OF ARTS AND SCIENCES. 229 



The substance is obviously metanitrobenzophenone, and its formation 

 may be ascribed to reactions analogous to those above given for the 

 corresponding ortho compound. The yield is much better than in the 

 case of the ortho body, and its purification is a comparatively simple 

 operation. It is insoluble in water, easily soluble in hot alcohol, ben- 

 zol, and chloroform. Ether also dissolves it readily. Absolute alcohol 

 is the best medium for its crystallization. 



III. Action of Phosphorpentoxide upon Paranitrobenzoic Anhydride 



in an Excess of Benzol. 



4.5 grams paranitrobenzoic anhydride, m.pt. 184°, prepared from 

 the acid and phosphoric anhydride, were heated in a scaled tube with 

 5 grams phosphorpentoxide and 30-35 c.c. pure dry benzol at 190-2(JO J 

 for about four hours. As with the ortho and meta anhydrides, the 

 product of the reaction consisted of two portions, a solid black deposit, 

 and a more or less dark-colored supernatant liquid. The latter gave, 

 after purification and concentration, a precipitate of anhydride, easily 

 identified by its crystal form and melting point. 



The residue was worked up in the same manner as described under 

 the ortho aud meta compounds. The chief product consisted of yellow- 

 ish white scales melting at 137-138°, and, as the analyses prove, is 

 paranitrobenzophenone. 



0.2408 gr. gave 0.6081 gr. C0 2 and 0.0902 gr. H 2 0. 

 0.1944 gr. gave 10.2 c.c. nitrogen at 16°. 4 and 749.3 mm. 



Paranitrobenzophenone crystallizes from hot absolute alcohol in 

 yellowish scales. It is readily soluble in ether, chloroform, ligroiue, 

 benzol, and carbon disulphide, insoluble in water. 



The course of the reaction by which the substance is formed i- 

 obviously the same as in the case of the isomeric ortho and metanitro- 

 phenones. The yield is much the best of the three, the raw product 

 is less gummy, and the quantity of anhydride recovered from the 

 benzol mother-liquor, very small. Traces of free acid arc: neverthe- 

 less pivseut in the alkali washings, so that the possibility of the forma- 

 tion of the paranitrophenone by the first of the reactions above given 

 cannot be entirely excluded. 



