SOME ASPECTS OF ZOOLOGY 43 



made clear directly, he gradually abandoned his belief in 

 the accumulation of variations by Natural Selection, and 

 laid more stress on the use and disuse of organs and the 

 direct and indirect effects of external conditions in producing 

 variation. 



Time is getting short, and it is impossible to give the 

 evidence at length. But I may say at once that it is proved, 

 beyond all doubt, that the small individual variations relied 

 upon by the selection theory are not unlimited. 



The late Sir Francis Galton, by the application of exact 

 statistical methods to the study of Heredity and Variation, 

 showed that, with regard to any one character, the range of 

 variability within the limits of a race is confined within 

 certain limits, which may be expressed by a curve of prob- 

 ability. Further, that the offspring of an individual chosen 

 from one extreme of the curve do not inherit to the full the 

 deviation of the parent from the normal, but tend to regress 

 towards the latter. This " law of regression " has been so 

 well established that I need not say any more about it, except 

 to give a striking instance taken from de Vries' famous work 

 on the " Mutation Theory." 



As long ago as 1851 the famous Belgian cultivator, Louis 

 Vilmorin, began a long course of improvement of the sugar- 

 contents of the sugar-beet by selection. He found that the 

 saccharine contents of single roots varied from 7 to 

 14 per cent ; and selecting seed only from the best beets, 

 in the second generation had some beets with 21 per 

 cent of sugar. Bv continued selection of seed from the best 



w 



beets, aided by greatly improved methods of determining the 

 sugar-contents, he and his followers have been able to raise 

 the average contents to a considerable extent, from 8 per cent 

 to 16 per cent, but this only by a rigorous selection of the 

 best plants in every generation. But they have not succeeded 

 in overstepping the upper limit of variability, namely, 21 

 per cent, a result obtained in the earliest days of the 

 experiment more than sixty years ago ! 



All these experiments on beets were undertaken for com- 

 mercial purposes. I will now give you a famous experiment 



