M. WlIELDALK 375 



whereas the percentage compositions actually found for the two forms 

 of Anth-rhinum anthocyanin are : 



Red Magenta 



C ... 51-81% C ... 50-50% 



H ... 5-01 7^ H ... 511 % 



O ... 43-18% O ... 44-39 7„ 



Moreover the artificial red product prepared from apigenin by 

 reduction with sodium amalgam does not give the qualitative reactions 

 nor has it the percentage composition of the natural anthocyanin. 



The red and magenta anthocyanins from Antirrhinum were not 

 obtained in crystalline form, nevertheless their purity was guaranteed 

 by the concordance of analysis results when the pigments were prepared 

 independently from entirely different varieties. Also the products 

 analysed were the pigments themselves and not the hydrochloric acid 

 salts. 



The analyses show that both red and magenta contain more oxygen 

 than apigenin. Moreover the red pigment is not an acid salt of the 

 purple or magenta : both are precipitated from acid solutions. Nor is 

 the magenta an alkaline salt of the red. They are different substances, 

 and cannot be converted the one into the other by such means. 



In the work on Antirrhinum, the problem which has been before 

 the author may be exjjressed thus : — What are the actual processes 

 which lead to anthocyanin formation in the living cell ? Only the 

 answer to this question will enable us to interpret the Mendelian 

 factors correctly. 



Willstatter's percentage formulae for several anthocyanins coupled 

 with the production, from quercetin by reduction, of a very small 

 quantity of a substance claimed to be identical with cyanidin are not 

 convincing reasons for regarding the natural process of pigment for- 

 mation as one of reduction. In this respect the important point is 

 whether cyanidin is formed from tfuercetin, delphinidin fi-om myricetin, 

 etc., in the living plant. Willstatter makes no mention of the flavones 

 accompanying the anthocyanins he has isolated. Myricetin which 

 should be the chromogen of delphinidin is at present known in Myrica, 

 Rhus, Haematoxylon and Arctostaphylos. From Delphinium consolida 

 kampherol has been isolated but it is quite possible that other flavones 

 may be present in addition or different flavones in other species. 



Further evidence is still needed of the connection between antho- 

 cyanin and flavone in the same plant, and between the natural 



