THE ANTHOCYAN PIGMENTS 605 



line pigment, Grafe described, as in the case of the hollyhock, an 

 amorphous compound, and here also considered the latter to be a 

 glucoside, the former sugar-free. The crystalline compound he 

 described as very unstable and deliquescent ; as this is not true 

 of the pure pigment, his crystals must have contained impurities 

 — for it he gave the formula C 18 H 2 60 13 , and considered that it 

 was a tribasic acid, having two hydroxyl and two carbonyl groups 

 within the molecule. To the amorphous substance he gave the 

 formula C 2 iH440 2 o, and from it by hydrolysis, obtained glucose ; 

 here again, he does not appear to have examined the resulting 

 non-glucoside pigment. Of the crystalline substance he obtained 

 10 grams, together with 15 grams of the amorphous compound, 

 from some 28 kilograms of fresh petals. 



Having thus overcome the experimental difficulties involved, 

 and having for the first time obtained considerable quantities of 

 an anthocyan in a crystalline condition, it is to be regretted 

 that Grafe drew such incorrect conclusions from his results. 

 Doubtless Grafe had in mind the conclusions of Heise and Glan, 

 that both glucoside and non-glucoside pigments were present in 

 the plants they examined, when, on finding that his amorphous 

 product reduced Fehling solution, but that his crystalline sub- 

 stance did not, he concluded that the former was a glucoside, 

 the latter not, and convinced himself of this by hydrolysis of 

 the amorphous product whereby he obtained glucose. Recent 

 work has proved that Grafe's amorphous product must have 

 been an impure specimen of the glucoside, containing reducing 

 sugars, whilst the crystalline substance was the glucoside in 

 practically pure condition ; the possibility that such glucosides 

 when pure do not reduce Fehling solution never appears to 

 have occurred to Grafe, and he never seems to have attempted 

 to hydrolyse his crystalline pigments. He concluded that his 

 work, together with that of Heise and Glan, proved the coexis- 

 tence, in the cases examined, of glucoside and non-glucoside in 

 the plants. This conclusion has, however, been proved erroneous 

 by the work of Willstatter and Everest, who have shown that 

 only the glucoside exists in these plants. Grafe's conclusions 

 led him to suggest that the two pigments he obtained were re- 

 lated to one another, and might be formed one from the other 

 according to the following scheme : 



C, 4 H 44 2 o + H 2 -> C 6 H 12 6 + C 18 Hs 4 0i6 (hypothetical) ; 

 then : C 18 H, 4 0i 5 - 4H 2 + 2 -> CisHseOis- 

 40 



