262 JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 
alone or in combination, express themselves as old or newly compounded 
forms or qualities—often gives the best materials for the plant breeder to 
work upon. Whether the cross is radical or mild, the centrifugal force 
seems to reach its climax in a few generations ; but the tendency to vary 
ofttimes continues for many generations, differing with the individual 
plant in close fertilised species, and with pairs of plants mated in open or 
insect fertilised species, or in case of artificial fertilisation. 
The practical point is that it is often wise to give free rein to varia- 
tion for at least a few generations. Not all the resulting plants need be 
kept, but it is wise to keep very many strong ones or those having special 
characteristics. Largeness of numbers in the progeny requires system 
in the selection, that the intrinsic qualities may be sought after in the 
most effective manner, and that the labour be not wasted upon many 
useless forms. The intrinsic or special qualities should first be sought, 
and then the stocks should be bred to uniformity of type afterwards. 
Much stress has been laid upon having the forms uniform in all their 
characters—-in matters of appearance, if you please—as well as in matters 
pertaining to special values. No doubt the longer time required to fix 
several characters makes the variety more stable ; but if the greater labour 
be nearly all directed for the longer time to the intrinsic or special 
qualities sought, no doubt time and labour would be expended to more 
purpose. ' For practical purposes many new hybrids produced from seeds 
are fixed in type by rigid selection for only three to six generations; and 
the stronger growth or fruiting character of most food-producing plants 
may be fixed in a comparatively short time, whether open or close 
fertilised, while in the case of plants propagated by division they are at once 
sufficiently well fixed in character for practical uses. In such cases as 
the open-fertilised Sugar-Beets, where the seeds are raised in one place 
and sold in many lands, the longer time spent in fixing characters, the 
attention to fixing the permanency of form of leaf and other non-intrinsic 
qualities, doubtless has much weight; but in the case of grains, forage 
crops, &e., bred in the general locality in which they are to be used, the 
stability of the characteristics of large yield, good quality, and hardiness 
is usually provided for by rigid selection during only a short term of 
years. The American farmer has wisely sought the large sound ears of 
corn, caring little whether the cobs were all of one colour or the leaves of 
similar form. 
Since selection affects crops to which, in the aggregate, are attached 
immense value, the selection should be from among such large numbers 
that the greatest increase in yields and qualities will surely result. M. 
Vilmorin and other breeders of Sugar-Beets have fully recognised this 
most important principle. Here, quality of sugar per acre, together with 
its availability, are the requisites, and these experimenters early took a 
direct road for results. They dealt first with varieties, securing the best ; 
then they sought the best roots; and then they tested the capabilities of 
each mother root in the ability of its seeds to produce roots with high 
yield of sugar and high purity of juice—in otber words, its prepotency. 
With the best roots from the best mother plants, the process was repeated 
indefinitely, each season seeing a gain of one-fourth to one-half per cent. 
of sugar, and little, if any, lessening of the gross yield per acre of the crop 
