principally confined to this country, which alone affords the necessary facts for a sufficient number of years past, by which to determine it.

(3) The official censuses of England and Wales present us with the number of the marriages annually contracted only, and not of those previously subsisting; and as it is obvious, that these are in no necessary or known proportion to each other, and as the number of the latter can alone decide the question, it has to be determined by calculation, which, if founded upon different data, mutually confirming and correcting each other, the results may be regarded as being equally satisfactory for the purpose of comparison, on which the argument solely rests, as though the computation were critically exact; especially since, by this uniform method, it is impossible to be betrayed into those false accommodations of facts, or partial selection of them, which, in enquiries of this nature, are often made to serve an argument at the expense of truth.

(4) In making these computations, there is another most important circumstance to be attended to, which, as far as I have observed, has escaped the attention of all writers who have calculated the prolificness of marriages, upon the facts which are given by registers; I mean, the far higher degree of fecundity (as shewn by the conceptions) in the first year of marriage, than in any of the succeeding ones to its termination on any general calculation, certainly more than a fourfold proportion, when so compared ; and calculated on the fruitful years of marriage, at least a double one, on the average. The number of the annual marriages, being exceedingly various, it is quite clear that the number of the conceptions of that year will vary proportionably; and it is equally plain that the differences so occasioned do not at all deter

mine the question, as to the comparative prolificness of the entire number of existing marriages. It has been from the want of attending to this essential consideration, that the error, which I am now refuting, as well as many similar ones, has been fallen into. In again referring to a register of the peerage which I have formed, with a view to the determination of this and many other interesting questions, I find that twothirds (very exactly) of the marriages produce a birth each before the termination of the year following that in which they take place; and that each of the ensuing fecund years are, on the whole, only about half as prolific; but if calculated on the entire duration of marriages, the remaining years are, on an average, each less than a quarter as productive. Now, I think none would carry the argument so far as to assert, that the price of grain influences the fertility of the first year of marriage. I shall, therefore, in the ensuing calculations, deduct, or rather leave out, the marriages of the current year from the whole number; and subtract, from the conceptions of that year, a number equal to two-thirds of such marriages: the remainder will then express, accurately enough, the regular prolificness of the rest of the standing marriages, freed from the fluctuations which arise from variations in the number of the annual weddings. The variations in the annual prolificness of these standing marriages, collated with contemporaneous differences in the price of food, or wheat, will thus determine the question.

(5) The number of the annual marriages and deaths, during the period about to be examined, being known, to find the sum of the standing marriages at the commencement of it, and the proportion of existing marriages dissolved by the annual deaths, is the question. The number of standing marriages assumed

as those existing in 1780, was the result of a series of calculations, founded on the population, and the proportion they bear to it in other countries, and in some districts of this, where that proportion has been actually ascertained: taking into consideration, also, the various circumstances which might affect it, namely, the average age of marriage, the degree of its prevalence, and the expectation of life. From the amount of the standing marriages thus obtained, I deducted 1 in 3.11 of the annual deaths, as being the proportion of them which would be yearly dissolved by death. There was no difficulty as to the annual accessions to the existing marriages, these having been published.

(6) Without presuming that the results thus obtained are exact, I conceive they are sufficiently so to decide the question; which, as before observed, is one founded on comparison only. I do not, however, think that they are very remote from the truth the various proportions they exhibit support each other; and, indeed, as a final proof that they are tolerably accurate, it will be found, that if the average annual fecundity of each marriage, as expressed in the ensuing table, be multiplied by its average duration, and two-thirds of a birth, or .666 be added to it, the sum will be very nearly that of the average prolificness of marriages, as given in the public documents for the same period. I assume, then, as the result of calculations formed, as explained above, that the existing marriages of 1780 amounted to 1,289,550, whereof 61,760 were weddings of that year, the conceptions belonging to which were 41,140, leaving 182,983, as resulting from 1,227,790, the couples married previously to 1780; the annual prolificness of each, therefore, was in that year, 18 of a birth each. The ensuing table then proceeds thus :—