tongued will fail where a number of the longer-tongued succeed. As every creature multiplies up to the limits set by the means of subsistence, herds of giraffes must be from time to time underfed. At such times the short-tongued ones must be more underfed than the long-tongued ones. The difference of feeding may not be such as to produce in a direct way greater mortality in the one class than in the other, but it may readily tell indirectly. Especially will there be more deaths of the weaker adults, and the less vigorous young they have produced, when the herd is chased by carnivores. Those which are a yard or two behind the rest lose their lives; and a very small defect in the constitutional state of the adults, or the strength of the young, may entail the slight difference in speed implied. So that, other things equal, more of the short-tongued and their offspring will die than of the long-tongued and their offspring. Hence, without any special choice of mates, it will result that in the next generation the average of length of tongue will be greater. Through subsequent generations the same process will go on increasing this advantageous variation, until some limit is reached at which disadvantages check it, or at which the life-sustaining advantages from some other variation become greater.
So that in the absence of any such improbable events as those Lord Salisbury supposes to be necessary, there are two co-operating ways in which survival of the fittest establishes in a species a useful modification of structure.
The great length of time required for the production of species by the evolutionary process, is supposed by Lord Salisbury to furnish a reason for disbelief. In support of his argument he cites Lord Kelvin's conclusion that life can not have existed on the Earth more than a hundred millions of years. Respecting Lord Kelvin's estimate it may be remarked that the truth of a conclusion depends primarily on the character of the premises; that mathematical processes do not furnish much aid in the choice of premises; that no mathematical genius, however transcendent, can evolve true conclusions out of premises that are either incorrect or incomplete; and that while putting absolute faith in Lord's Kelvin's reasonings, it is possible to doubt the data with which he sets out. Suppressing criticism, however, let us accept in full the hundred million years, and see what comes of it. Lord Salisbury argues:—
