PERMANENCE OF HYBRIDS. 587 



yet had time to disperse and multiply. This conception would not, however, accord 

 with the actual condition of affairs. In point of fact, the floral contrivances which 

 exist for promoting crosses between different species occasion a constant origination 

 of hybrids, but it is certainly not the case that they all have the prospect of becom- 

 ing new species. Many are called, but few are chosen. In only a fraction of the 

 total number of fresh plant-forms produced yearly by inter-specific crosses do we 

 find the power to survive and multiply. The first condition that must be fulfilled if 

 a hybrid is to become a species is that it be fertile, i.e. that its flowers yield seeds 

 capable of germination as a consequence of fertilization with their own pollen. By 

 " their own pollen " is here meant not only that which is developed in the same 

 flower as the stigma which receives it, or in some flower on the same plant, but also 

 pollen belonging to other plants provided they belong to the same hybrid-formation. 

 To this condition another is added in the case of dioecious, pseudo-hermaphrodite, 

 and completely dichogamous flowers, viz.: that several individuals of the hybrid 

 must make their appearance at the same time, and that of these at least one must 

 bear male flowers and one female flowers. If we suppose the case of a Willow 

 hybrid, of which all the individual-plants bear catkins of male flowers only, 

 obviously no propagation by means of fruits is possible. If none but female flowers 

 are borne, these may be crossed with the parent-species and give rise to goneoclinic 

 hybrids (cf. p. 559), and perhaps, in addition, ternary hybrids may be produced, but 

 no unmodified descendants can be expected from the fruits of a Willow of the kind. 

 The same thing applies in the case of Cirsium, the separate individuals of this genus 

 being differentiated into those bearing pseudo-hermaphrodite male flowers and those 

 bearing pseudo-hermaphrodite female flowers (see p. 294). This affords sufficient 

 explanation of the fact that although Willows and Cirsiums are continually develop- 

 ing numberless hybrids, few instances are known which one can affirm to be the 

 beginnings of new species. It usually happens, in fact, in the case of these hybrids, 

 that all the plants which arise together at a particular spot are furnished exclusively 

 either with true or pseudo-hermaphrodite male flowers, or else with true or pseudo- 

 hermaphrodite female flowers. The goneoclinic hybrids produced from the latter 

 are for the most part represented by greater numbers of individuals. Moreover, 

 amongst those individuals both sexes much more frequently make their appearance; 

 hence, they have a far better prospect of being preserved. 



The development of a hybrid into a species is also dependent on the conditions 

 determined by the habitat. When a species thrives well at a particular place, is 

 represented by a large number of individual plants, and renews itself in descendants 

 which are in the main unchanged, it may be assumed that the organization of that 

 species is suited to the soil and climate of the habitat in question. If there were no 

 such harmonious relation there could be no question of the species flourishing, but 

 on the contrary is would sooner or later die out. This suitability of the climate and 

 soil to the organization manifested in the plant's external form must also exist in 

 the case of the newly developed hybrid if the few individuals which spring up at 

 any particular place are to survive in their original settlements, and to give rise to 



