52 CALIFORNIA ACADEMY OF SCIENCES. 



be selected. We find then, that before variations can 

 occur in the phylogenic series they must take place in 

 the ontogenic series, or in other words, natural selection 

 is strictly conditioned upon individual variation. It 

 therefore becomes a matter of no small moment to en- 

 quire into this subject, from the standpoint of observa- 

 tion to determine what are the possibilities of variation, 

 and from the theoretical aspect to establish the condi- 

 tions of variation. Wallace, more than any other 

 scientist, probably, has given attention to the record- 

 ing of individual variations, and his most recent work, 

 Darwinism, furnishes us with a large number of facts of 

 this sort. An inspection of the instances there given* 

 and of cases adduced by other writers, as Semper in his 

 Animal Life, discloses the fact that they are all quanti- 

 tive and not qualitative. Variations may occur in size, 

 in shape, in position, in number and in color. Varia- 

 tions in size are most numerous and most marked, per- 

 haps, but such variations if in any special direction 

 entail variations in shape. Variation in the position 

 of parts is almost equally great, perhaps, although less 

 readily observed and consequently less frequently re- 

 corded. Variations in number may be trivial and 

 unnoticeable, or may become monstrous as with six- 

 fingered men and two-headed calves. Besides the 

 above mentioned deviations I know of none, except 

 the occasional suppression of parts, either internal or 

 external. 



The above variations might at first sight appear suf- 

 ficient to produce changes in structures already in ex- 

 istence, but the old difficulty still remains of accounting 

 for the origination of new parts. Wallace's answer to 

 this is certainly apologetic. t Have we a right to assume 



* Darwinism, pp. 41-82. 

 U. 0., pp. 128-131. 



