172 



PRINCIPLES OF PLANT-TERATOLOGY. 



legia), and the globe-flower (Trollius), doubling is 

 caused by transformation of stamens and the petal- 

 representatives into sepals ; in Aqui legia this only 

 applies to the variety stellata. Braun observed the 

 same thing in Delphinium orientale. 



Sepalody of the stamen is a progressive pheno- 

 menon ; the normal sepals, in the plants named above, 

 arose from a transformation of the outermost stamens ; 

 the extra sepals are formed in the same way from the 

 remaining stamens. 



Petalody. — Probably by far the best-known of 

 all plant-abnormalities, occurring as it does so uni- 

 versally in most natural orders, and usually as a result 

 of cultivation. 4 6 Doubling " of flowers is largely, 

 though not exclusively, caused by this phenomenon. 

 It is to this that we mainly owe the exquisite scent 

 and luscious splendour of coloration in such flowers 

 as the garden rose. 



While it is true that, as so many metamorphosed 

 stamens show, the anther is the equivalent of the leaf- 

 blade and the filament of the petiole, the formation of 

 petals, normal or abnormal, may take place in various 

 ways. De Candolle points out that in Clematis, 

 doubling, or petalody, is due to expansion of the 

 filament, in Ranunculus it arises from dilatation of the 

 anther, in HeUeboriis the petals arise from both 

 filament and anther, and in Camellia all three types 

 occur on the same plant. Hence, knowing as we do 

 from the leafy transformations of the stamen that, as 

 a rule, anther corresponds to lamina and filament to 

 petiole, it would be incorrect to conclude that every 

 petal, wheresoever occurring, represents an entire 

 leaf consisting of both petiole and lamina ; the morpho- 

 logical nature of any given petal can only be deter- 

 mined from a study of the various transitional stages 

 of the metamorphoses through which it has passed in 

 its evolution from a stamen. Here again we see the 

 great value of the study of abnormal structures. Yet, 

 as showing that all parts of a leaf are really homo- 



