463,040, and in the next 474,603, the mean number being, consequently, 468,821; whereas, in the most healthy years, in which there were only 1,274,282 deaths, the marriages amounted to 470,131. In different other countries, whose registers he gives for a series of years, I find the deaths in the most mortal years in the several divisions to amount to 895,269, in which years 232,879 marriages were contracted; in the year ensuing these, 269,247, mean number 251,063; whereas the average number of marriages for the same periods is 261,331; and those years in which the least room" had been made, were again precisely those in which most had crowded into that state, the deaths being only 846,134, the marriages 276,392. In towns in which we are assured the preventive check is the most operative, (another deduction, by-the-bye, of this reasoning system, which is totally contradicted by facts,) in the years of the greatest mortality, when the deaths amounted to 371,954, the marriages were 78,179; those of the year following 81,531, the mean number 79,855, being 5,044 under the number of the celebrations in the healthiest years, in which 289,021 only had died.

(5) These numbers, if correctly stated, it will be admitted, are quite sufficient to set this controversy at rest; yet, justice to the argument requires that these results should have further consideration, in order to give them their due weight. As, in selecting from every section in the whole tables, where the years are regularly given, that, in the first place, in which the greatest number of deaths occurred, it must be very obvious, that in an increasing population, which seems to have been the case in most of the instances given, supposing that increase to be regular, the largest number would naturally fall in the last year of each division, and

would have, of course, the largest number of marriages opposite; on the contrary, the smallest number of deaths would be expected to stand in the first year of every section, attended by the smallest number of marriages this difference between the marriages and deaths of the first and last years being regulated by the rate of the increase of the population. Susmilch usually divides his table into averages of five years; and supposing 1 per cent. to be the annual rate of increase in any country, the registers of whose population he has given; 5 per cent., it is clear, would be the difference between the number of the marriages and deaths in the first and those in the last year of each lustrum. Now, without meaning that this natural arrangement of the numbers constantly occurs, I contend that there is always a " tendency" to it; such is obvious on the slightest inspection of the tables, and must, indeed, be the fact on the whole, wherever population is advancing. It is clear that some calculations should be made on this principle, if we mean to deal fairly by the proofs brought forward. If, however, they already substantiate the facts at issue, without the considerable addition which is due to this increase throughout, the results must have been still more striking, had the correction been supplied: but they are already decisive.

(6) But on this important, and as it respects the main question, determining point, the facts just presented may be objected to by some, as the only means left of defending a system, which they are as tenacious to maintain as they were eager to embrace. Though I can hardly suppose many will suspect an author of wilfully falsifying printed documents, still, I should have been glad to have presented at full length the tables thus examined; but as it would

have involved the printing of some hundred additional pages of figures to have done so, I was obliged to decline it. All I can therefore do, is to offer the documents from which the preceding sums are obtained to the examination of any who may be disposed to controvert their exactness; or, rather, to refer such to the original tables themselves. In the mean time, I can assure the reader, that I had an inducement to be as exact as possible, even beyond what the determination of the subject in dispute supplied. I had, as I believed, previously discovered, from the examination of other records, a law of nature, far more mysterious, and not less certain than the one just examined, and this I found fully confirmed by these very registers, as well as by all others to which I have hitherto had access. A law, were the principle of population I am opposing true, which necessarily presents Providence in the perpetual act of malignantly frustrating the very attempts which Nature makes to mitigate the evils which that principle inflicts: but, if the system developed in this work is that of truth, which exhibits the same gracious power as the unwearied instrument, not only of promoting the general prosperity of mankind, but of instantly repairing those incidental calamities to which they are exposed. More of this, however, hereafter, when I shall, like Muret, "betake myself to a miracle1;" that is, venture to attribute the undoubted operations of Nature to a principle of intelligent and active benevolence.

(7) But to return. The reader may, perhaps, after all, refuse his confidence to my deductions from tables thus mutually appealed to by us both; but not easily to be met with, at least in this country, especially after having seen the incorrect use that has been made of

'Malthus, Essay on Population, p. 272.

them though I cannot but think that mine are of a very different character from those founded upon mere vague and general references, still, on a subject of so much moment, every reasonable objection ought to be considered; and this shall be entirely removed by transferring the proof to the registers of those other countries, which are generally referred to throughout the entire argument; the statistics of which are published here, and are sufficiently familiar to all who attend to inquiries like the present.

(8) Respecting Sweden (including Finland), Wargentin has given the marriages of fifteen successive years, from 1749 to 1763 inclusive'; dividing this period, after the manner of Susmilch, into three sections of five years each, these are the results.

TABLE XXXIV.

Shewing the EFFECT OF MORTALITIES ON THE NUMBER OF MARRIAGES IN SWEDEN.

This table also is conclusive as it regards Sweden. In the mortal years, the deaths exceeded those in the healthful ones by above 22 in the 100; but the marriages in the latter exceeded those in the former by

1 Wargentin, Kongl. Vetens. Academ. Handl. 1766, Q. In Sir Joseph Banks's Library.

This sum is the average of the last

five years, the year after 1763 not being given, the table terminating with that year.

6010; and even if the years after the mortal ones be added, still their mean number falls short by 2529.

(9) The census of France, as given by the Bureau of Longitude, for the years 1817 to 1824 inclusive, presents a less remarkable variation in the number of annual deaths than any of the documents hitherto referred to. Still it establishes the fact for which I contend; though, as it ought to do, admitting the deduction I have made to be a law of Nature, by a difference less striking. This term must of course be divided into four years only, as, indeed, some few of Susmilch's are. The facts in point are then as follow1:

TABLE XXXV.

SHEWING THE EFFECT OF MORTALITIES ON THE NUMBER OF MARRIAGES IN FRANCE.

France, therefore, may be confidently appealed to in contradiction of the notion that deaths make room for marriages. The marriages in the most fatal years are fewer than in those the least fatal; and, even if the year following be added to the former, the mean

1 Since writing this, I have seen the annual report for 1825, and this gives 798,012 for the deaths of that year, the largest number that has yet occurred;

but the marriages are only 243,674fewer by 18,346, than those which took place in the least mortal year of the rine now given. See the Annuaries.