DARWINISM ATTACKED. 71 



advocates of the selection theory have to admit, says Wolff, 

 that they face a serious matter here. Weismann 3 is quoted 

 as follows : "Wenn man sich die Umwandlung deshalb in 

 grosseren Schritten und durch Variationen von qualitativer 

 Natur geschehend denkt, so wird man uber dieses Hindernis 

 nicht wegkommen. Ich glaube aber, dass man von den 

 Variationen grosseren Betrages, wie sie bei domestizierten 

 Tieren und Pflanzen nicht selten vorkommen, bei den Pro- 

 zessen der Artumwandlung, wie sie in der freien Natur vor 

 sich gehen, vollstandig abgesehen hat, dass hier uberhaupt 

 nicht qualitative sondern nur quantitative Unterschiede der 

 Individuen das Material der Naturziichtung bilden, solche 

 aber sind immer vorhanden." Wolff holds that there can 

 simply be no doubt that if natural selection can do any 

 modifying at all it has at most to limit itself solely to work 

 possible on a basis of quantitative variation of already ex- 

 isting structures. 



Variation also of whatever kind is subject to Galton's 4 law 

 of regression. This is, put briefly, that the young of 



Galton's law parents varying from the mean of their species 

 of regression, or race tend to vary also in the same direction 

 but less so than the parent, so that the mean or mode among 

 the young is nearer the species or race type (mean or mode) 

 than the parental type was. Or as Morgan 5 has stated it : 



'The facts of observation show that when a new variety ap- 

 pears its descendants are more likely, on the average, to 

 produce proportionately more individuals that show the same 

 variation, and some even that may go still farther in the 

 same direction. If these latter are chosen to be the parents 

 of the next generation, then once more the offspring may 

 show the same advance ; but little by little the advance slows 

 down, until before long it may cease altogether. Unless, 

 then, a new kind of variation appears, or a new standard of 

 variation develops of a different kind, the result of selection 

 of fluctuating variations has reached its limit. Our experi- 



