OBTAINING VARIATIONS 113 



Multiplication of petals may result, as we 

 have already noticed, from the transformation 

 of stamens into petals, or it may come about 

 from the springing into being of new petals de 

 novo, rather than as modifications of any pre- 

 existing part of the flower. The latter appears 

 to be the case with the new rows of petals of the 

 balloon-flower. 



At the present stage the flower has a triple 

 corolla, constituting a very striking modification. 

 The ultimate limits of its variation can be abso- 

 lutely determined only by further series of 

 experiments. 



STIMULATING VARIATION 



The modification of the balloon-flower has 

 somewhat exceptional interest, because there is 

 only a single species of the genus Platycodon, to 

 which it belongs, anywhere in the world. 



In other words, this genus is what is called a 

 monotype, and it is a well-recognized fact that 

 flowers belonging to a genus having only a single 

 species, and even to genera having half a dozen 

 species, are relatively little subject to variation. 

 Rightly considered, this is almost axiomatic; 

 because the very fact that there are many species 

 in a genus proves that the representatives of that 

 genus have been variable; else they would not 



