Hume, in his essay on the populousness of ancient and modern nations, when he intermingles, as he says, an inquiry concerning causes with that concerning facts, does not seem to see with his usual penetration how very little some of the causes he alludes to could enable him to form any judgement of the actual population of ancient nations. If any inference can be drawn from them, perhaps it should be directly the reverse of what Hume draws, though I certainly ought to speak with great diffidence in dissenting from a man who of all others on such subjects was the least likely to be deceived by first appearances. If I find that at a certain period in ancient history, the encouragements to have a family were great, that early marriages were consequently very prevalent, and that few persons remained single, I should infer with certainty that population was rapidly increasing, but by no means that it was then actually very great, rather; indeed, the contrary, that it was then thin and that there was room and food for a much greater number. On the other hand, if I find that at this period the difficulties attending a family were very great, that, consequently, few early marriages took place, and that a great number of both sexes remained single, I infer with certainty that population was at a stand, and, probably, because the actual population was very great in proportion to the fertility of the land and that there was scarcely room and food for more. The number of footmen, housemaids, and other persons remaining unmarried in modern states, Hume allows to be rather an argument against their population. I should rather draw a contrary inference and consider it an argument of their fullness, though this inference is not certain, because there are many thinly inhabited states that are yet stationary in their population. To speak, therefore, correctly, perhaps it may be said that the number of unmarried persons in proportion to the whole number, existing at different periods, in the same or different states will enable us to judge whether population at these periods was increasing, stationary, or decreasing, but will form no criterion by which we can determine the actual population.