498 TRANSMISSION 



is not now growing less (see chap, xii) leads to the conclusion 

 that the upper limits in this breeding experiment will be set by 

 some factor other than the failure of variability ; indeed, it is 

 the opinion of the writer that the principle of progression is able 

 always to afford all the material the breeder will need, and that 

 the limits of improvement will be set, when they are set, by 

 some biological, mechanical, or other considerations entirely 

 aside from the failure of variability to present new upper values 

 on which to base selection. 1 The remarkable fact about progres- 

 sion is that the distributions are not distorted but are all typical 

 of the race (see table, page 496). 



This fact of progression betrays a unique principle in heredity, 

 or rather in variability, because progression is over against and 

 in spite of the " drag of the race." Those individuals that have 

 overleaped the limits of the race have not only exceeded their 

 own parents, but by much more have they exceeded the compara- 

 tive mediocrity of their other ancestors. 2 Progression cannot, 

 therefore, be explained by any principle of " mean parentage." 

 It rests upon a principle fundamentally distinct, and is to be 

 regarded as the result of those fortuitous combinations of physio- 

 logical units which we may expect to occur from time to time in 

 the complicated processes attending reproduction and differenti- 

 ation, and on which more will be said in Section XII of this 

 chapter. 



This suggests what will be later found to be a fact, namely, 

 that in a large sense heredity follows the laws of probability, and 

 in process of time we may expect all possible combinations of the 

 elements that make up characters, and of the characters that 

 make up the race, the largest proportion of which will cluster 

 about a common point, which we call the type, but a few of 

 which will inevitably appear at the extreme limits of the range 

 of possibilities. 



1 The greatest menace to extreme improvement is, as has already been said, 

 lessened fertility. According to Pearson all evidence points to the fact that varia- 

 bility will never be reduced by more than about 1 1 per cent. See Grammar of 

 Science, p. 483. 



2 The student will not fail to note that the ancestors of the exceptional parent 

 are of necessity more mediocre than that parent. They will then exert their influ- 

 ence against, not in favor of, progression. 



