170 Proceedings of the Royal Irish Academy. 



theoretically possible stereoisomer^ forms — cis-cis, eis-trans, trans-trans — 

 of dihydroxy-dichalkone. 



In the case of the dibenzylidene-succinic acids, Stobbe (A^erh. d. Ges. 

 deutseh. Naturforsch. u, Aerzte, Munich, 1899, p. 88) prepared three sub- 

 stances corresponding to the three possible stereoisoinei-ides, the phenomenon 

 observed by him being very similar to that met here by us. 



The conversion of the three stereoisomeric modifications of dihydroxy- 

 dichalkone into the same diacetate-tetrabromide is probably due to changing 

 of the two labile isomers into the stable one during the acetylation and 

 addition of the bromine. 



We found, indeed, that the a isomeride, in chloroform solution in the 

 presence of a trace of bromine, on prolonged exposure to light changes into 

 •the /3 isomeride — the latter being probably the most stable of the three 

 forms. 



The -y modification seems to be a structural isomeride of the other three. 

 When subjected to the process of acetylation and bromination, it formed an 

 oily bromine derivative, which did not give diflavone when treated with 

 alcoholic potash. It cannot, therefore, be a dihydroxy-dichalkone. As the 

 quantity of the substance (about one gramme) at our disposal was very small, 

 we were not able to examine it as fully as is desirable. 



Its colour is lighter and its melting-point higher than those of the 

 dihydroxy-dichalkones ; and since flavanones are always lighter in the colour 

 and generally higher in the melting-point than the corresponding chalkones, 

 we were at first inclined to regard the substance as diflavanone (n), all the more 

 as it is nearly insoluble in dilute alkali, and gives only a very faint coloration 

 with alcoholic ferric chloride. 



An attempt to convert a small quantity of it by bromination and treat- 

 ment with alcoholic potash into diflavone not being successful pointed to 

 the chalkone-flavanone formula (in) rather than the diflavanone formula (n), 

 as that which must be given to the compound. 



Although -y-dibenzylidene-diacetoresorcinol is probably not diflavanone, 

 we have found that diacetoresorcinol condenses readily with aldehydes in 

 the presence of alcoholic hydrochloric acid to form derivatives of diflavanone. 

 Thus with benzaldehyde it forms dibenzylidene-diflavanone (v), and with 

 piperonal we obtained dipiperonylidene-3'. 4'. 3'. 4'.-dimethylenedioxy-difla- 

 vanone (vi). 



The method, which has not hitherto been employed for the preparation of 

 similar substances in the monoflavone group, we have found equally applicable 

 to the latter group, and is probably the best and simplest method for the 

 preparation of flavindogenides. Gallacetophenone-dimethylether, under the 



