EDITORIAL. 167 



cohols for the aliphatic he obtained hydrocarbons. Thus, for example, 

 with benzyl alcohol and sodium benzylate, he did not obtain phenyl 

 benzyl carbinol, but stilben, diphenyl, and toluol. The publication of 

 these results induces me to call attention to some experiments made in 

 connection with my studies on sodium benzhydrol and sodium benzylate.^ 

 I have found by many experiments in which I heated sodium benzhydrol 

 with methyl and ethyl alcohol to between 250° and 300° in sealed tubes 

 that it was not possible to substitute hydrogen in these alcohols under 

 the given conditions. There were obtained with ethyl alcohol and sodium 

 benzhydrol, instead of the expected diphenyl propyl alcohol the follow- 

 ing: Benzol; diphenyl methane; sodium benzoate; tetraphenyl ethylene 

 and ethane; ethylene; ethyl ether, and acetaldehyde condensation pro- 

 ducts; in other words, simply the decomposition products, at this tem- 

 perature, of sodium benzhydrol and of sodium ethylate. It would also 

 seem theoretically improbable that 77 diphenyl propyl alcohol could be 

 formed at this temperature, for the very similarly constituted benzhydrol 

 very readily decomposes at 300°. The results of Guerbet are self-evident 

 from the work of Nef on sodium alcoholates, and from my work on 

 sodium benzhydrol and sodium benzylate. The methylene hypothesis of 

 Nef, assuming an equilibrium between di- and tetra-valent carbon, is not 

 only a very satisfactory explanation for the chemical behavior of alkyl 

 halides and of metallic alcoholates, but makes it possible to predict with a 

 considerable degree of accuracy the reaction between these compounds 

 under various conditions. Fromm ^ obtained symmetrical tetraphenyl 

 butane aa one of the products of the dissociation by heat of benzyl 

 sulphide. According to the theory of ISTef, this must have been formed 

 by the addition of phenyl methylene to dibenzil, the latter resulting from 

 the union of two phenyl methylene particles and the subsequent reduction 

 of the stilben, thus formed, these reactions are represented by equations 

 as follows: 



(1) (CoH,CH,)3S >aH,CH<+H,S 



( 2 ) 2Cc,H„ . CH < > CcH; . CH : CH . CcHs 



( 3 ) C„H,CH : CHC„H5-f-H,S J CeH^ . CH, . CH, . C„H,-f S 



CijHs . CH, HaC . CcHs 



I I 



( 4 ) 2C0H5 . CH <-|- CaH, . CH, . CH, . C5H5 > C„H-, . C C . C.Hj 



H H 



To confirm this method of the formation of symmetrical tetraphenyl 

 butane I chose the benzyl ester of an acid which on decomposition would 

 jdeld products which could reduce the stilben first formed in the reaction, 



'Am. Ohem. Journ. (1905), 33, 68. 



'Ber. d. deutschen chem. Ges. (1903), 36, 534. 



