64 COLOUR IN NATURE chap. 



an hour the part of the epidermis still in contact, 

 with the leaf became decolorised, while the severed 

 portion completely retained its colour. As the 

 chlorophyll can hardly be supposed to have a direct 

 effect on the anthocyan, Molisch concludes that the 

 " conditions for the formation of alkaline substances 

 must be especially favourable in chlorophyll-con- 

 taining cells." This is a conclusion of great interest 

 and importance. If cells which contain much chloro- 

 phyll also contain strongly alkaline substances, and 

 therefore, as we may reasonably suppose, tend to 

 destroy any anthocyan pigment which may be 

 formed, then this is equivalent to saying that the 

 conditions which are favourable to the development 

 of chlorophyll are unfavourable to the development 

 of anthocyan, and vice versa. This is a very tempt- 

 ing conclusion, for it seems to explain many points 

 in the distribution of anthocyan which have hitherto 

 been very puzzling. Broadly speaking, anthocyan 

 tends to appear conspicuously in many leaves and 

 shoots in early spring, in many leaves in autumn, in 

 flowers, and in fruits. Its function is often said in 

 the case of vegetative organs to be the protection of 

 chlorophyll from the injurious effects of excessive 

 light, but this does not account for its appearance in 

 autumnal leaves. If, however, we may assume that 

 chlorophyll and anthocyan are in antithesis to one 

 another, we can readily understand why the latter 

 should appear in spring, before the power of assimi- 

 lation is completely established, in autumn when it 

 is beginning to disappear, in ripening flowers and 

 fruits where it is more or less completely lost, and 

 also in the leaves examined by Stahl which grew 



