23 FLORA INDICA. 



number of species capable of being impregnated even by skil- 

 fal management, is very few ; and in nature the stigma exerts 

 a specific action, which not only favours and quickens the 

 operation of the pollen of its own species, but which resists and 

 retards the action of that of another ; so that the artist has 

 not only to forestall the natural operation, but to experience 

 opposition to his conducting the artificial one. 



2. Even when the impregnation is once effected, very few 

 seeds are produced, still fewer of these ripen, and fewest of aU 

 become healthy jolants, capable of maintaining an independent 

 existence ; this is a very important point, for under the most 

 favourable influences the average number of seeds that are 

 shed by a healthy plant in a state of nature come to nothing, 

 chiefly owing to the pre-occupation of the soil and the wants 

 of the animal creation. 



3. The offspring of a hybrid has never yet been known to 

 possess a character foreign to those of its parents; but it 

 blends those of each, whence hybridization must be regarded 

 as the means of obliterating, not creating, species. 



4. The offspring of hybrids are almost invariably absolutely 

 barren, nor do we know an authenticated case of the second 

 generation maturing its seeds. 



5. In the animal kingdom hybrids are still rarer in an ar- 

 tificial state, are all but unknown in a natural one, and are 

 almost invariably barren. 



On the other hand, it is often argued that hybrids are com- 

 mon in gardens, and that their occurrence in a state of nature 

 cannot be denied; and that if the permanence of one such 

 hybrid be admitted, the whole fabric of species is shaken to its 

 foundation. Such summary conclusions are however opposed 

 to philosophical caution : the whole subject is one that cannot 

 be cleared up by a consideration of exceptional cases ; it must 

 be argued upon broad principles, and unfortunately no argu- 

 ment has ever been adduced that has not been taken in evi- 

 dence on both sides of the question. This is especially the 

 case with hybridization, which, in so far as it can produce a 



