6o SOUTH AFRICAN FLOWERING PLANTS. 



Orra'niuhi., can 1)0 reached from any parts ; l)Ut in the 

 Labiates, etc., a single gland is often developed at only 

 one spot, at the bottom of the canal or passage, down 

 which the proboscis of the insect finds itself conducted. 



When, therefore, one finds not one only, but many 

 points of coincidence in all the four whorls of a flower, 

 as well as in the position of the honey- glands, one is 

 quite justified by " inductive evidence," i.e. the accumu- 

 lation of many points of agreement, in concluding that 

 some one common cause has brought about these many 

 correlations or adaptations in unison together. 



Moreover, since Evolution forbids the idea of any 

 structure being formed at first in anticipation of the 

 needs of another being, and as there is no evidence of 

 spontaneous evolution without external agencies, we are 

 driven to the conclusion that it is the injects tlicmi^elvcs 

 which first stimulate or irritate the flower, and then 

 all the above adaptations follow suit in response to the 

 irritations. Such is the theory based on these and 

 many other facts. It is that the living Frofojjfasvi is 

 endowed with a power to respond to the mechanical 

 forces brought to bear upon the several organs of the 

 flower by the insect itself, i.e. by its weight and thrusts.^ 



We know from Nature and from experiments that 

 such a response is common enough everywhere in the 



• This is more fully explained in my works, " The Origin of Floral 

 Structures " (International Scientific Series : Kegau Paul, Trench, & 

 Co.). and a little book, " The Making of Flowors" (S.P.C.K.). 



