1898] 



THE STUDY OF VARIATIONS 235 



" I. Darwin asserts that Natural Selection has no relation what- 

 ever to the primary cause of any modification of structure. 



" II. A changed environment — especially that of cultivation — 

 stimulates variability — i.e., the innate capacity of varying, which 

 results in variations of structure. This fact is recognised by Darwin, 

 Weismann, Spencer, and all other bicjlogists. 



" III. Under cultivation variations, especially after several years, 

 are often indefinite, as may be seen in wheat, maize, and in numerous 

 garden plants (but not in all, as sea-kale and asparagus). Hence 

 artificial selection is absolutely necessary. 



"IV. In nature variations are always definite, and not excep- 

 tionally so, as Darwin thought. The consequence is that ' all or 

 nearly all individuals become modified in the same way.' 



" V. The result of the preceding is that a new variety, and 

 thence a new species, would be produced ' without the aid of Natural 

 Selection.' " 



I propose to deal with the last conclusion first ; the statement 

 is made that if definite variations occur " a new species would be 

 produced ' without the aid of Natural Selection.' " 



And that this meaning should be perfectly clear, and that the 

 total exclusion of Natural Selection in the formation of species is 

 intended, he elsewhere places the words " a seedling survives solely 

 because it is vigorous " in italics. 



If for the sake of argument it is granted " that variations are 

 always definite," and also that definite variations are unexplainable 

 by Natural Selection, and therefore that "all or nearly all the 

 individuals become modified in the same way," it is still necessary 

 to show not merely that all are similarly modified, but also 

 that they are all equally thoroughly so, otherwise the 

 variation that is most adapted will probably or at least 

 possibly be selected, and Natural Selection will thus become 

 a factor of some importance. 



Again, it is stated that " In nature, variations are always definite," 

 and it apparently is assumed by Henslow and other Lamarckians 

 that this of necessity upsets the Darwinian position. 



It does not appear to me seriously to weaken the theory of 

 Natural Selection, because more or less direct variations would 

 ultimately be produced by this factor alone. 



Assume that at any period, however remote, variations were 

 completely indefinite, and suppose that an individual A gave rise 

 to 11 varieties, cO- a^ a^ up to a^^ ; assuming that the odd numbers 

 were more or less unadaptable varieties, 2, 4, 6, 8 and 10 would be 

 selected, and then on the simple assumption of hereditary trans- 

 mission of tendencies, which all biologists admit, these varieties 

 2 to 10 would give rise to still more adaptable varieties, until each 



