LINNBAN 800IBTT OF LOKDON. XI 



and the authors of some of our most carefully prepared modern 

 European floras have thought that, by external characters alone, 

 they could not only recognize a hybrid, but even determine Avhich 

 parent supplied the male and which the female element, and have 

 accordingly adopted a nomenclature expressive of their decision 

 as of a proved fact. As a test of specific identity and diversity, 

 the relative perfection of the hybridizing power is also still re- 

 garded by them as absolute, in opposition to the conclusions of 

 Darwin. It may therefore be useful to consider how far the pro- 

 positions, on this head, of that distinguished biologist have been 

 confirmed or modified by subsequent observers, and what are the 

 points which are specially in need of fiu"ther illustration. 



The principal facts, bearing on this question, established in the 

 admirable chapter on Hybridism in the ' Origin of Species,' appear 

 to be : — 



That generally, in order to produce the most fertile offspring, 

 that is, beings most perfect in constitution as to the reproductive 

 system, there must be near affinity, although not always too close 

 a consanguinity, between the two parents. 



That generally, as the affinity is more remote, the fertilizing 

 process is less successful, not ceasing abruptly at a definite stage 

 of remoteness, but diminishing by various gradations. 



That generally, also, the fertility of the offspring of plants or 

 animals nearly enough related to be considered as slight varieties 

 will be unimpaired, but frequently diminishes if the parents be- 

 long to more constantly distinct races. "Where the distinctions be- 

 tween the parents have obtained constancy and importance enough 

 to be admitted as specific, iuterbreeding usually produces sterile 

 offspring, and all power of interbreeding ceases as the differences 

 stiU further increase so as to be deemed generic or ordinal. 



But that exceptional cases more or less interfere Avith any uni- 

 versal regularity in any of these rules. 



And that, although the generality of these rules has practically 

 the effect of maintaining the distinctness of wild races once sepa- 

 rated by a wide interval, yet that their indefiniteness — the nume- 

 rous exceptions observed — forbid our considering them as a special 

 provision for the preservation of any fixed relations of varieties, 

 races, species, or genera. Nor can the greater or less perfection 

 in the power of interbreeding, or the absolute want of it, be taken 

 as positive tests of affinity or diversity. They are constitutional 

 characters, standing more or less in correlation wdth external and 

 structural characters, as these do with each other ; and their value 



