312 



PROCEEDINGS OF THE AMERICAN ACADEMY 



TABLE IV. — Experiments with Alcohol and Benzol. 



An examination of Tables III. and IV. shows that the percent- 

 ages of nitrogen obtained with the ethylate are the only ones which 

 agree among themselves, the numbers with other alcoholates often 

 differing by about six per cent, in one case by nearly ten per cent. 

 Under these circumstances it is obviously unwise to attempt to 

 draw any definite inferences from these results; we therefore dis- 

 miss them with the general remark that the amount of nitrogen 

 removed seems to show a tendency to diminish as the molecular 

 weight of the alcohol increases. 



The length of time during which the mixture of tribromtrinitro- 

 benzol and the alcoholate stood seemed to have but little influence 

 on the reaction, as is shown by Experiments V., XXIV., XXVII. , 



