Who painted the Flowers ? 31 



tentionally, to become a mere piece of special pleading. 

 It is very easy to collect all the instances that tell one 

 way, and to forget those which tell the other way : it is 

 easy for a man who has too hastily assumed the truth of 

 his hypothesis to see all facts through its medium, and 

 to make them mean something which on more critical 

 examination would be seen not to be their meaning. 

 It seems to me that a conspicuous example of such 

 a process is afforded here, when from the undoubted 

 usefulness of insects to some flowers it has been 

 inferred that all flowers have been entirely modified 

 by insects in all those respects which bring them 

 into connexion. It seems also that even so earnest 

 and so painstaking an investigator as Sir John Lubbock 

 has not escaped the danger above indicated, and has 

 in many instances seen his facts with pre-determined 

 eyes. 



In his work, British Wild Flowers in their Relation to 

 Insects^ from which I have already quoted the general 

 conclusion which he seeks to draw, he runs through 

 the whole British flora, and endeavours in the case of 

 each family to establish the truth of his hypothesis. It 

 seems truer to say that we need go no further than his 

 book to find convincing proof that insects can not do 

 all that is claimed for them. It is not easy to arrange 

 in very logical order the points which arise from the 

 examination of many separate examples. Having in- 

 dicated my general drift, I shall consider it enough to 

 arrange my strictures very much in the order which his 

 work suggests. * 



He tells us, 1 with regard to anemophilous^ or wind- 

 fertilized flowers, that "it is an advantage to these 

 plants to flower before the leaves are out, because the 

 latter would greatly interfere with the access of the 

 pollen to the female flower." Now it is true that 

 Hazels, Poplars, and the like, flower before the leaves 

 appear, and that they are wind-fertilized ; but no less so 



1 Op. cit. p. 8, 



