RESULTS OF EXPERIMENTS. 165 



and good seeds, but the latter can vary as to number very 

 widely. As proof of his proposition respecting sexual 

 affinity, Nageli adduces many cases in which species very 

 closely resembling one another, and believed to be most 

 nearly related, cannot be made to cross, wliile they will cross 

 with species apparently far more remote. He calls attention 

 also to the fact that we may be able to cross one species (A) 

 by another (B), and yet not be able to cross B by A. For 

 instance, Gartner fertilized in five years seventy-nine flowers 

 of Nicotiana panicidata, L., by pollen from Nicotiana Lang%- 

 dorfii^ W. : sixty-six set fruit, and all had a good many seeds. 

 But he tried to fertilize N. Langsdorfii by pollen from N. 

 2Janicidata, without any effect. Kolreuter impregnated Mi- 

 rabilis Jalapa^ L., readily by the pollen of 3IiraMlis longijlora^ 

 L. ; but, in two hundred trials to fertilize M. longijiora by M. 

 Jcdapa, he utterly failed to obtain any seed. These latter 

 experiments covered a period of eight years. You will also 

 notice, in an examination of Mr. Parkman's paper, that lilies 

 afford excellent examples of this peculiarity. There are 

 some lilies which will serve only as the female parent. The 

 list of cases may be greatly extended ; but I shall only point 

 out the fact, that differences — the same in kind, though less 

 in degree — have been observed among varieties of the same 

 species ; for instance, some varieties of Indian corn. 



3. The fertility of hybrids and the number of good seeds., 

 depend upon the degree of sexual affinity. Species-hybrids 

 are, as a rule, less fertile than variety-hybrids. Without 

 dwelling upon Nageli's proofs adduced in support of this 

 proposition, I will quote a single statement in passing : " We 

 have good evidence that there are many grades in the fertility 

 of hybrid forms, and that, in this respect, there is no sharp line 

 to be drawn between variety-hybrids and species-hybrids." 



4. That the sexual affinity is greater, and therefore hybrid- 

 impregnation more readily ensues, yielding a larger number 

 of seeds, and that the hybrids therefrom resulting are the 

 more productive when self-fertilized, in those cases where 

 the parent-forms are closely allied both in structure and in 

 form, is a rule true only within certain limits. Self-imp)reg- 

 nation^ as a general thing., appears to give fewer seeds., and 

 plants of less vitality and fertility, than does cross-impregna- 

 tion with a nearly allied variety. This will be recognized by 



