was as 1 in 36 in Holland generally, was as 1 in 22 in those villages. The additional marriages did not really increase the population. They were caused by the high number of deaths which provided openings for the living; and the high number of deaths was caused by the unhealthiness of the region and of its prevailing industries, which were manufacturing rather than agricultural. The choice in every large population is between having many lives which end soon, and few which last long. Greater healthiness in the conditions of life will result in the latter. We find as a matter of fact that, where there has been the sanitary improvement as well as simply the “replenishment” of an old country, the marriage rate goes down at the expense of the death-rate, and there is an economy of human life and suffering. Putting the parts of his exposition together, we get something like a deductive scheme of the growth of population in old countries under an industrial revolution like that of the eighteenth century. The first effect of the discovery of new minerals, and even (with some qualifications) of the invention of new machines, is to provide new employment for working men, and many new opportunities for marriage; the proportion of marriages therefore becomes at once greater without any alteration (from this cause at least) in the death-rate. But, when the first burst of progress has passed, and the succeeding improvement is not by leaps and bounds, but at a uniform rate, then the proportion of marriages will decrease, as the new situations are filled up and there is no more room for an increasing population. Once the country is really “old” in the sense of fully peopled and unprovided with new sources of employment, then the marriages will be regulated principally by the deaths, and (the habits of the people remaining the same) will bear much the same proportion to each other at one time as at another. It is not, however, exactly the same proportion for all old countries, simply because the habits and standards of living are different, to say nothing of healthiness or unhealthiness of climate and occupation. For similar reasons it is not the same for towns as for country districts.[279] “A general measure of mortality for all countries taken together” would be useless if procurable; but it cannot be procured.[280] Habits, however, are sufficiently fixed to make us certain that “any _direct_ encouragements to marriage must be accompanied by an increased mortality.”[281] They spur a willing horse. Montesquieu and Süssmilch, although they both enlarge on the evils of over-population, still think it a statesman’s duty to be, like Augustus and Trajan, the father of his people by encouraging their marriages. But, if many marriages mean many deaths, the princes or statesmen who should really succeed in this patriotic policy might more justly be called the destroyers than the fathers of their people.[282]