38 E. E. TURNER. 



than in the still hypothetical reaction : 

 PhMg I + MeMgl + 2 Ou01 2 = Ph.Me + Ou 2 01 2 + Mgh + MgCl 2 , 

 both being apparently quadrimoleeular reactions, it is prob- 

 able that the Fittig reaction is of a far more complex nature 

 than is generally supposed. It is not commonly realised, 

 moreover, that the Fittig reaction is by no means capable 

 of universal application, and that actually it is useless in 

 practice for the preparation of several of the more common 

 diaryl-hydrocarbons in any quantity; in these cases either 

 the chromic or cupric chloride method gives good yields, 

 and (which is equally important) very pure products. 



It would be expected that cupric chloride would react 

 with the magnesium derivatives of halogen fatty acids to 

 give succinic acids. Although, owing to lack of material, 

 it has not been possible to investigate this matter fully, it 

 is concluded that little success is likely to result from a 

 more exhaustive examination. Thus in some preliminary 

 experiments neither bromoacetic ester nor /2-iodopropionic 

 ester reacted in ethereal solution to give the expected 

 succinic or adipic esters. With the simpler ester, a vigor- 

 ous reaction was observed, but this was probably due to 

 interaction between carbethoxymethyl magnesium bromide 

 and unchanged bromoacetic ester. The iodopropionic ester, 

 on the other hand, was recovered, for the most part, un- 

 changed, a result which can be attributed, no doubt, to the 

 slight reactivity of a halogen situated in the ^-position with 

 respect to a carbethoxy-group. 



a-Bromobutyric ester, on the other hand, gave an appre- 

 ciable quantity of s-diethylsuccinic diethyl ester when 

 allowed to react in ethereal solution with magnesium and 



cupric chloride. 



Experimental. 



Preparation of 1 : l 1 Ditolyl. 



Anhydrous cupric chloride (60 grams) was added to a 



Grignard reagent made from 51*3 grams of o-bromotoluene, 



