l8 9*-] Insects a?id Flowers. 



177 



much importance as fertilization by the pollen in an early 

 . stage 1 in securing ultimate fruit. 



5. The floral parts are modified leaves — modified by a 

 process that lessens their vital power — and color in these 

 floral parts is an attribute of weakening vital power, having 

 no relation to the visits of insects. 



6. Plants wholly dependent on insects for fertilization are 

 all perennials. An innumerable number of the flowers of these 

 plants fall unfertilized, and but for their being perennials, 

 many species so dependent would have long since disappeared. 



7- All annuals, though in some cases so arranged that 

 cross-fertilization may occur, can self-fertilize when cross- 

 fertilization fails. In fact annuals are in a general sense self- 

 fertilizers. In almost all cases annuals have every flower 

 fertile. 



8. Flowers do not abhor own-pollen, as the proposition 

 once enthusiastically ran. No flowers are so truly fertile as 

 those of the cleistogene class, while the nearly allied class 

 of plants which fertilize before the corolla expands are also 

 certainly fertile. The list in these two classes has grown so 

 large as to render the proposition cited untenable. 



9- It is conceded now that variety or variation is an essen- 

 tial condition in the order of things, — and that there is no 

 more reason why special forms or colors in flowers should be 

 made dependent on the accident of an insect's visit, than are 

 forms and colors of minerals. The forms and colors of 

 flowers must have had an extensive range had not an insect 

 appeared on the stage. 



We must not forget that what we call the kingdom of nature 

 is a vast organization in which a great number of smaller and 

 inferior powers are working apparently independently, but 

 actually in co-operation or accord with the greater ones. No 

 one phenomenon can be fairly placed to the credit of any one 

 direct cause. The forgetfulness of this fact leads to many an 

 error in our theoretical deductions. At any rate the unchal- 

 lenged propositions I have enumerated, show how many things 

 have to be considered before we accept the wide generaliza- 

 tions presented by those who tell us of the relation of insects 

 to the forms and characters of flowers. 



Germantown, Phila. 



1 This proposition is commended to the author of the note on Cephalanthus, 

 at P. 66 of this volume. 



