94 



The Journal of Heredity 



Table II. — Line No. 23; Per Cent of Spikehls ivith T'lco Kernels 



barley. As in my recent trials with 

 oats and my earlier trials with peas, a 

 performance-character or internal char- 

 acter was chosen, viz., the protein con- 

 tent of the kernel, a character that is 

 quantitatively modifial^le to a very 

 high degree. Five generations of ptire- 

 line selection to increase or diminish 

 this character failed to produce any 

 change. Some transmission, in the 

 sense in which I ha\-e just defined it — 

 what is also called nachwirkung — after- 

 effect — has been found but no real 

 lUfxlification of an inherited factor. 



Critical scientists may, i)erhaps, object 

 that if these various experiments had 

 been continued much longer, some 

 hereditary effects would be produced. 

 In such a case, I admit that changes 

 could occur; but I am of the opinion that 

 thev would not 1k' tlif effects of sek'c- 



tion. They would be mere spontaneous 

 variations (mutations) whose i)roduc- 

 tion would be wholly independent of 

 the selection; but whose preservation 

 would naturally be result of selection. 

 Com])aring these experiments \\'ith 

 the results of Hopkins and Smith, I 

 think the practical breeder may con- 

 clude that there is a marked difference 

 l)etween selecting in a self-pollinated 

 plant and selecting in a cross-pollinated 

 l)lant. In the selection ^\^th corn — a 

 plant exceedingly subject to cross- 

 fertilizalion — there is a marked liercdi- 

 tary effect. There is no hereditary 

 effect in a self-fertile j)lant, such as I 

 iiave used — save for a little transmission 

 of immediate parental influence, which 

 is not hereditary and cannot be made 

 the basis for a change in the character 

 of the pure h'ne. 



