120 PROCEEDINGS OF THE AMERICAN ACADEMY 



to stand for several days, when it was found that pale yellow crystals 

 mixed with potassic bromide had been deposited ; the dark-colored 

 liquid was poured off, and the solid residue freed from parabi'omben- 

 zylbromide by repeated treatment with hot ligroine, and from cui-cumin 

 by boiling with successive portions of aqueous potassic carbonate, 

 until it ceased to give a red solution. The essentially inire ester thus 

 obtained was boiled several times with alcohol, which dissolved a 

 small portion of it, while the residue melted to a reddish-black tar; 

 upon dissolving this in glacial acetic acid and precipitating with water, 

 yellowish flocks were thrown down, the melting-point of which was 

 compared with that of the similarly-colored indistinct crystals obtained 

 by cooling the alcoholic extract. As both these substances melted, 

 or, more accurately, drew together, at the same temperature, the ester 

 seemed to be essentially pure, and after drying at 50°-60° was 

 analysed. 



I. 0.1796 gr. of substance gave 0.3980 gr. of CO^ and 0.0817 gr. 



of H,0. 

 II. 0.2506 gr. of substance gave according to Carius 0.1156 gr. of 

 AgBr. 



Found. Calculated for 



Ci4Hi3(C,H,Br)0, Ci6H,,(C,HeBr)04 



Carbon 60.43 60.72 62.58 



Hydrogen 5.05 4.57 4.76 



Bromine 19.63 19.26 18.14 



From these results there can be no doubt that Cj^Hj^O^ is the true 

 formula of curcumin. 



The ester consists of indistinct crystals, grouped in forms like 

 cauliflowers, of a much paler yellow color than curcumin ; it mt_dts at 

 7G°-78°, beginning to draw together at 76°, and becoming thoroughly 

 liquid at 78°; we have not succeeded in obtaining it with a perfectly 

 sharp melting-point. It is more soluble in glacial acetic acid than in 

 alcohol, nevertheless the latter is to be preferred as a solvent for obtain- 

 ing crystals, since the substance is apt to separate from the hot glacial 

 acetic acid in a fused tarry condition. It is readily soluble in ether 

 and benzol, but does not crystallize well from these solvents ; slightly 

 soluble in carbonic disulphide ; essentially insoluble in ligroine ; not 

 attacked by a solution of potassic carbonate, but soluble in potassic 

 hydrate, although without the red color characteristic of curcumin. 

 As the analysis of this ester establishes our formula, we have not 

 continued the study of the esters. 



