THE ANTHOCYAN FIGMENTS 607 



interesting papers upon the formation of anthocyans, and the 

 hypothesis which they support is very similar to that of Miss 

 Wheldale, but they part company with that author in regard 

 to the process necessary subsequent to hydrolysis of the 

 glucosides, for they maintain that the oxidation must be pre- 

 ceded by reduction of the non-glucoside flavone or flavonol 

 derivative. 



In 191 3, Willstatter and Everest {Annalen, 401, 189) published 

 an account of investigations upon the anthocyan pigments, and 

 in particular of the pigment of the cornflower. In this commu- 

 nication some important conclusions were arrived at. It was 

 proved that the blue form of cornflower pigment was a po- 

 tassium salt, the free pigment being violet in colour, whereas 

 the red form was not, as had always been assumed by previous 

 workers, the free pigment, but an oxonium salt in which the 

 pigment was combined with an equivalent of some mineral or 

 plant acid. The anthocyans were found to be most stable when 

 in the form of these oxonium salts. It was definitely proved 

 that the decolourisation in solution, so often mentioned by other 

 workers, was not due to reduction. 



Having obtained the cornflower pigment pure and crystalline 

 in the form of its chloride, they proved that it was a disaccharide 

 from which, on hydrolysis, two molecules of glucose were split 

 off from each molecule of pigment ; the sugar-free pigment was 

 obtained in a finely crystalline condition. Microphotographs of 

 the crystals were given. 



The first step in their preparation of the pure glucoside 

 consisted in repeated rapid fractional precipitation of the blue 

 pigment from aqueous solution by means of alcohol. By this 

 means a pigment about 30% pure was obtained, at which stage it 

 was converted into the chloride by solution in hydrochloric acid, 

 nearly all the impurities (chiefly pentosans) were precipitated by 

 the addition of alcohol, but the pigment remained in solution; 

 on addition of much ether to the clear solution, the pigment was 

 precipitated. After further fractional precipitation from alcoholic 

 solution by means of ether, the pigment was taken up in alcohol, 

 some dilute hydrochloric acid added, and slow evaporation 

 allowed to take place, when the glucoside separated as chloride 

 in well-formed crystals. The pure glucoside does not reduce 

 Fehling solution. By boiling the crystalline glucoside with 20% 

 hydrochloric acid for three minutes, it was quantitatively hydro- 



