142 



PROCEEDINGS OP THE AMERICAN ACADEMY. 



firmed by the behavior of the substance. Hydroxylamine does not act 

 on the body ; but with phenylhydrazine products are obtained, to which 

 we have devoted much attention with little result, because they are so 

 unstable. They seem to resemble the isoamylamine compounds de- 

 scribed later, and, if this is the case, tell against a ketone formula for 

 the substance. This evidence, therefore, so far as it goes, is unfavorable 

 to the ketone formula, but not conclusive against it. The action of a hot 

 solution of sodic hydrate on our substance converted it into the sodium 

 salt of chloranilic acid, and this result cannot be brought into harmony 

 with a ketone formula, for in that case the sodic hydrate must have con- 

 verted the ketone group C-0 into a phenol group C-OH. The ketone 

 attachment of the oxygen must, therefore, be rejected. 



We have not succeeded in constructing any formula for our substance 

 with the oxygen present in a hydroxyl group, as this would give seven sub- 

 stituting radicals to be arranged upon the benzol ring ; and we may add 

 that the existence of a hydroxyl group is inadmissible on experimental 

 grounds, since cold sodic hydrate has no action upon the substance. 



After we had thus proved that the oxygen is not attached to a single 

 atom of carbon, a molecular weight determination was made, which 

 showed that the body was derived from only one molecule of the sub- 

 stituted quinone. We next considered the different ways in which the 

 atom of oxygen could be attached to two atoms of carbon, and this led 

 to four possible formulas for the substance : — 



COCeHs 



II. 



CH,-0 0-coaH 



'6"6 



CH,0 0-coaH 



e'^iff 



m. 



O O-COCcHg 



ciij y 0CH3 



CHoO o-coa,H 



CH,0, 



IV. 



O-COC.Hs 



C^^b 



CH3O O-COCuHg 



