two lines of thinkers or artists, then the general truth of this principle is abundantly clear.
Supposing such small functionally-produced modifications to be always taking place, it will be obvious that they must take place most in the most differentiated societies, and least in the least differentiated. A race of hunting savages will perform a certain number of routine acts, which will be for the most part the same for all members of the tribe, and will remain pretty much the same from generation to generation. In the particular direction of hunting and fishing, the cleverness at last attained will be very remarkable; but in most other directions there will be little excellence and still less variety. On the other hand, in a tribe which is also made a trading and navigating one by the accident of a maritime position, a new set of activities will be specially cultivated, and will give rise to new functional modifications in a different direction. Suppose some of the tribe, in this latter case, to be mainly inland cultivators and hunters, while others of the tribe are mainly seaboard traders or pirates, then each of these sections will tend to develop certain special hereditary brain-modifications of its own. But if a man of the inland section marries a woman of the maritime section, or vice versa, then the offspring will tend to reproduce more or less the structural peculiarities of both parents. And here comes in an important corollary. For though, under such circumstances, the children may none of them fully reproduce all the brain-gains of their father's line, nor all the brain-gains of their mother's line, they will yet on the average reproduce a fair share of the former and a fair share of the latter. Accordingly, they will usually turn out, on the whole, persons of higher general brain-power than either ancestral series; they will partially unite the strong points of both.
It seems to me that this principle is one of very great importance. From it we can deduce the conclusion that in any complex society many children represent directly a convergence of two unlike lines of descent, and indirectly a convergence of innumerable unlike lines, with corresponding gain to the species. Two parents, possessing distinct points of advantage of their own, produce children, some of whom resemble rather the one, and some the other; but many of whom will at least tend to resemble both in their stronger points. Of course, one must allow much for the idiosyncrasis as well as for the crasis. This child may fall below both its parents in most things; that child may reproduce the weakest elements of both; yonder other child may attain the average or may surpass them in everything. But, on the whole, the principle of convergence seems to imply that in a fairly complex society there will always be an average of mental improvement from generation to generation, due to the constant intercrossing of brains specially improved in particular directions. This improvement will, it need hardly be said, be increased and favored by natural