1873.] 



Mr. F. Galton on Blood-relationship, 



399 



animal with respect to any particular spot are, I will not say in every case, 

 but certainly on the average of many cases, much more numerous than if that 

 spot had been purely a personal characteristic, without the concurrence of 

 any latent elements. Bearing this argument in mind, we shall more justly 

 estimate the import of the statistical evidence to be obtained from breeders of 

 animals. I should judge, from the impression left 

 by many scattered statistics, that it is perfectly 

 safe to affirm that breeders, when they mate two 

 animals, each having the same unusual charac- 

 teristic, not through known hereditary trans- 

 mission, but by supposed variation, would con- 

 sider themselves fortunate if one quarter of the 

 progeny inherited that quality. Now these suc- 

 cessful cases are, as I have shown, on the average, 

 the produce of parents having the peculiarity 

 not only in a personal but also, to some degree, 

 in a latent form. We may therefore reasonably 

 conclude that, had the latter portion been non- 

 existent, the ratio of successful cases would have 

 been materially diminished. 



I should demur, on precisely the same grounds, 

 to objections based on the fact of the transmis- 

 sion of qualities to grandchildren being more 

 frequent through children who possess those 

 qualities than through children who do not ; for 

 I maintain that the personal manifestation is, on 

 the average, though it need not be so in every 

 case, a certain proof of the existence of some 

 latent elements. 



Having proved how small is the power of 

 hereditary transmission of the personal elements, 

 we can easily show how large is the transmission 

 of the purely latent elements, in the case (3), by 

 appealing to the well-known facts of Reversion ; 

 but into these it is hardly necessary for me to 

 enter at length. The general and safe conclusion 

 is, that the contribution from the patent ele- 

 ments is very much less than from the latent 

 ones. 



If we now combine our results into a diagram 

 (fig. 2), showing the fainter streams of hereditv 

 by italic lines, and indicating those processes by 

 asterisks (*) which were described at length in 

 the previous figure, we shall easily recognize the complexity of hereditary 



.- u 



In 



* 



p 



<- o 

 c - 



1-3 



t 1 j; 



