122 PROCEEDINGS OF THE AMERICAN ACADEMY 



Incomplete Oxidation. — When curcumin was dissolved in aqueous 

 potassic hydrate, a solution of potassic permanganate added, not in 

 excess, and after the oxidation had ceased, the liquid acidified with 

 sulphuric acid, a strong smell of vanilla was observed. The liquid 

 was accordingly filtered, and the precipitate thoroughly washed with 

 boilino- water, the filtrate and wash-water extracted with ether, and 

 the extract treated with acid sodic sulphite, as directed by Tiemann 

 and Haarmann.* The product was an oil, having a strong smell of 

 vanilla and gradually solidifying in circular groups of radiating 

 needles ; the amount, however, was extremely small, and none of this 

 product was obtained with an excess of potassic permanganate or 

 when the quantity of curcumin was much more than half a gramme. 

 The same substance was obtained with various weak oxidizing agents, 

 such as bleaching powder and water, potassic ferricyanide with potassic 

 hydrate, and even the action of atmospheric oxygen on curcumin dis- 

 solved in potassic hydrate. Of these the mixture of potassic hydrate 

 and potassic ferricyanide gave the best yield, but even this was ex- 

 tremely small, — in fact after uniting the product from all the oxida- 

 tions made by us, in which over eight grammes of curcumin were used, 

 the quantity was not enough for complete purification. By sublimation, 

 however, and subsequent crystallization from boiling water, it was 

 obtained in white needles resembling in appearance and odor the 

 vanillin from the vanilla-bean and melting at 79". Vanillin melts at 

 80°-81°. 



In addition to the vanillin there were formed carbonic dioxide, a 

 black amorphous substance with feebly acid properties, perhaps the 

 aldehyd resin of vanillin, as it appeared in largest quantity when no 

 vanillin was obtained, and an acid volatile with steam. We have not 

 as yet made any complete study of these secondary products, because 

 the properties of the humus-like substance are far from inviting, and 

 the amount of the volatile acid is so minute that its isolation in quan- 

 tity sufficient for analysis would be extremely laborious. We shall, 

 however, return to these substances if we fail in finding easier methods 

 for studying the nature of the side-chain. 



As the small yield of vanillin was undoubtedly due to the presence 

 of the phenol hydroxyl, which ofi'ered a point of attack for the oxidiz- 

 ing mixture, we next tried to increase our yield by replacing the 

 hydroxyl hydrogen with some radical which would protect it from 

 oxidation, and in this way not only prove that the substance was really 



* Ber. d. cli. G. 1875, p. 1115. 



