65 



would interfere with the reaction by which the salicylic acid should finally be esti- 

 mated in the iv.-idiir. Owing to its rapidity and convenience, the ferric chlorid 

 reaction has usually been employed for determining the amount of salicylic acid 

 :it. It is, therefore, important . that the solvent employed should not extract 

 tannin from the food. 



lu order to determine what solvents should be most advantageously employed as 

 far a- delicacy "f reaction and freedom from tannin or other interfering bodies is con- 

 cerned, Mr. i 'harle- S. A-h extracted 50 cc portions of claret containing salicylic acid 

 in amount.- varying from 0. <)_>:> mg to 0.5 mg and treated the residue obtained by 

 evaporation of tin- solvent with ferric chlorid in the usual manner. The results are 

 given in the following table: 



Comparative efficiency of solvents on salicylic acid dissolved in 50 cc of claret (Ash}. 



Mr. A-h found that the tir-t live .-olvrnts u'iven in the table extract tannin in the 



in which they are im>nti'ne<l -that i-, ether extracts the greatest amount and 



tn. hl.. r acetylene the least. The last three solvents carbon bisulphid, carbon 



hloriil. and toluene do not extract tannin, and the ferric chlorid reaction in the 



re-iduo obtained by them from wine is clear and characteristic of ferric salicylatc. 



It will be Q0ted that the residue from the ether extraction gave no reaction whatever 



witli ferric chlorid. This was due to the presence of tannin, which entirely obscured 



the reaction. With chloroform much better results were obtained, but even here the 



reaction was partially ol>scured by tannin, which was also true of the residue from 



dichlor-acetylene. 



The data given in the column headed " Kther and hydrogen peroxid" were deter- 

 mined by oxidi/.in- with ammoniaand hydrogen peroxid the residue obtained by evap- 

 orating the eth- troys the tannin and also partially converts ben/o- 

 nd -accharin when present into salic\ lie acid. The salicylic acid was then again 

 tel with ether, the ether extra- t evaporated, and the residue tested with ferric 

 id. As previoii.-ly -tated, the greatest freedom from interfering substances 

 attended the use of carbon tetruchlorid and toluene, the latter appearing to extract 

 slightly !nre -alicylic acid than the former, and thus affording a better test in the 

 l-r.-ence of a small amount of that substance. Chloroform was also very satisfactory, 

 being inferior to carbon tetrachlorid and toluene in respect of dissolving interfering 

 Ubfltancee, though apparently slightly superior in the amount of salicylic acid 

 extracted. 



T::;T:J r.uii. I-JL- 



