A Family History. 233 



head to head by bees or butterflies. The two flowers 

 grow also exactly where we should expect them to do. 

 The salad-burnet loves dry and wind-swept pastures 

 or rocky hill-sides, where it has free elbow-room to 

 shed its pollen to the breeze ; the stanch-wound takes 

 rather to moist and rich meadows, where many insects 

 are always to be found flitting about from blossom to 

 blossom of the honey-bearing daisies or the sweet- 

 scented clover. 



Perhaps it may be asked, How do I know that the 

 salad-burnet is not descended from the stanch wound, 

 rather than the stanch-wound from the salad burnet ? 

 At first sight this might seem the simpler explana- 

 tion of the facts, but I merely mention it to show 

 briefly what are the sort of grounds on which such 

 questions must be decided. The stanch-wound is 

 certainly a later development than the salad-burnet ; 

 and for this reason — it has only four stamens, while 

 the parent plant has several, like all the other roseS. 

 Now, it would be almost impossible for the flower 

 first to lose the numerous stamens of the ordinary 

 rose type, and then to regain them anew as occasion 

 demanded. It is easy enough to lose any part or organ, 

 butitis a very different thing to develop it over again. 

 Thus the great-burnet, having once lost its petals, has 



never recovered them, but has been obliged to colour 

 11 



