214 PROCEEDINGS OP THE AMERICAN ACADEMY 



oxycliloride diluted with ligroine and afterward washed repeat- 

 edly with ligroiue ; the dark purple viscous mass thus obtained was 

 dried in a desiccator over sulphuric acid, liine, and paraffiue. This 

 treatment, the best which we could devise, was usually far from 

 effectual, as in every case but one the substance contained phosphorus, 

 and was invariably converted during drying into a black mass having 

 very different properties from the purple viscous substance at first 

 obtained. The analysis of the substance free from phosphorus gave 

 results less than one per cent higher than those required by curcumin, 

 and this was confirmed by the analysis of one of the preparations 

 containing phosphorus, after the phosphorus present had been calcu- 

 lated as phosphoric acid, and subtracted from the weight of the 

 substance. 



The properties of the purple product are quite as striking as its 

 formation from curcumin ; for the addition of water converts it again 

 into curcumin, and the change from purple to orange-yellow is instan- 

 taneous. That curcumin was formed in this case was shown by the 

 melting-point, 178°. Alcohol also changes the color instantly from 

 purple to yellow ; but the product is much more soluble in alcohol 

 than curcumin, and is possibly its ethylether. Ether acts in the same 

 way, leaving, on evaporation, a viscous red mass ; in ligroine and benzol 

 it is essentially insoluble. By standing even in desiccator it is grad- 

 ually decomposed — more rapidly at 100° — into a black mass, which is 

 unaffected by water, but soluble in alcohol, or sodic hydrate forming 

 dark solutions. Owing to the uninviting properties of this decomposi- 

 tion-product it was not studied further. 



It is highly probable that the blood- red color imparted to strong 

 sulphuric acid by curcumin is due to this substance, as curcumin is 

 deposited when this solution is diluted. As to the nature of the 

 purple substance, our analyses show nothing ; but its easy conversion 

 into curcumin by the addition of water indicates that it is an anhy- 

 dride, and we are inclined to believe that the carboxyl of the curcumin 

 alone is involved in the reaction, because we obtained a similar but 

 somewhat redder color when monacetcurcumin was treated with 

 phosphoric oxychloride. 



Id our first paper on curcumin we assigned to it the formula 



CJLjCOH) (OCH,) (CellcCOOH). 



Our work since then lias been directed toward the determination of 

 the structure of the side-chain CgllyCOOII, and, although we have 



/ 



