30 THE ROYAL SOCIETY OF CANADA 
vessel. It was shaken and allowed to settle alternately for an hour, 
at the end of which time 50 cb.c. of the ether solution (20 cb.c4 as 
in the original directions, would have been too little) was withdrawn 
for titration. To this ethereal solution an excess of water was added 
and the whole was titrated with approximately N/5 sulphuric acid 
(1 cb.c. = 0-03382 gm. nicotine) using lacmoid as the indicator. 
Acid required 1-4 cb.c. = 0-04734 gramme nicotine = 0-1136 
gramme in total sample or 1-89%. 
Experiment 4, Kissling’s Method. For comparison of our own 
with this method, a sample of the same tobacco weighing 20 
grammes was mixed with 10 cb.c. of sodium hydroxide solution 
(6 grammes of sodium hydroxide in 40 cb.c. of water and 60 
cb.c. of alcohol) transferred to.a Soxhlet apparatus and extracted with 
ether for five hours. The ethereal solution was then evaporated at a 
gentle heat, the residue was mixed with 50 cb.c. of N. sodium hy- 
droxide solution and transferred to a distillation flask with the aid of 
water. After distilling in a current of steam till about 400 cb.c. of 
distillate had been collected the portion which still passed over was no 
longer alkaline. 
The whole distillate was titrated with sulphuric acid, using lac- 
moid as indicator, and required 10-3 cb.c. for neutralisation. Since 
1 cb.c. of acid = 0:03382 gramme of nicotine, this corresponds to 
0-3483 gramme of nicotine in the sample, or 1-74%. 
Discussion and Comparison of Methods 
It will be noticed that the two modifications of our method give 
results as nearly as possible identical with one another, but higher 
than those given by the methods of either Téth or Kissling. We have 
devoted careful consideration tothe possible errors of our own method. 
All of them tend, however, in the direction of giving a result that would 
be too low rather than too high, and we are left with a preference for 
our own procedure both on considerations of accuracy and ease of 
working. 
Our main objection to the method of Téth is that the original 
weight of the sample worked on is necessarily small; even when a 
large aliquot portion of the solution is titrated, only a small quantity 
of acid is needed, so that there is a correspondingly great chance of 
error. Had 0-25 cb.c. more acid been used in the experiment de- 
scribed above, the corresponding increase in the apparent amount of 
nicotine would have been 0:36%, bringing the Téth determination 
into exact agreement with our own. 
