IV] THE FLOWER AND METAMORPHOSIS 55 



through the foliage region and the bracts to the 

 flower, and finally to the perfected fruit. He noted 

 the modification of one form into another in a regular 

 succession, till the " acme of Nature " is reached in 

 reproduction by seed. On the other hand he designated 

 as "retrogressive metamorphosis" the process by 

 which that succession appears to be reversed, as 

 for instance in abnormal or doubled flowers, when a 

 stamen or a carpel developes as a petal, or even as 

 a foliage leaf. Lastly, he styled as "irregular or 

 accidental metamorphoses" those changes of the 

 normal parts which often follow upon the attacks 

 of insects, or other external agents. These general 

 ideas of the relation of the vegetative and floral 

 regions were amplified and made more definite by 

 subsequent writers, and were for a long time widely 

 held. Thus it became a general belief that the flower 

 had resulted from changes wrought in some pre-ex- 

 istent vegetative shoot. 



So long as we direct our attention solely to the 

 Seed-bearing (or Flowering) plants, and are prepared 

 to pass lightly over all comparisons with Vascular 

 plants lower in the scale, this opinion may stand. 

 But as the 19th century drew on, the knowledge of 

 the lower forms was greatly widened, especially in the 

 case of such plants as the Ferns and Club- Mosses. 

 This supplied the material necessary for a revised 

 theory of the origin of the flower. Moreover, about 



