DETERMINATION OF NICOTINE IN ToBACCO. 67 



off from the tobacco Avithoiit bringing the hitter ou the fiUer, and 

 further i)ortions are pre^^sed ont by means of a glass rod flattened at 

 one end. The fihration shoukl be carried out as rapidly as possible 

 in order to avoid loss from evaporation, and as soon as ail the extract 

 can be brought on the filter the funnel should be covered with a 

 watch glass. An aliquot portion of the filtrate — conveniently 75 

 c. c, corresponding to 4.5 grams of tobacco — is measured into a dry 

 separatory funnel, -which is left unstoppered and allowed to stand for 

 about an hour in order to remove the ammonia. As has already been 

 stated, no loss of nicotine is to ]>e feared by this treatment. Instead 

 of allowing the extract to stand in the funnel a slow current of air 

 may be drawn through it for five minutes by means of a stopper sup- 

 Dlied with inlet and outlet tubes, the latter of which is connected with 

 a filter pump. After the ammonia has been removed 10 c. c. of fifth- 

 normal sulphuric acid in about 50 c. c. of water are introduced into 

 the funnel and the contents thoroughly shaken. After draAving off 

 the aqueous layer the gasoline is washed twice with small volumes of 

 water and the excess of sulphuric acid titrated with tenth-normal 

 alkali, using cochineal as the indicator. 



This method of titrating the nicotine may be modified as follows: 

 The measured portion of the filtered extract is placed in a 250 c. c. 

 Erlenmeyer flask and allowed to stand an hour. The standard acid 

 is then run in and a convenient volume of water added. The contents 

 of the flask are thoroughly shaken, and after the two layers have 

 separated the greater portion of the ligroin is drawn off by means of 

 a pipette.<» The excess of acid is then titrated as before. One cubic 

 centimeter of fifth-normal sulphuric acid is equivalent to 0.0324 

 gram of nicotine. 



COMPARISON OF THE NEW METHOD WITH THAT OF KISSLING. 



As has already been stated, the only reliable method for the deter- 

 mination of nicotine heretofore available is that of Kissling. and we 

 have used this as the standard for comparison. The nicotine content 

 of a large number of samples. of tobacco of different types has been 

 determined both by the Kissling process and by the new method, and 

 the figures given in the following table are fairly representative of 

 the results obtained. 



It will be seen that the nicotine content in the different samples 

 included in the table varies from 5 to less than 1 per cent, and these 

 samples represent practically all classes of domestic cigar tobacco of 

 both w^rapper and filler types. Most of the samples had been fer- 



alt is necessai-y to remove the greater portion of tlie gasoline before titrating 

 because of the yellow coloring matter which it contains in solution. 



102 — VII 



