OF ARTS AND SCIENCES. 95 



last is the best medium for its crystallization. Cold water con- 

 verts it slowly, boiling water rapidly, into benzoic acid. Alkalies 

 transform it into benzoic acid salts, from the solutions of which 

 mineral acids precipitate benzoic acid in its characteristic leaflets. 



The gelatinous phosphoric anhydride residue from the original 

 benzol solution gave on extraction with benzol only a trace of an- 

 hydride. As this extract did not appeal to contain any other sub- 

 stance, the gelatinous residue was taken up in absolute alcohol, 

 with a view of thus isolating any benzophenone which might have 

 been formed. As the alcohol was poured upon the residue, con- 

 siderable heat was manifested. The alcoholic solution was concen- 

 trated to a small volume, but no crystals could be obtained. After 

 some days, the alcohol having entirely disappeared, there remained 

 in the beaker a small quantity of a pleasant-smelling, oily liquid. 

 This was taken up in water to remove the phosphoric acid, washed 

 with dilute potash to remove benzoic acid, and extracted with ether. 

 Finally there remained a few drops of a brownish-colored oil, which, 

 from its odor, we concluded was benzoic ethyl ester. Unfortunately 

 the amount was too small and too impure to admit of analysis. Its 

 formation may be ascribed to the following reaction: — 



C 6 H 5 CO 



) + 2 C 2 H 5 OH = 2 C 6 H 5 COOC 2 H 5 + H 2 0. 

 C 6 H 5 CO 



It is probable that the anhydride necessary for this reaction was 

 retained, mechanically enclosed, in the gelatinous phosphorous resi- 

 due, from which it is but incompletely extracted by treatment with 

 benzol. Phosphoric anhydride being readily soluble in alcohol, 

 the formation of the ester is easily accounted for. In subsequent 

 experiments we obtained some more of this oil, which boiled be- 

 tween 215° and 225°, (benzoic ethyl ester boils at 211-212°,) but 

 could not get it in sufficiently large quantity to purify for analysis. 

 The formation of esters of the nitrobenzoic acids, as described 

 hereafter, under similar conditions, render it practically certain 

 that the substance we have just described is benzoic ethyl ester. 



II. Action of Phosphorpentoxide upon Orthonitrobenzoic Acid 

 in an Excess of Benzol. 



10 grams o-nitrobenzoic acid, melting point 147°, were dissolved 

 in about 200 c.c. of pure, dry benzol, 10 grams phosphorpentoxide 



