DETECTION AND DETERMINATION OF ALCOHOLS, ETC. 5 



however, that there are many mixtures from which the alcohol can 

 not be separated by the general method here given, and such special 

 cases must be treated individually. After the alcohol is separated 

 by some method applicable to the mixture in hand the general 

 methods herein described may be applied. They have been used in 

 this laboratory by H. C. Gore in following the decomposition changes 

 in bananas, with quite satisfactory results. 



METHYL ALCOHOL. 



A convenient and reliable method of detecting methyl alcohol is 

 that of Mulliken'and Scudder, 1 in which the alcohol is oxidized to 

 formaldehyde by means of a spiral of oxidized copper wire. A careful 

 study of this method showed that to obtain a positive test it is neces- 

 sary to have a solution containing over 0.1 per cent of methyl alcohol. 

 It seemed that greater delicacy would be obtained in such a test by 

 conducting the oxidation with certain substances in solution instead 

 of using the solid copper oxid in wire form as proposed by Mulliken. 2 

 A. Vorisek 3 proposed such a method, but as he did not determine its 

 delicacy and the apparatus used by him seemed cumbersome, the 

 subject was further studied. The method proposed by the writer is 

 as follows: 



To 100 cc of aqueous methyl alcohol in a 200 cc distilling flask add 

 from 5 to 8 grams of chromic acid; collect 10 cc of distillate and test 

 for formaldehyde by the Leach or Hehner methods. 4 If other oxidiz- 

 able substances, such as acids, sugars, starches, proteids, etc., are 

 present the solution under examination is exactly neutralized with 

 sodium hydroxid and one-third of it is distilled, the distillate being 

 oxidized and redistilled as has just been described. The use 

 of the relatively large quantity of chromic acid indicated is found to 

 increase the delicacy of the reaction. By the proposed method it is 

 very easy to detect one part of methyl alcohol in 100,000 parts of 

 water, and by making one preliminary distillation to concentrate the 

 methyl alcohol, collecting the first 100 cc from a liter and then pro- 

 ceeding as before described, a very strong test was obtained from a 

 solution containing only one part of methyl alcohol per million. 



When testing for methyl alcohol in the presence of strong ethyl 

 alcohol it is advisable to dilute the alcohol with water to obtain 

 approximately a 20 per cent solution, as otherwise the action of the 

 chromic acid on the ethyl alcohol may become explosively violent. 

 It is easy by this method to detect 0.1 per cent of methyl alcohol in 

 80 per cent alcohol, and by a little fractionation it was found possible 



i J. Amer. Chem. Soc., 1899, 21: 266. 



Loc. cit. 



J. Soc. Chem. Ind., 1909, 28: 823. 



* U. S. Dept. Agr., Bureau of Chemistry Bui. 107, Revised, p. 185; Analyst, 1S95, 20: 155. 



