OF ARTS AND SCIENCES. 



207 



Before considering the inferences which can be drawn from these 

 results, it is necessary to examine the sources of error which affect 

 the accuracy of the process ; and, for this purpose, all the results 

 obtained under the same conditions are collected in Table II. 



TABLE II. 



Toluol at 100°. 



No. of Experi- 

 ment. 



I. 

 II. 



Time in 

 Minutes, 



G5 

 20 



5 grs. used. 

 Per cent. 



72 

 54 



Toluol at 120°. 



No of Experi- 

 ment. 



VI. 

 VII. 



Time in 

 Minutes. 



20 



20 



10 grs. used. 

 Per cent. 



87 

 74 



Toluol at 100°. 10 grs. used. 



III. 60 55 



IV. 55 G5 



Toluol at 120°. 5 grs. used. 



V. 10 82 



Parabromtoluol at 120°. 



XIV. 25 95 



XV. 30 90 



Orthobrojitoluol at 100°. 



XIX. 30 21 



XX. 45 31 



XXI. 55 32 



The want of agreement between these results can be ascribed to 

 two causes: first, differences in the rate of adding the bromine; second, 

 unavoidable defects in the method of bromiring or of treating the pro- 

 duct. To the first class belong the difference of 18 per cent, corre- 

 sponding to 45 minutes difference in rate between Nos. I. and II., and 

 that of 10 per cent, corresponding to 15 minutes between Nos. XIX. 

 and XX. ; but these are the only differences that come under this class, 

 a difference of less than 10 minutes having apparently no effect upon 

 the amount of benzylbromide formed (see XX. and XXL). The 

 differences due to the second cause are that of 10 per cent between 

 III. and IV., 13 per cent between VI. and VII., and 5 per cent be- 

 tween XIV. and XV., which show that our method is incapable of 

 giving the exact quantitative results that we had hoped for. The 

 most probable cause of these variations is the difficulty of washing and 

 handling the liquid or pasty product of the bromiring. In the follow- 

 ing discussion, Experiment I. will be rejected, because of the very slow 

 rate at which the bromine was added, and differences of less than 10 

 per cent will be accepted, if at all, only with the greatest caution. 



Effect of the Temperature on the Amount of Benzylbromide formed. 

 — To make this comparison easier, the results given in Table I. (re- 

 duced to the mean where there are several observations at the same 

 temperature) are collected in the following table, and in the second 



