Ryan axd O'Neill — Studies in the Diflavone Group. 175 



The crystals are coloured orange by concentrated sulphuric acid, in which 

 they dissolve, giving an orange solution, which has a strong green fluorescence. 

 When this solution is diluted with water, a yellowish-white precipitate 

 forms. 



Ferric chloride imparts a slight brownish-red coloration to boiling alcohol 

 which has been saturated with the substance. 



Mixtures of the § modification with the a, (3, or y-compounds soften about 

 185° C, and gi-adually melt as the temperature rises. Thus a mixture of 

 the S-isomeride with the a-compound from which it was formed began to melt 

 about 185° C, and was almost completely melted below 195° C. 



With a view to determining the structural relationship of it and the three 

 dibenzylidene-diacetoresorcinols described above, we acetylated it by heating 

 0'3 gramme of the substance with 03 gramme of anhydrous sodium acetate 

 and 3 ccs. of acetic anhydride. The acetyl derivative, which was oily, was 

 treated with bromine in the manner described above. About 0-3 gramme of 

 the diacetate-tetrabromide was obtained. It crystallized in colourless needles, 

 which melted at 176-178° C. A mixture of it with the tetrabromide of 

 )3-dibenzylidene-diacetoresorcinol-diacetate also melted at 176-178°, and, like 

 the latter tetrabromide, with which it is identical, it was converted by alcoholic 

 potash into diflavone. 



Of the four modifications of dibenzylidene-diacetoresorcinol, the /3-com- 

 pound is the most soluble in alcohol, and gives a deep red coloration with 

 alcoholic ferric chloride ; the coloration got from the a modification is 

 much fainter, that from the S-modification still fainter, while the y isomeride 

 gives almost no coloration with the ferric chloride. 



Sulphuric acid colours the crystals of the a modification dark red, the 

 (3 and y modifications orange, giving an orange solution, while the solution 

 in the case of the y compound has a yellow colour. The melting-points of 

 mixtures of the aand-y, « and c, /3 and y, /3andS, y and o isomerides were 

 not sharp, and in each case the mixture melted below the meltiug-point of 

 that component in the mixture which had the lower melting-point. 



^Notwithstanding the marked difference between the crystalline form of 

 the a and j3 isomerides and the differences in their behaviour towards concen- 

 trated sulphuric acid and alcoholic ferric chloride respectively, a mixture of 

 the a and /i isomerides showed no appreciable lowering of melting-point. 



-i. Conversion of the a into the /3 Isomeride. 



We dissolved about a gramme of a-dibenzylidene-diacetoresorcinol in 

 chloroform, added a couple of drops of a 40 per cent, solution of bromine in 

 chloroform, and let the solution stand for several days in a bright place. The 



K.I.A.. PROC, VOL. XXXII., SECT. B. '_2 C] 



