Principles of selection.
Selection is the principal factor of breeding, both in the improvement of races and in the pro duction of new races or varieties. The keynote of selection is the choice of the best, and a factor of the highest importance is the examination of very large numbers in order to secure the maximum. Galton, writing on this subject, says : "One gene ration of 99-degree selection is seen to be more effective than two generations of the 90-degree selection, and to have about equal effect with the the 80-degree selection, carried on to perpetuity. Two generations of the 99-degree selection are more effective than four of the 95-degree, and than the perpetuity of the 90-degree." The use of de grees in representing the perfection in which a character is shown may not be possible, but it is possible for any breeder to examine large numbers and to find one or two plants which produce in the greatest degree the character desired. It is these plants that should be preserved as mother plants in starting the selection.
In the production of new races, it is of interest to us to know whether by pure selection we can lead plants to vary so greatly that they may be considered to have passed beyond the bounds of the race, and thereby the breeder to have estab lished a new and distinct race. It is certain, of course, that, by careful observation and selection from any particular race, ultimately a new race may be produced. The question is whether the individual or individuals selected in producing the new race have not varied by mutation or seed sporting rather than being simply representative of the cumulative result of the selection of slight individual variations. The sugar-beet furnishes an interesting illustration in this direction. It will be remembered that Louis Vilmorin started the selec tion of sugar-beets for richness in sugar, between 1830 and 1840, selecting first by means of specific gravity, the method being to throw the beets into solutions of brine strong enough so that the great majority of them would float, the few which sank being of greater specific gravity and presumably of greater sugar content. Considerable improve
ment was produced by this method. About 1851 the method of chemical analysis was introduced to determine the exact sugar content. At this time the sugar content was found to vary from 7 to 14 per cent, and in the second generation of selection individuals with 21 per cent of sugar were found. The selection based on sugar content, using the beets highest in sugar content as mothers, has been continued regularly since that time, and the indus try has come to rely entirely on careful selection for high sugar content. It would be expected that under these conditions the sugar content would have increased sufficiently so that the selected plants could be considered a different race or strain. Yet, after fifty years of selection, the highest sugar content found is only about 26 per cent, and this in a very few instances, seldom over 21 per cent being found. At the present time many thousand analyses are made every year, so that abundant opportunity is afforded to find individuals producing a high sugar content. On the contrary, when Vilmorin's work was started the determina tion of sugar content was by very laborious meth ods, and was limited to comparatively few indi viduals. It is not improbable that if Vilmorin had been able to make analyses of the sugar content in many thousands of roots he would have found cer tain individuals producing as high as 26 per cent. The inference from this illustration would be that the limitations of the variation within the race have not been surpassed as a result of selection. It may be argued, however, that in this case we are dealing with a physical impossiblity, as it is clearly evident that it would be impossible for a plant to produce a root containing a proportion of sugar beyond a certain percentage, and it is thus possi ble that 26 per cent, or thereabouts, represents the maximum.