366 DARWINISM TO-DAY. 



alleged case of sudden appearances of modified animals or plants 

 that seem to breed true. But these infrequent prepotent sports, 

 or discontinuous variations, do little to furnish any convincing 

 foundation for de Vries's theory. Far better than all of them are 

 de Vries's own long and carefully observed primrose mutations. 

 Here, besides a few single mutations, were several that appeared 

 in considerable numbers, which is a condition almost imperatively 

 necessary for the successful propagation of a new organic type. 



A recent record of an alleged case of mutation is Schaffner's, 

 "A Successful Mutant of Verbena without External Isolation," in 

 Ohio Naturalist, Vol. VII, pp. 31-34, December, 1906. 



As Davenport ("The Mutation Theory in Animal Evolution," 

 Science, N. S., Vol. XXIV, pp. 556-558, November, 1906) puts it : "The 

 real argument for discontinuity in evolution is the occurrence of 

 characteristics in nature that are discontinuous and which never 

 show intergrades. The mere fact of discontinuity between species 

 of the same genus is not sufficient to prove that they have arisen 

 by mutation. It must be shown that the differential characters are 

 in essence discontinuous. The practical way to get at the true 

 nature of characteristics, whether continuous or discontinuous, is 

 by their behaviour in inheritance. If, in cross-breeding, a char- 

 acter tends to bjend with the dissimilar character of its consort it 

 must be concluded that the character can be fractionised and inter- 

 grades are possible. If, on the contrary, the characteristic refuses 

 to blend, but comes out of the cross intact, as it went in, the con- 

 clusion seems justified that the characteristic is essentially integral 

 and must have arisen completely formed, and hence discon- 

 tinuonsly. 



"Using this criterion, I have of late been testing the application 

 of the mutation theory to animals and have had an opportunity to 

 examine the experiments of others. Some of the work has been 

 done on the characteristics of domesticated 'races,' others on wild 

 varieties. There seems to be no difference in the behaviour of 

 characteristics of domesticated and wild varieties. The result is 

 that most characteristics, but not all, fail to blend and are strictly 

 alternative in inheritance. I interpret this to mean that the char- 

 acteristic depends on a certain molecular condition that does not 

 fractionise. The inference is that if the characteristic is incapable of 

 gradations now it has always been so and hence must have arisen 

 without gradations, i. e., discontinuously. Examples of such discon- 

 tinuous characteristics are the spots in the elytra of certain beetles, 

 the crest on the canary, the form of the comb in poultry, extra toes, 

 black plumage, and colour of iris. One who sees the striking 

 failure of these characteristics and many others to be modified in 



