THE ORIGIN OF THE "FLOWER" 257 



an interesting comparative study of the Ranunculaceae, showed 

 that in all probability both petals and sepals are derived from 

 stamens. On the other hand, Prantl x held that the corolla, 

 along with the calyx, arose from the metamorphosis of bracts 

 or bracteoles situated below the flower. Goebel, 2 while admit- 

 ting that in the majority of cases petals have arisen from stamens, 

 inclines to the belief that in other cases petals, as well as sepals, 

 have taken origin from bracts. In a paper 3 published four years 

 ago I supported Celakovsky's view on the matter, viz. that both 

 sepals and petals are in all cases derived from stamens. 



The phenomena presented by. many double flowers, e.g. the 

 garden rose and daffodil, clearly disclose the staminal origin 

 of the petals, for all transitions may be found between a sharply 

 defined and typical stamen and an ordinary petal, and the 

 stamens undoubtedly existed prior to this latter. 



So far, then, investigation by means of the study of the 

 comparative morphology and teratology (especially the latter) 

 of organs has revealed to us that the " flower " is the homologue 

 or morphological equivalent of a vegetative leafy shoot, of which 

 the foliar organs, along with the axis on which they are borne, 

 have been peculiarly modified for the purpose of subserving 

 the special functions characteristic of the " flower." Inasmuch 

 as the great majority of " sports " or teratological phenomena 

 represent reversions, or attempts at reversions, to a primitive 

 condition of the organs concerned, we may, perhaps, infer 

 that the flower which has proliferated into a leafy shoot is 

 at the same time yielding us a clue to the ancestral con- 

 dition of that structure. Yet this is by no means necessarily 

 the case ; and I think it would be quite unwise at this stage 

 of our knowledge to draw such a conclusion from such a terato- 

 logical structure. For if a "flower" and a leafy shoot are 

 homologous structures, this can only mean, in the long run, 

 that both have been derived from a common ancestor; and 

 this being the case, it is surely possible and, I would add, 

 probable, that each of these structures contains within it the 

 potentialities of the other, and, given the appropriate environ- 

 mental stimulus, is capable of exhibiting its characteristics. 



1 "Beitrage zur Morphologie und Systematik der Ranunculacere " (Engler's 

 Bot. Jahrbiicher, voL ix. 1887). 



2 Organographie der Pflanzen, vol. ii. 1901. 



3 " The Origin of the Perianth of Flowers " {New Phytologist, vol. ii. no. 2, 1903), 



