222 PROCEEDINGS OF THE AMERICAN ACADEMY 



XIY. 



CONTRIBUTION FROM THE SALISBURY LABORATORY OF 

 THE WORCESTER POLYTECHNIC INSTITUTE. 



ON THE FORMATION OF CHLOR AND BROMBENZOIC 



ANHYDRIDES. 



By George D. Moore and Daniel F. O'Regan. 



Presented June 14, 1893. 



In a previous paper * entitled " On the Formation of the Anhy- 

 drides of Benzoic and substituted Benzoic Acids," we have shown that 

 by the action of phosphorpentoxide upon benzoic and mono-nitroben- 

 zoic acids in the presence of an excess of benzol at the boiling tem- 

 perature the anhydrides of these acids are produced. Further 

 investigration has shown that the monochlor and monobrombenzoic 

 acids behave in a similar manner. 



The process by which the chlor and bromanhydrides are prepared 

 is essentially the same as that already described under the nitro com- 

 pounds. The only difference worthy of note consists in using rather 

 a larger quantity of phosphorpentoxide, and in boiling the mixture 

 somewhat longer. Thus, whereas in the case of the nitro anhydrides 

 we employed equal weights of acid and phosphorpentoxide, boiling 

 with benzol for four hours, the halogen compounds require about one 

 fifth excess of phosphorpentoxide and five to six hours boiling. 



I. Orthochlorhenzoic Anhydride. 



The orthochlorhenzoic acid necessary for the preparation of this 

 substance was made according to the method described by Anschiitz 

 and one of us,t by treating salicylic acid with phosphorpentachloride, 

 and, after rectifying the product in vacuo, decomposing it by distilla- 

 tion at ordinary atmospheric pressure. 



10 grams of orthochlorhenzoic acid, prepared in this manner and 

 melting at 137°, were heated with 12 grams of phosphorpentoxide 



* These Proceedings, XXVII. 93. t Ann. Chem., CCXXXIX. 326. 



