62 PROCEEDINGS OF THE AMERICAN ACADEMY 



from methyl alcohol containing different amounts of acetone. Their 

 results seemed to sliow that the yield was very considerably dimin- 

 ished if the quantity of acetone exceeded one per cent. Since crude 

 wood spirit always contains a much higher percentage, it must be 

 specially purified for the preparation of dimethylaniline. 



Various methods, dependent either upon the physical or the chemical 

 properties of methyl alcohol, have been proposed for testing its purity. 

 Krell * suggested a determination of the amount of methyl iodide 

 which could be obtained by treating a given quantity of the alcohol 

 with phosphorous iodide. In the hands of a skilful analyst the per- 

 centage of methyl alcohol may be accurately determined by this 

 method ; but, as Kramer remarked, it does not account for the nature 

 of the impurities, and it is not sufficiently expeditious. To overcome 

 these objections, Kramer f took advantage of the formation of iodo- 

 form from acetone. In his modification of Lieben's reaction, J the for- 

 mation of iodoform from ethyl alcohol and acetic acid is avoided, and 

 acetic acid is only partially decomposed. 



According to Kramer's method a mixture of 10 c.c. of a normal 

 sodic hydrate solution, 5 c.c. of a double-normal iodine solution, and 

 1 c.c. of the alcohol to be tested, is shaken in a graduated cylinder with 

 10 c.c. of ether. 5 c.c. of the ether solution are withdrawn with a 

 pipette, evaporated on a weighed watch-glass, and the residue we'ghed 

 after drying in a warm desiccator. To obtain the weight of acetone 

 in Ic.c. of the alcohol, the weight of iodoform is multiplied by ,28 if 

 there are 9.5 c.c. of the ether solution. 



The percentage of acetone in several fractions, as determined by this 

 method, is given in the following table.§ Since in each determination 

 from 9 c.c. of the ether solution 5 c.c. were evaporated, the weight of 

 iodoform was multiplied by .27 for the weight of acetone. 



* Berichte deutsch. chem. Gesellsch., 1873, 1310. 



t Ibid., 1880, 1000. 



t Ann. Cliem. Pliarm., Suppl. VII., ccxviii. 377. 



§ In several instances the fraction was boiled with silver oxide before test- 

 ing for acetone ; but this seemed to have no appreciable effect on the quantity 

 of acetone indicated by the iodoform reaction. 



