Production of Neiv Varieties. 23 



made from the first of these, and so on in continued succession ; 

 the best and soonest in bearing were uniformly chosen. He thus 

 obtained fruit from the eighth generation ; each successive experi- 

 ment yielding an improved result on the preceding. He had, in the 

 early part of this series of experiments, no less than eighty thousand 

 trees ; hence, in selecting from so large a number, his chance for 

 fine sorts was far greater than from a small collection ; and hence 

 too the reason why, after seven or eight improving generations, he 

 had obtained so many good varieties. In the early stages of his 

 operations, he found "that twelve or fifteen years was the mean 

 term of time, from the moment of planting the first seed of an 

 ancient variety of the domestic pear, to the first fructification of the 

 trees which sprang from them." When his seedlings were at the age 

 of three or four years, he was able to judge of their appearances, 

 though they had not as yet borne ; such only were taken for further 

 trial as exhibited the strongest probability of excellence. It is 

 hardly necessary to remark that in all these trials, the young trees 

 were kept in the highest state of cultivation. 



Van Mons maintained that by selecting and planting the seeds of 

 the first crop on the young tree, the product would be less liable to 

 run back to the original variety than where the seeds were taken 

 from the fruit of an old-bearing or grafted tree ; and to this practice 

 he chiefly ascribed his success. The many instances, however, of 

 fine seedlings from old grafted sorts, throw a shade of doubt over 

 this theory. There is scarcely a question that the same extent of 

 labor expended in crossing varieties, would have given greater 

 success. 



NEW VARIETIES BY CROSSING. 



New varieties are produced in crossing by fertilizing the stigma of 

 one with the pollen from another, as described in the preceding chap- 

 ter. The simplest instance which occurs is that of the strawberry, 

 the pistillate varieties of which must always be impregnated with 

 pollen from staminate sorts. Thus the seed obtained from the ber- 

 ries of every pistillate strawberry are crosses, and if planted will 

 produce new varieties. In fruit-trees, the stamens and pistils are in 

 the same flower, and the chances of accidental mixture from other 

 trees become very small, unless effected by insects, which, becom- 

 ing thickly dusted with powder from one flower, plunge into the 

 recesses of another, and effect a cross-fertilization. Where many 

 varieties grow in one garden, in close proximity, cases of promiscu- 



