IMPROVEMENT OF THE POTATO. 399 



Hence if we consider a potato variety as a perennial individual, 

 though divided, we may conclude that following these laws the 

 variability lessens as the variety becomes older. 



Admitting this law to be sufficient reason for lessened variation 

 in the variety as it ages; there is still variation, and as Bailey (4) 

 has shown considerable variation, both continuous and discontinu- 

 ous, or with De Vriesian names, both fluctuations and mutations. 

 In the progress of evolution, such variations must have been suf- 

 ficient either as mutations or as accumulated fluctuations to have 

 created varieties and even species. In no other way could the 

 numerous species and varieties of the asexually propagated fungi 

 have originated, as well as numerous varieties of higher plants of 

 various families, as sugar cane, banana, weeping willow, sweet po- 

 tato, olive, fig and date which seldom or never are propagated by 

 seeds. But as most biologists now accept the doctrine of discontin- 

 uous evolution we cannot a priori conclude that partial fluctuations 

 (using the terms described below) are inherited even in the tem- 

 porary Galtonian way in which individual fluctuations are inher- 

 ited. De Vries (27) divides fluctuations into two heads which he 

 says "obey quite the same laws," but which with respect to ques- 

 tions of heredity should be carefully separated. "They are desig- 

 nated by the terms 'individual' and 'partial' fluctuation. Individual 

 variability indicates the differences between individuals, while par- 

 tial variability is limited to the deviations shown by the parts of 

 one organism from the average stature." Fluctuations, he says, 

 take place in only two directions, the increase or decrease of what 

 characters are already available, and in this way are fundamentally 

 different from mutations which take place in all directions, and if 

 progressive produce new characters. He concludes that partial fluc- 

 tuations are usually far smaller than individual and partial fluctua- 

 tion together, and that partial variations do not appear to offer im- 

 portant material for selection.* Multiplication by buds, however, 

 of high extremes of individual fluctuation, he says, is what the 

 breeder desires to obtain. 



From De Vries' work, we might conclude that although partial 

 fluctuations obey the same laws as individual fluctuations, there is 

 not a great chance for improvement through their selection, because 

 of their narrowness. Theoretically the fluctuations of the whole of 

 any variety of potatoes belong to this class. Still the variability 



*De Vries, however, admits the possibility of the commercial value of the selection of par- 

 tial variations, when he says (P. 766): "Potatoes for the factory have even been selected for 

 their amount of starch, and in this case at least, fluctuating- variability has played a very im- 

 portant part in the improvement of the race." This is an admission of something- that cannot 

 be regarded as an undisputed fact. E. M E. 



