THE SUDDEN ORIGIN OF NEW TYPES 401 



reduction to the optimum number oi five has taken place in 

 the floral leaves and arms respectively, it seems only consistent 

 to infer that in both cases the same cause has been operative, 

 viz. the use of a minimum amount of material consistent with 

 the maximum amount of enclosed space. 



It has usually been held that floral pentamerism is a result 

 of the I phyllotaxis, which is so frequent in Dicotyledons ; 

 but since it has now been proved that whorled flowers (at 

 any rate flowers with staminate whorls) occur in the unbranched 

 or only slightly branched Bcnnettitece, and since free-branching 

 was a habit coincident with the rise of Angiosperms, it appears 

 to me at least conceivable, if not altogether demonstrable, that 

 the I phyllotaxis and the consequent free-branching may be 

 the result, not the cause of the floral pentamerism. The 

 possibility of this view is supported by Rendle's ^ observations 

 on a specimen of heather {Erica cinerea), in which the flowers 

 were replaced by dark-red leaf-buds of about the same size 

 as the flowers ; in these the leaf-arrangement resembled that 

 of the flower, not that of the foliage-leaves. 



Dr. Scott - has cogently argued that evolution by reduction 

 in the number of parts seems to have taken place with com- 

 parative rapidity and suddenness, so that it often " set in at 

 a relatively early stage of evolution." Thus such groups as 

 the Ameiitifcrce, with simple reduced flowers, may indeed be 

 very ancient, though not primitive in the exact sense of the 

 word.^ Whilst in some cases reduction may occur mainly as 

 the result of the natural tendency of organisms to economise 

 material, it seems to me not impossible that in the case of 

 the early Aincntiferce it took place partly as a result of a 

 migration to the Temperate zone of their ancestral types from 

 the Tropics, where variability is always greatest owing to the 

 increased rate of growth induced by heat ^ ; and partly as a 



' Proc. Linn. Soc. Nov. 4, 1909. 



^ Adaptation in fossil plants. Presid. Address, Linn. Soc. 1909. 



^ A striking instance of a remarkably rapid and early reduction is furnished 

 by the Chelonians, for the earliest known example of this class, viz. from the 

 Upper Keuper sandstone of Wiirtemberg, is a complete shell (2 feet long) of 

 Proganochelys belonging to the peculiarly specialised group of the Pleurodira. 

 Yet this Triassic form cannot, geologically speaking, be very far removed from 

 the ancestral Chelonian, considering that no reptiles are known in pre-Permian 

 rocks. 



* Blaringhem, L., Mutation et Tfauntaiisine, etc. Paris, 1908. 



