70 



If the volume of salt solution taken in this work is equivalent to 100 grams of food 

 extracted as in the case of the subsequent studies, the error caused by the extraction 

 of interfering bodies by means of chloroform is less than 0.001 per cent, and is negligi- 

 ble. It would appear, therefore, that chloroform possesses the following advantages: 

 First, after saturating a solution with sodium chlorid a practically complete extraction 

 of benzoic acid may be made; second, the benzoic acid extracted is not accompanied 

 by other bodies that interfere with its determination by means of titration; third, 

 the solvent i not combustible; fourth, it is inexpensive. In connection with subse- 

 quent work on the determination of benzoic acid, therefore, no study was made of 

 any other solvent than chloroform except as described under Method IV. The detail 

 of the methods studied is as follows: 



LA WALL AND BRADSHAW METHOD (METHOD l). 



This method was described in full by the authors in the American Journal of Phar- 

 macy (volume 80, pages 171-172). The principle depends upon the method outlined 

 by F. X. Moerk in an article published in the Proceedings of the Pennsylvania Phar- 

 maceutical Association for 1905, page 181. The details of the method are as follows: 



To 20 grams of the substance under examination add 2 grams of sodium chlorid, 5 cc 

 of hydrochloric acid, and 25 cc of a saturated solution of sodium chlorid. Shake 

 thoroughly for five minutes, transfer to a moistened filter, and wash with a saturated solu- 

 tion of sodium chlorid until 100 cc of the filtrate are collected. Transfer the filtrate 

 to a separatory funnel and shake with three portions of chloroform, using 25, 15, and 

 10 cc respectively. Evaporate the chloroform at room temperature. If the residue 

 is white and crystalline, dry over sulphuric acid in a desiccator and weigh. If yel- 

 lowish and oily, dissolve in 10 or 15 cc of weak ammonia acidified with dilute sul- 

 phuric acid and again extract with chloroform. The residue is dissolved in 3 to 5 cc 

 of neutral alkali and titrated with twentieth-normal alkali solution, using phenol- 

 phthaleiii as indicator. The titration should agree closely with the gravimetric deter- 

 mination, the difference being rarely more than 1 or 2 mg. 



As will be seen by a comparison of the figures obtained by various methods given 

 in the table on page 71, the results by weighing the benzoic acid, given under Method I, 

 are much higher than those obtained by titration. All collaborators report difficulty 

 in securing an adequate filtration. F. W. Heyl obtains very satisfactory results by 

 titration, whereas the figures of all other analysts were low. In the Division of Foods 

 the method was applied to the examination of a considerable number of commercial 

 ketchups. With the low-grade product, made from skin and core pulp, filtration was 

 possible though somewhat slow. With many of the high-grade ketchups the material 

 was so finely divided that filtration was almost impossible. In this connection should 

 be considered the limited solubility of benzoic acid in a saturated solution of sodium 

 chlorid to be mentioned subsequently. 



MODIFICATION OF LA WALL AND BRADSHAW METHOD (METHOD II). 



This method is identical with the one just described, except that no attempt is 

 made to obtain a complete filtration. 



To 200 grams of ketchup are added 20 grams of finely powdered sodium chlorid in 

 a liter flask, and enough of a saturated solution of sodium chlorid is added to make a 

 liter. The contents of the flask are thoroughly mixed and allowed to stand overnight, 

 when they are filtered, and 500 cc of the filtrate are transferred to a separatory funnel, 

 treated with 5 cc of sulphuric acid (1 to 5) and extracted repeatedly with chloroform, 

 using 100, 50, 50, and 25 cc of chloroform, respectively. 



In this extraction a troublesome emulsion was formed which it was found could be 

 broken up to a considerable extent by centrifuging and by stirring with a glass rod. 

 Where a centrifuge is not available much may be accomplished by swinging the sep- 

 aratory funnel with the hand. It is important that after each extraction the chloro- 

 form be removed as completely as possible, and at the same time the utmost care 

 must be exercised to prevent any emulsion passing through with the chloroform. 



