APPENDIX 225 



zonale, and a dark-violet variety of Dahlia, both of which contain 

 pelargonin and cyanin. In such cases variation might be due to the 

 loss of either pigment. 



It is important to keep in mind two considerations in connection 

 with colour-variation. On the one hand, the same anthocyanin may 

 be present in several species, but may not give the same colour owing 

 to the presence in the cell-sap of other substances which are peculiar 

 to the respective species, and which may enter into combination with 

 the pigment to form various glucosides, salts, etc. On the other hand, 

 there are the changes in colour in the Mendelian varieties of a species, 

 and these may be due to the formation of different pigments from 

 that in the type : or to the formation of one pigment only, should the 

 type contain two : or to the modification of the pigment of the type 

 owing to loss or gain of some other substance in the cell-sap. 



Willstatter has also shown by analyses that the amount of pigment 

 in flowers of different shades and colours may vary. The following 

 are some of his results expressed as percentages of dry weight of flowers : 



Cyanin 

 Blue Centamea Cyanus ... ... O7 % 



Deep purple Centaurea Cyanus ... 13-14% 

 Rosa gallica ... ... ... 2 % 



Dark red Dahlia 20% 



Pelargonin 



Pelargonium peltatum 1 % 



P. zonale 6-14 % 



Scarlet Dahlia 4-5-6% 



Pink Centaurea 4 % 



Paeonin 

 Paeonia ... ... 3-3-5 % 



Malvin 

 Malva sylvestris ... ... ... 6-4% 



Many phenomena in colour-variation still remain unexplained. 

 One of the most striking anomalies is the absence of blue from, some 

 genera, for example Rosa. If Rosa gallica and purple Centaurea Cyanus 

 both contain the same anthocyanin, it is difficult to imagine why Rosa 

 should not give a blue derivative. Since, in Willstatter's opinion, 

 potassium replaces one of the hydroxyls of the anthocyanin molecule 



w. P. 15 



