Improvements Made in Plants by Man 441 



reproduce at all by seeds and only badly by bulbs. Such of these 

 species hybrids, however, as can be reproduced by cuttings or 

 grafting can be preserved indefinitely, and this is the case with 

 hybrid trees, because the seedling can be grafted into either of the 

 parent trees, or into some allied kind, and thereafter can be multi- 

 plied with rapidity and certainty. It is obvious from these con- 

 siderations that specific hybrids can only rarely be made the 

 foundation of a race, and equally plain that they can pla}- no 

 appreciable part in the natural evolution of plants. 



(6) He can pollinate a given stigma by pollen from a plant of 

 another but allied genus. Though some such generic hybrids 

 have been made as a matter of scientific experunent, this has 

 only been possible with genera extremely closely related, and 

 moreover the hybrids are unstable and of no horticultural im- 

 portance. Nor have any attempts at hybridization over wider 

 limits ever succeeded; and the occasional newspaper accounts of 

 crosses between members of different plant families are lies, when 

 they are not obvious jokes. 



It is evident from this discussion that plant-breeders make use 

 of in-breeding, crossing, and hybridization for various purposes in 

 accordance with the results which they wish to attain. By suit- 

 able combinations it is possible to keep races close to their type 

 and thus preserve desirable characteristics; to break the type and 

 thus provide a basis for the development of new characteristics 

 through selection; to eliminate undesirable features out of a race; 

 to combine the desirable features of two or more races into one; 

 and in general to promote vigor and productivity. I think it will 

 now be evident why crossing and hybridization are so prominent 

 in plant improvement. 



It will interest the reader, at this point, to learn in what way 

 crossing and hybridization are effected in practice. It is no 

 trouble at all to transfer the pollen from any ripe stamens to any 

 ripe stigma. It is only necessary to pick the fine pollen dust from 

 the opened anthers by some dry utensil to which it will cling (for 



