Lwalker] physico-chemical RESEARCHES 31 



alcoholic solution. For example, when equimolecular quantities of 

 methyl bromide and ammonia were thus allowed to interact at 100° C, 

 the crv'stalline deposit obtained when the reaction was complete had 

 the composition CHglsrHaBr. Without further investigation, he pro- 

 nounced it to be the pure hydrobromide of methylamine. But, since 

 none of the five salts which might possibly have been produced are 

 soluble in benzene, the crystalline mass, however complex, could have 

 no other percentage composition than that represented by the above 

 formula. A similar error is found in many of the other cases described 

 in the communication. Experimentally we have found working under 

 approximately the same conditions that, although the solvent does have 

 an effect upon the nature of the product of the reaction, this product 

 is still far from simple, since in every instance examined ammonium 

 bromide was present in the reaction mixture. This can only be the 

 case if an amine higher than the primary is formed. The only instance 

 in which the reaction seems to yield a single alkylated ammonium salt 

 is tliat between ammonia and methyl iodide in which ammonium 

 iodide and tetramethyl-ammonium iodide are the sole products. A 

 d}Tiamical study of this reaction in alcholic solution showed it to be one 

 of the second order. This is in agreement with the observation of Mens- 

 chutkin, that the velocity of the action of a methjd halide on the amines 

 is vastly greater than on ammonia. 



