who are largely free from the industrial and social influences which evidently, to a great extent, control the conception-rate.  It seems probable that the seasonal influence would here be specially well shown.  The only attempt I have made in this direction is to examine a well-filled birthday-book.  The entries show a very high and equally maintained maximum of conceptions throughout April, May and June, followed by a marked minimum during the next three months, and an autumn rise very strongly marked, in November.  There is no December rise.  As will be seen, there is here a fairly exact resemblance to the yearly ecbolic curve of people of the same class.  The inquiry needs, however, to be extended to a very much larger number of cases.

Mr. John Douglass Brown, of Philadelphia, has kindly prepared and sent me, since the above was written, a series of curves showing the, annual periodicity of births among the educated classes in the State of Pennsylvania, using the statistics as to 4,066 births contained in the Biographical Catalogue of Matriculates of the College of the University of Pennsylvania.  Mr. Brown prepared four curves:  the first, covering the earliest period, 1757-1859; the second, the period 1860-1876; the third, 1877-1893; while the fourth presented the summated results for the whole period. (The dates named are those of the entry to classes, and not of actual occurrence of birth.) A very definite and well-marked curve is shown, and the average number of births (not conceptions) per day, for the whole period, is as follows:—­

        Jan.  Feb.  Mar.  Apr.  May June July Aug.  Sept.  Oct.  Nov.  Dec.
        10.5 11.4 11 8.3 10.2 10.5 11.5 12.6 12.3 11.6 12 11.7

There is thus a well-marked minimum of conceptions (a depression appearing here in each of the three periods, separately) about the month of July. (In the second period, however, which contains the smallest number of births, the minimum occurs in September.) From that low minimum there is steady and unbroken rise up to the chief maximum in November. (In the first period, however, the maximum is delayed till January, and in the second period it is somewhat diffused.) There is a tendency to a minor maximum in February, specially well marked in the third and most important period, and in the first period delayed until March.

A very curious and perhaps not accidental coincidence might be briefly pointed out before we leave this part of the subject.  It is found[157] by taking 3000 cases of children dying under one year that, among the general population, children born in February and September (and therefore conceived in May and December) appear to possess the greatest vitality, and those born in June, and, therefore, conceived in September, the least vitality.[158] As we have seen, May and December are precisely the periods when conceptions in Europe generally are at a maximum, and September is precisely the period when they are at a minimum, so that, if this coincidence is not accidental, the strongest children are conceived when there is the strongest tendency to procreate, and the feeblest children when that tendency is feeblest.