i;6 THE MUTATION THEORY 



connected by intermediate links with smaller variations. Doubt- 

 less in this also they are right. The mutations which have been 

 discovered in man all display this peculiarity. Yet again they 

 claim that domestic varieties of plants and animals have arisen, not 

 through the fixation of fluctuating variations by continued selection, 

 but through the artificial selection of mutations. Here, also, there 

 is good reason to believe that they are right, at least in large 

 measure. Breeders of plants and animals are no more able than 

 experimental workers to distinguish the smaller differences between 

 mating individuals. Moreover, they are in haste to get good 

 results. Of necessity they have selected large mutations rather 

 than small fluctuations. 1 This, however, has not always been the 

 case. For example, when evolving speed in the American 

 trotting horse, a favourable mutation, involving as it would a 

 co-adaptive change in a thousand structures that is a thousand 

 co-ordinated mutations would be an exceedingly rare, indeed, an 

 almost impossible occurrence. Therefore the breeder must depend 

 on fluctuations, which in this case, though imperceptible to the 

 eye, can be measured by the test of speed. It is admitted that the 

 inheritance of fluctuations is blended, and that, as regards them, 

 offspring display a strong tendency to return to the ancestral 

 type. For this reason, according to Mendelians, do trotting horses 

 tend to retrogress unless carefully selected, and when crossed with 

 inferior types to transmit blended characters to descendants. 



293. Experiments in crossing have been limited in great 

 measure to domestic plants and animals. Wild types are less 

 accessible, more difficult to manipulate, and often more or less 

 sterile when crossed. When, however, the cross is fertile, the 

 hybrid, like man, usually blends the parental characters and 

 transmits the blend to its descendants. In hardly a single instance 

 has the crossing of natural varieties revealed a latent ancestral 

 character? Darwin noted long ago that " Gartner further states 

 that reversions rarely occur with hybrid plants raised from species 

 which have not long been cultivated, whilst, with those which have 



1 " He [Man] often begins his selection by some half-monstrous form, or at 

 least by some modification prominent enough to catch the eye or to be plainly 

 useful to him." Origin of Species, 6th ed., p. 60. 



2 1 know of only one to which my attention was called by Sir William Thiselton- 

 Dyer. When Kalanchce flammed was crossed at Kew with K. Benin, the hybrid 

 displayed characters which can hardly have been other than ancestral traits 

 which have remained latent in one or other species (see Morphological Notes, by 

 William Thiselton-Dyer, K.C.M.G., F.R.S., Annals of Botany, vol. xvii. No. Ixvi., 

 March 1903, pp. 435-41- 



