THE ORIGIN OF PETALS. 



15 



flattened petaloid filaments, as in the starof Rethleheiri 

 and many other flowers. The curious scales on the 

 petals of Parnassia palustris are now known to be 

 altered stamens. Looking at the question as a whole, 

 we can see how petals might easily have taken theii 

 origin from stamens, while it is difficult to understand 

 how they could have taken their origin from ordinary 

 leaves — a process of which, if it ever took place, no 

 hint now remains to us. We shall see hereafter that 

 the manner in which certain outer florets in the 

 compound flower-heads of the daisy or the aster 

 have been sterilised and specialised for the work of 



\J i, ^J ^0 1 



Fig. 4.— Transition from stamen («) to petal {U) and sepal (c) in flower of doubl 



e rose. 



attraction, affords an exact analogy to the manner in 

 which it is here suggested that certain stamens may 

 at an earlier date have been sterilised and specialised 

 for the same purpose, thus giving rise to what we 

 know as petals. 



In a few rare instances, petals even now show a 

 slight tendency to revert to the condition of 

 fertile stamens. In Monaudra fistulosa the lower lip 

 is sometimes prolonged into a filament bearing an 

 anther: and the petals of shepherd s-purse {Capsella 

 Imrsa-pastoris) have been observed anthcriferous. 



The hypothesis upon which we shall hereafter 



