2 Proceedings of the Royal Irish Academy. 



Alizarin, the best-known of the natural red colouring matters, contains 

 two complex chromophores - C = C - CO - C = G - together with ortho- 

 hydroxyls, whilst the active principle of turmeric (curcumin), whose colour 

 is intermediate between that of the red alizarin and the yellow flavone 

 derivatives, has had the formula 



/CO - CH = CH - C e H 3 (OCH 3 XOH) 

 CH 2 

 \CO - CH = CH - C 6 H 3 (OCH 3 )(OH) 



assigned to it by von Kostanecki. 1 



According to Werner* the position of the methylene group between 

 two earbonyl groups will enable this substance to act as a mordant dye. 

 Curcumin is also a substantive dye, and this latter property receives a 

 plausible explanation from the structural analogy of the above formula with 

 that of other well-known substantive dyes 



/chromophore - X 

 E 

 \chromophore - Y. 



This formula also receives support from the facts that curcumin reacts 

 with hydroxylamine hydrochloride to form an isoxazol, and on distillation 

 with aqueous potash yields ferulic acid. 



If this formula be correct, curcumin is a derivative of a typical 

 unsaturated dike tone 



C 6 H 5 CH = CH - CO - CH 2 - CO - CH = CH . C«H 6 

 di-cirmamylniethane which up to the present has not been prepared. Indeed, 

 so far as we are aware, with the single exception of some derivatives of 

 mesityloxide, such as mesityloxide-oxalic ester, obtained by Claisen,' no 

 hormatic unsaturated diketones have hitherto been synthesized. 



We have attempted the synthesis of compounds of this type by two 

 different methods : — 



1. The condensation in presence of metallic sodium or sodamide of 

 the ester of an unsaturated acid, cinnamic, with a saturated ketone, and 

 vice versa. 



2. The condensation of an alkyl diketone with an aldehyde in 

 presence of a dehydrating agent. 



In the present paper we deal only with the first of these methods, and 

 we may remark at the outset that, of the two condensing agents mentioned 

 above, we found metallic sodium in the form of wire much more convenient 

 to work with, and nob inferior in yield to sodamide. 



' Ber. 43 (1910), p. 2163. - Ber. 41 (1908), p. 1067. 3 Annalen, 291, p. 122. 



