17 
there will be variations in some characteristics which will 
produce little alteration in the whole vitality. Thus 
amongst wild oxen probably no varieties without horns 
would exist, for they affect the vitality. Amongst pro- 
tected races they do not, and so hornless varieties arise. 
Still these varieties are but varieties, and are not steps 
towards a new maximum which a gulf of lesser vitality 
still separates them from. 
Or let us consider the varieties that we try to make by 
select breeding. These are least of all likely to produce 
new species. We simply by main force depress vitality 
in removing individuals as far as we can from the normal 
type, and when the vitality is sufficiently depressed we can 
go no further. As for altering the province, the inde- 
pendent variables, so to speak, we know so little how to do 
it, and certainly could not do it gradually enough, that we 
have no chance in this way of effecting anything. 
How then can new species arise ? Apparently in some such 
way as this, by what we may call the bifurcation of a 
maximum. If we drew a horizontal line along which the 
variation of the organs of an animal were expressed and the 
corresponding vitality were drawn by ordinates, we should 
get a curve we might call the vitality curve whose maxima 
values would be species. As time elapses and the conditions 
of the earth, &c., altei’, the constants, so to speak, of the 
curve alter, and we get our curve to vary and the maxima 
shift; and as the curve alters, one maximum may separate 
into two or more others, and thus inthe lapse of time one species 
may separate into two or more others. Roughly to illustrate 
it, suppose some species developed free from the influence 
of carnivora, and that,owingto various causes, size little effects 
its vitality, it may vary all through, from little and swift to 
big and heavy. Now, introducing carnivora, we can see how 
a bifurcation of our maximum would take place. The very 
light and swift would preserve themselves by their agility, 
the strong and heavy by their strength, whilst the inter- 
