71 



()winu r t" the mineral acid present in the aqueous liquid, the presence of a slight 

 amount of thi> emulsion in the chloroform layer causes serious error in a titration of 



llli- IVMillle. 



Th.- refills obtained by this method, both by weighing and by titrating the chloro- 

 form n-idue, are given in the table below, under Method II. It will be observed 

 i luii in all caaea the results obtained by weighing are obviously too high, while the 

 results obtained by titration with this method are on the whole very satisfactory. It 

 would appear therefore that the method is fairly satisfactory for the determination 

 of ben/.oii- aid in tomato ketchup, which presumably offers as great difficulties for 

 thi< determination a.- any food to which this preservative is commonly added. 



Detenu ii'.itin,, o/.sW/,/m benzoate in ketchup by three methods. 



& 100 grains ketchup diluted with water to 200 cc, mixed, and filtered, 50 cc portions of filtrate, diluted to 

 100 cc, saturated with sodium chlorid. acidified, and extracted with four portions of 100 cc each of ether. 



e Same as f oxrept extraction is made with chloroform instead of ether, three portions being 



used of 100, 50, and 50 cc, respectively 



In some cases where the mixture of tomato ketchup and a saturated salt solution 

 was tillered soon after it was prepared, the amount of benzoic acid determined by 

 t h is met IKK! was too low. This suggests the possibility that the solution of the benzoic 

 arid in the saturated salt solution may not have been complete. More satisfactory 

 results were obtained in all cases where the mixture was allowed to stand overnight 

 before filtering. 



