The Production of Anthocyanins and Anthocyanidins. 259 



In the comparison of the author's anthocyanin chloride with Willstatter 

 and Weil's violanin chloride, the following points were examined : — crystalline 

 form, colour and reflex of the crystalline chloride ; colour of solution in 

 aqueous acid and in alcohol ; colour change on addition of ferric chloride to 

 an aqueous, or alcoholic, solution ; colour change on addition of alum to an 

 aqueous solution ; solubility in acidified methyl alcohol, in acidified ethyl 

 alcohol, in water, in aqueous hydrochloric acid (0"15, 0'3, 0-5, 5, 20, and 

 30 per cent. HCl respectively), in aqueous sulphuric acid (0"05 and 7 per cent.) : 

 the distribution of pigment between amyl alcohol and dilute aqueous acid ; 

 and finally the crystalline form, colour, and solubility in water of the picrate. 



A determination of the water of crystallisation carried out by Mr. A. J. Hall, 

 B.Sc, A.I.C., showed that when dried to constant weight at the temperature of 

 a boiling water-bath and pressure 10 mm., the air-dried crystalline chloride 

 lost 15'68 per cent, of its weight (used 0"5658 grm. air-dried product ; loss = 

 0"0887 grm.). Willstatter and Weil found that violanin chloride (air-dried) 

 when dried in vacuo over sulphuric acid lost 15'29 and 16'15 per cent. 

 (2 separate estimations). 



Whilst the author found very slight divergences from Willstatter and 

 Weil's descriptions in respect of colour and reflex of the crystalline chloride, 

 which the author would describe as red-brown and golden respectively, and 

 the solubility in very dilute (0'05 per cent.) sulphuric acid, in which it is 

 rather more soluble than the description of WiUstatter and Weil would lead 

 one to assume, the solubility being nevertheless very small, in every other 

 point, save the alum reaction, which was very pronouncedly different, the 

 pigments possessed identical properties, and there can be no doubt that the 

 pigment isolated by the author is violanin chloride. The striking difference 

 found in the alum reaction, which at first led the author to believe that his 

 product was different from violanin chloride, has been satisfactorily explained, 

 as will be seen from the results below. 



The Alum Beaction of Anthocyan Pigments. 



In a number of instances Willstatter and his collaborators have included 

 among the characteristic reactions of anthocyan pigments described by them, 

 beyond a colour reaction with ferric chloride, one also with alum or, as in the 

 case of malvin and malvidin, they have stated that pigments give neither 

 colour reaction with ferric chloride, nor with alum. 



In the case of violanin chloride, tliey state that alum gives a blue coloration, 

 whereas when the author examined the pigment obtained by him from the 

 purple-black Viola, and which is doubtless identical with violanin chloride, 

 he was unable to obtain any such coloration. 



