THE SUDDEN ORIGIN OF NEW TYPES 397 



structure in an organism would appear to lie in its capacity for 

 variation, A primitive type, that is to say an ancestral type, 

 must a priori have been pre-eminently plastic and variable, 

 otherwise it could not have given rise in the course of evolution 

 to a divergent series of descendants. On the other hand, a 

 modern existing animal or plant showing a marked simplicity 

 of structure has evidently reached this condition by an 

 extreme degree of adaptation to its environment, and it will 

 be found on examination to be due to a process not so much 

 of degeneration as of specialisation by reduction from a more 

 complicated condition. 



Now it was noticed long ago by Isidore Geoffroy St. Hilaire^ 

 that when any part or organ is repeated many times in the same 

 individual {e.g. the vertebrae in snakes, the arms of some star- 

 fishes, the stamens in polyandrous flowers, etc.) its number is 

 variable ; whereas this remains constant in cases when the 

 same part or organ occurs in few numbers. In other words, 

 multiple parts are extremely liable to vary in structure. The 

 high variability, which is necessarily implied by the capacity of 

 a primitive type to give rise to a series distinctly higher in the 

 scale than itself, indicates that the first step requisite to start the 

 beginning of this new series would seem to be the multiplication 

 or repetition of some special part of its organisation. This 

 increase in the number of parts is correlated with an increased 

 variability of the part or organ, and under the sifting influence 

 of natural selection upon the numerous variations a greater 

 efficiency and elaboration of the organ would be rapidly secured, 

 leading to a correspondingly rapid rise and predominance of the 

 new type. 



Strong support of this view has recently been furnished 

 by the investigations of Wieland - on American Bennettitcce ; 

 his results have been still further amplified and pushed to their 

 logical conclusions by Scott ^ and Arber.^ In face of these 



' Hist. Gen. ct Part, ties Anomalies de forgmiisation., t. i. 1832. 



^ Wieland, G. R., A Study of some American Fossil Cycads, Amer. Joiirn, 

 Set. Ser. 4, vii. (1899), pp. 219, 305, 383, xi. (1901) p. 423; and A/ncrican 

 Fossil Cycads, 1906. 



^ Scott, D. H., The Flowering Plants of the Mesozoic Age, etc., Joiirn. Roy, 

 Microsc. Soc. April 1907, p. 129 ; Studies in Fossil Botany, vol. ii. 1909 ; Adapta- 

 tion in Fossil Plants, Pres. Addr. Linn. Soc. 1909. 



■* Arber, E. A. N.,and Parkin, J., On the Origin of Angiosperms, /t^^/v/. Linn. 

 Soc. xxxviii. p. 29, 1907. 



