70 THE COLOURS OF FLOWERS 



enclosed in a pericarp or envelope), which include the 

 vast majority of flowering plants, Mr. Allen differed 

 from the view which, at the time he wrote, was gener- 

 ally accepted by evolutionists, namely, that flowers with 

 green corollas were the most primitive, and that other 

 colours were developed as a change from green. He 

 suggested, that, on the contrary, green corollas had lost 

 their original colour through degeneration. But he 

 agreed with other phyto-biologists in tracing the origin 

 of the more or less showy flowers of Angiosperms to 

 the more primitive bloom of Gymnosperms (that is, 

 plants bearing naked seeds as do conifers, cycads, and 

 a few others). The flowers of such plants have no 

 corollas, and are mostly composed of green or greenish 

 scales or leaves. Upon this theory Professor G. Hens- 

 low comments thus : 



'When we remember that the spore-cases and spores of 

 Lycopodium, the anther-cells of Cupressus, the whole anther- 

 scale of Pinus, and all the pollens of Gymnosperms are 

 yellow ; again, when we come to Dicotyledons and find the 

 prevailing tints of stamens is the same, we gather proba- 

 bilities in support of that view.' * 



Assuming, then, as may be pretty safely done, that 

 the petal is an altered stamen, 2 and that the all but 

 universal, or at all events the largely predominant, 

 colour of stamens is yellow, it seems probable that the 

 original colour of all, or nearly all, petals was yellow. 

 Allen and Henslow agree in opinion that the next 

 colour to be developed was red, and thence through 



1 The Structure of Flowers, p. 179 (1888). 



2 Altered bracts or leaves do not develop into petals but into 

 coloured calyces or involucres. 



