^ The Production of Anthocyanins and Anthocyanidins. 263 



after separation, was washed six times witli fresh quantities of aqueous acid 



of varying concentrations to remove all traces of anthocyan. The last four ^ 



washings were quite devoid of colour, and, on making alkaline with sodium 



carbonate, gave but the faintest tinge of yellow. The ether layer, after 



receiving the above treatment, was filtered through double filter papers to 



remove all traces of suspended plant matter, then shaken with dilute 



sodium carbonate solution, whereupon a fine deep, clear olive-green aqueous 



layer was produced. On standing, this colour soon passed through a rather 



bluer shade of green to a brown-green, then to orange-yellow. These 



observations, together with the fact that the glucoside gave yellow, not 



green, with dilute alkalis, gives a strong indication of the possible presence 



of myricetin. 



Further confirmation of the presence of a myricetin (or less probably 

 gossypetin) glucoside in the Violas under investigation was obtained in the 

 following way : — 



When in the original precipitation of the anthocyan pigment the 

 extract obtained from the flower petals was precipitated by means of the 

 addition of 3^ times its volume of ether instead of 2\, the process above 

 described gave an ether-alcohol liquor which was so completely devoid 

 (after the CaCOs treatment) of anthocyan that its presence could be 

 neglected. When such a liquor was freed from ether by distillation, it ' 

 showed but the faintest pink coloration on acidification with cone. HCl, 

 but, on addition of a small amount of Mg ribbon, the solution rapidly 

 developed a fine red colour, as would be expected if a flavone derivative 

 were present. The main bulk of the liquors were deprived of their alcohol, 

 hydrolysed by boiling with acid, and treated exactly as described above. 

 In this case, when the ether solution obtained was examined, it was found 

 to contain only a flavone derivative which gave a yellow colour with dilute 

 sodium carbonate solution, whilst, from the yellow aqueous layer so obtained, 

 a small deposit of a yellow powder was obtained after acidification and 

 standing. Hence there is present in the Viola under investigation not only a 

 myricetin glucoside, but also another yellow sap-pigment, which does not 

 give a green coloration (in sugar-free condition) with dilute alkalis. 



Having failed to detect the presence of myricetin in the liquors prepared 

 in the above manner, the crude anthocyan precipitate from which they were 

 obtained was examined, and it was very readily shown that, by the more 

 complete precipitation of the anthocyan pigment by use of 3^ volumes of 

 ether in place of the 2\ previously used, the myricetin glucoside had been 

 thrown down with the anthocyan, from which it was readily recovered. 



The crude, undried, anthocyan pi-ecipitate, obtained in the above way, was 



