130 



SCIENCE 



[N. S. Vol. LIII. No. 1363 



from which to select; that individual varia- 

 tion is a qualitative factor, giving the differ- 

 ences which make a selection possible; and 

 that heredity is a conservative factor, holding 

 fast those characters which better fit the 

 organism to its environment. 



Wow it seems to me that there is no possible 

 outcome of the necessary action and inter- 

 actions of these three factors .that would not 

 be a selection of some sort. Darwin thought 

 it comparable in a large way to the selection 

 by which the stock-breeder improves his herd, 

 and therefore called it "natural selection," 

 carefully guarding the phrase from misinter- 

 pretation from the teleological angle as well 

 as from a too close parallelism between arti- 

 ficial and natural selection. And I believe no 

 one has suggested a more acceptable term for 

 the process of selection resulting from the 

 interplay of natural laws. 



Three outstanding theories have been ad- 

 vanced since the publication of the " Origin," 

 each involving an advance in our knowledge 

 of the mechanism of heredity on the one hand 

 and of the origin of variations on the other. 



"Weismann's theory of the continuity and 

 stability of the germplasm was of immense 

 importance in its discussion of the mechanism 

 of heredity, and his amphimixis gave a 

 plausible explanation of the origin of varia- 

 tions. His results were almost universally re- 

 garded as confirming and greatly extending 

 the scope of natural selection. 



Mendel's theory regarding the purity of the 

 gametes, their segi'egation in the sex cells, 

 and the whole complex Mendelian mechanism 

 so admirably described by Morgan; all of 

 these, fascinating and important as they are, 

 deal with the mechanism rather than the fact 

 of heredity. In my opinion their acceptance 

 or rejection does not affect the status of nat- 

 ural selection as a theory of organic evolution. 



But it is the theory of mutation that has 

 furnished most of the ammunition for the 

 opponents of natural selection; and this in 

 spite of the fact that De Vries, the originator 

 of the mutation theory, expresses himself with 

 great clarity as follows: 



My work claims to be in full accord with the 

 principles laid down by Darwin and to give a thor- 

 ough and sharp analysis to some of the ideas of 

 variability, inheritance, selection and mutation 

 which were necessarily vague in his time. 



In 1904, when these words were published, 

 there did seem to be a sharp distinction be- 

 tween the ideas of Darwin and those of De 

 Vries. The former believed that natural 

 selection acted upon many small variations 

 and accumulated them imtil the differences 

 were sirfficient to constitute new species ; while 

 De Vries claimed that new species were 

 formed by the sudden appearance by muta- 

 tions of forms specifically distinct from the 

 parents. That mutants ivere new species! 



It seems evident that Darwin did not re- 

 gard " saltatory evolution " as the common 

 method, while De Vries did. 



Darwin believed that individual, usually 

 small, variations furnished the material on 

 which selection acts; while De Vries thought 

 that mutants, usually large variations, fur- 

 nished the material. Both, however, believed 

 thoroughly that natural selection was a vera 

 causa of evolution. 



But things have changed greatly since 1904. 

 The work of Morgan, Castle, Jennings and a 

 host of others has shown that many mutations 

 are so small, from a phenotypic standpoint, 

 that they are quantitatively no greater than 

 the individual variations of Darwin; and that 

 they are heritable in the mendelian way. 



Castle produced a perfectly graded series of 

 hooded rats which exhibits almost ideally the 

 steps by which a new form might be produced 

 by natural selection. He says: 



If artificial selection can, in the brief span of a 

 man's lifetime, mould a character steadily in a 

 particular direction, why may not natural selection 

 in unlimited time also cause progressive evolution 

 in directions useful to the organism? 



Jennings says : 



Sufficiently thorough study shows that minute 

 heritable variations — so minute as to represent 

 practically continuous gradations — occur in many 

 organisms: some reproducing from a single parent 

 others by biparental reproduction. ... It is not 

 established that heritable changes must be sudden 



