368 ACTION OF NATURAL SELECTION 



The upper of the two accompanying figures represents 

 a (roughly) normal curve of distribution of 324 meas- 

 urements (represented by dots and crosses). As is in- 

 dicated on the base line, these vary in size from the 

 general mean by from 1 to 10 per cent. Sup- 

 posing that the larger individuals (the crosses) were 

 better adapted to their environment, and were being 

 gradually selected by the agency of Natural Selection, 

 whilst the shorter ones were gradually being weeded 

 out, then it is probable that the selective process would 

 act only very slowly, as the range of variation is so 

 slight. Thus differences of 2 or 4 per cent, from the 

 average are so small as to be almost inappreciable, 

 whilst greater differences, as of 7 or 8 per cent., are 

 exhibited by only a very small proportion of the whole 

 (only 12 out of the 324 measurements being 7 per cent, 

 greater than the average, and only 6 of them 8 per 

 cent, greater). Now let us suppose that this group of 

 324 individuals is acted on by a variable environment, 

 so that the slight range of blastogenic variations is en- 

 hanced by the superposition of somatic modifications. 

 Out of every ten organisms of any particular size, let 

 one be increased by 2 per cent., another by 4 per cent., 

 another by 6 per cent., another by 8 per cent., and an- 

 other by 10 per cent., owing to the action of a favour- 

 able environment, whilst the other five of the ten 

 organisms are diminished by similar amounts, owing to 

 the action of an unfavourable environment. The lower 

 figure shows the new distribution of the organisms ac- 

 cording to their altered magnitudes. For instance, of 

 the 40 crosses representing individuals 1 per cent, 

 larger than the average, 4 are increased by 2 per cent., 



