1 1 8 THE COLO URS OF FLO WERS. 



of the rarer and obsolescent purple form, L. 

 maculatum. 



It would, of course, be impossible to treat every 

 similar instance at equal length without swelling this 

 volume to an unreasonable extent. But if the reader 

 will carefully examine, at first hand, all cases of what 

 seem to him adverse examples, he will usually find 

 some such hint of the true relation, surviving in the 

 flower itself. Excellent studies may thus be made of 

 Teucrium scorodonia, compared with our three other 

 British Teucriums (where the calyx suggests the rela- 

 tive stages of development) ; of Ajuga chamcepitys 

 with A. reptans ; of our three Melampyrums ; of 

 Rhinanthus and Pedicularis, and of the various 

 Linarias. The Orobanches are also full of instruc- 

 tiveness, as are likewise Pinguicula and Utricularia. 

 The descriptions given in Floras and other botanical 

 works, and even the best coloured plates, supply very 

 inadequate ideas of the minute observation involved 

 in the study of this subject from the evolutionary 

 point of view. Dried specimens are of course almost 

 useless. The investigation must be conducted upon 

 the living corolla in all stages of its development. 

 Those who will take the trouble thus to watch the 

 actual growing flowers for themselves will soon learn 

 to recognise many other little marks of relative pro- 

 gress or retrogression which cannot all be set down 

 definitely in black and white without unnecessary 

 and tedious prolixity. 



If the general principle here put forward is true, 

 the special colours of different flowers are due to no 

 mere spontaneous accident, nay, even to no mean- 

 ingless caprice of the fertilising insects. They are 



