92 THE COLOURS OF FLOWERS. 



of angiosperm, we may, with some confidence, assert 

 that wherever a flower possesses a rudiment of a 

 perianth in any form, it is descended from coloured 

 and entomophilous ancestors. 



Those who have read Professor Ray Lankester's 

 able little work on Degeneration will not be surprised 

 to find that this retrograde agency has played as 

 large a part in the vegetable as in the animal world. 

 We will begin by examining some of the most certain 

 cases, and then will proceed to those in which the 

 evidence is more remote, and the traces of the original 

 petaliferous structure more completely obliterated. 



The Composites are, perhaps, in some respects, the 

 very highest family of entomophilous flowers now 

 existing on the earth. Their very structure implies 

 the long and active co-operation of insect fertilisers. 

 They could not otherwise have acquired the tubular 

 form, the united corolla, the sheathed anthers, the 

 compound heads of many-clustered florets. That 

 originally green flowers could attain to this stage of 

 development, and yet remain green, is simply incon- 

 ceivable. But the Composites contain also some of 

 the most degraded flower types in all nature. Be- 

 ginning with such forms as the common groundsel 

 {Senecio imlgaris), which has an inconspicuous yellow 

 rayless head, specially adapted to self-fertilisation, 

 we go on to plants like the Artemisias, with small 

 greenish florets, which have taken, or are taking, to 

 wind-fertilisation. Still more degraded are the Gna- 

 phalmmSf Filagos, and Antennarias, whose mode of 

 fertilisation is problematical. And at the very bottom 

 of the scale we get the little green XantJiium ; so 

 degenerate a form that its connection with the other 



