It explains at once why contagious or epidemic diseases are most fatal at their first appearance, and less so afterward: not by the dying out of a virus--for, when the disease reaches a new population, it is as virulent as ever (as, for instance, the smallpox among the Indians)--but by the selection of a race less subject to attack through the destruction of those that were more so, and the inheritance of the comparative immunity by the children and the grandchildren of the survivors; and how this immunity itself, causing the particular disease to become rare, paves the way to a return of the original fatality; for the mass of such population, both in the present and the immediately preceding generation, not having been exposed to the infection, or but little exposed, has not undergone selection, and so in time the proportion liable to attack, or to fatal attack, gets to be as large as ever. The greater the fatality, especially in the population under marriageable age, the more favorable the condition of the survivors; and, by the law of heredity, their children should share in the immunity. This explanation of the cause, or of one cause, of the return of pests at intervals no less applies to the diminution of the efficacy of remedies, and of preventive means, such as vaccination. When Jenner introduced vaccination, the small-pox in Europe and European colonies must have lost somewhat of its primitive intensity by the vigorous weeding out of the more susceptible through many generations. Upon the residue, vaccination was almost complete protection, and, being generally practised, small-pox consequently became rare. Selection thus ceasing to operate, a population arises which has not been exposed to the contagion, and of which a considerable proportion, under the common law of atavism, comes to be very much in the condition of a people invaded for the first time by the disease. To these, as we might expect, vaccination would prove a less safeguard than to their progenitors three or four generations before.