important that a 
sufficient quantity of alkali be added in the distillation to liberate com¬ 
bined nicotine. The association method should emphasize this point 
more clearly, since it is possible in following their directions in the dis¬ 
tillation to have the solution test alkaline due to free nicotine and yet 
an insufficient quantity of alkali may have been added to liberate that 
in combination. 
Chapin describes the various procedures which have been proposed for 
the estimation of nicotine and the defects he found in their application, so 
it is not necessary to mention them here. Unfortunately, however, 
Chapin, in his work on the silicotungstic acid method, did not include 
tobacco in his studies, otherwise he probably would have found that 
errors may often occur in the use of this method on tobacco, as will be 
mentioned later. The method as outlined by him was proposed for 
tobacco extracts and when so applied in the present work has been found 
to be more satisfactory than for tobacco itself. 
While it is generally recognized that the Kissling method will giv» 
good results in the hands of an experienced analyst, provided no inter¬ 
fering substances are present in the determination, it is nevertheless, 
conceded that the silicotungstic acid method will probably give more 
reliable and concordant results in ordinary use. 
After the work reported in this paper was practically completed, the 
writer learned for the first time, through a mutual acquaintance, that a 
similar method had been published by Rasmussen,^ in Denmark. As no 
reference to it could be found in the literature at hand, a copy was obtained 
J Accepted for publication Jan. 22, 1923. Published by permission of the Director of the Kentucky 
Agricultural Experiment Station. 
* The author desires to thank Dr. A. M. Peter, Head of the Department of Chemistry, for helpful criticisms 
offered during the progress of this investigation. 
3 Chapin, R. M. thb determination op nicotine in nicotine solutions and tobacco extracts. 
In U. S. Dept. Agr. Bur. Anim. Indus. Bui. 133, p. 21. 1911. 
* Assooation op Official Agricultural Chemists, official and tentative methods of ^^naly- 
SIS. AS COMPILED BY THE COMMITTEE ON REVISION OF METHODS. Revised tO NoV. I, I919. P- 65H36. 
Washington, D, C. 1920. 
6 Rasmussen, Hans Baggesgaard. om bestemmelse af nikotin. In K. Danske Vidensk. Selsk. 
Skr. Naturvidensk. ogMath. Afd., Ser. 8, Bd. 1 , p. 66. 1916. 
Journal of Agricultural Research, 
Washington, D. C. 
afb 
(961) 
Vol. XXIV, No. II 
June 16, 1923 
Key No. Ky.-i3 
