586 



the number of equivalents of these two elements not being divisible 

 by 2, many chemists were inclined to double the formulae of both 

 bodies, and to represent them by the following expressions : 



Azobenzol C 12 H 10 N 2 



Benzidine C 12 H 12 N 2 



This view received the first experimental confirmation in the forma- 

 tion of the nitro-derivatives of azobenzol, which were examined in 

 1849 by Gerhardt and Laurent. The formation of 



Nitrazobenzol .... C 12 H 9 N 3 O 2 =C 12 (H 9 NO 2 ) N 2 , of 

 Dinitrazobenzol . . C 12 H 8 N 4 O 4 =C 12 [H 8 (NO 2 )J N 2 , 

 and of several derivatives of these bodies, having established the 

 C 12 -formula of azobenzol, but little doubt could be entertained re- 

 garding the formula of berizidine, which is as readily obtained from 

 azobenzol by reducing agents, as it may be reconverted into azobenzol 

 by nitric acid*. 



The molecular value of benzidine being thus almost exclusively 

 fixed by the determination of the formula of the compound from 

 which it originates, it was of some interest to obtain additional experi- 

 mental evidence for the molecular weight of azobenzol. 



With this view I have determined the vapour- density of azoben- 

 zol. This body boiling at a rather high temperature, I have availed 

 myself of the method of displacement lately proposed by Professor 

 Hofmann. Experiment proved the density of the azobenzol-vapour 

 to be 94 referred to hydrogen as unity, or 6'50 referred to air. The 

 theoretical vapour-density of azobenzol, assuming that one molecule 

 of this compound furnishes, like the rest of well-examined sub- 



1 82 

 stances, 2 vols. of vapour f, is = 91 referred to hydrogen, and 



6-32 referred to air. 



The determination of the vapour-density, then, plainly confirms the 

 higher molecular weights proposed for azobenzol and for benzidine. 



When determining the vapour-density of azobenzol, I had occasion 

 to observe that, probably in consequence of a typographical error, 

 the boiling-point of this compound is misstated in all the manuals 

 which I could consult, and even in the original memoirs of Mit- 

 scherlich himself. The boiling-point is stated to be 193 C., whilst 

 it is in reality 293 C. 



* Noble, Journal of the Chem. Soc. vol. viii. p. 293, f H 2 = 2 vols. 



