162 HORTICULTURAL SOCIETY OF NEW YORK. 



contact, losing all affinity for other nuclei, and start at once a new life, the 

 infant germ soon to become the mature seed. In some of my accidental 

 hybrids, at times, when trying to analyze them by their specific markings, it 

 would seem impossible to explain them as resulting from only two of any 

 known species, some of them apparently embodying three and four distinct 

 sets of specific markings, while their known mothers were apparently pure 

 representatives of only one species, which stood surrounded by numerous 

 other species, several flowering at the same time. Yet as nearly all our 

 species of grapes show that they contain traces of mixture with other 

 species at some distant past, reversion may explain these apparent complex 

 hybrids by one generation. The theory of mutation, which appears to de- 

 mand a causeless effect, I dare not appeal to until better proven. I throw 

 out these thoughts for some of our younger, more scientific hybridizers to 

 work upon. 



If it should be found that double or treble fecundation can occur with 

 the female nucleus, and can be practiced by the hybridizer, he could accom- 

 plish in one lifetime what would take generations to do by single pollination. 

 The rate of development would be high geometrical, instead of arithmetical, 

 or a low geometrical speed, as by present methods. 



THE LIMITS OF USEFULNESS IN HYBRIDIZATION AMONG GRAPES. 



In considering the first part of our subject, the objects to be sought were 

 briefly pointed out. In general, we may say, that when all those objects 

 served by hybridization of pure species are reached, we should cease such 

 direct intermingling of pure species and proceed to select and cross-pollenize 

 among the varieties already produced, to constitute a separate family, until 

 we could work it into more desirable new varieties, all of a sufficient 

 homogeneity to give a fixedness of type, that will yield uniform results, as 

 in types of stock breeding (although live stock breeding does not make a per- 

 fect comparison, on account of lack of species involved, unless we should 

 include in it several species, as the musk-ox, the bison, the Indian and 

 African cattle, etc., in the genus Bos). 



By thus cross-pollinizing we eventually can build more wonderful families 

 of grapes for every season and use than the Old World has ever known 

 among its thousands of varieties. 



There are some fundamental facts with reference to pure and mixed 

 blood that, to keep in mind, will greatly aid the hybridizer in the proper 

 conduct of his work of development. 



The purer the specific blood, known by the greater uniformity of the 

 individuals of the species, and especially season of flowering, character and 

 ripening of fruit, the more persistent to its type will it be in cultivation and 

 in showing forth in hybrid forms. The hybrids vary much less from the 

 parent taken from the homogeneous species than they do from the parent 

 from the variable species. This law, or method of action, is well illustrated 

 in quite a number of hybrids of V. rotundi folia (one of the most homogeneous 

 species), with less homogeneous species and their hybrids, produced by me. 

 Several French hybridizers, among them A. Millardet* of the Faculty of Bor- 



*Deceased since this paper was read, in 1903. 



