dissolved 

 II. 76 



Hydride of Bensoyle. 523 



amount of sodium in the product being ascertained also by a 

 direct determination in the whole quantity made from a known 

 weight of benzoic aldehyde. Of the following analyses, the most 

 accordant of a large number, the two first were performed on the 

 compound as prepared with pure sodium ; the two latter on the 

 body owing its origin to the action of sodium-amalgam. Although 

 the flasks were supplied with a slow stream of nitrogen in order 

 to prevent oxidation during the experiment, yet small quantities 

 of benzoate of sodium were invariably produced. This formation 

 explains the large discrepancies in my results, amounting in some 

 cases to a difference in the percentage of sodium of more than 3 

 per cent. 



I. 3-13 grm. of € 7 H 5 0, H f*695 grm. of Na by analysis. 



\'67l „ „ difference. 

 / *169 „ „ analysis. 

 \ '149 „ „ difference. 



III. 1*045 „ „ *218 „ „ analysis. 



IV. 1-019 „ „ '22 „ „ analysis. 



These numbers correspond to the following percentages of 

 sodium in the product : — 



III. IV. 



17-18 17-75 



which seem to point to the empirical formula 



C 7 H 6 0,Na 



for the sodium-compound, which requires 17*83 per cent, of 

 sodium. This formula is scarcely admissible on several grounds. 

 To double it merely would not be to make it more intelligible ; 

 but I think a reasonable rearrangement of the doubled formula 

 is suggested by the decomposition which the new sodium -body 

 undergoes when treated with water (or acids), — benzoic alcohol 

 (hydrate of benzyle or toluenyle), benzoic aldehyde, and hydrate 

 of sodium being formed, according to the equation 



2 (G 7 H 6 9, Na) + 2H 2 = Q 7 H 7 , II, O 4- G 7 H 5 O, H + 2 NaHO. 



Now, if we assign to the original sodium- compound the formula 



Sodium-benzoic Benzoylide 

 alcohol. of sodium. 



then the reaction with water may be thus explained : — 



I. 



II. 



Mean. 



Mean. 



17*83 



17*3 



