120 JOHN BERRY HAYCRAFT ON 
May, = 3290 births (legitimate). 
June, = 3289 x 3 
July, = 3107 5 » 
August, = 2915 53 . 
September, = 2979 f 3 
October, = 3072 i % 
‘November, = 3001 Me ie 
December, = 3050 a a 
Average, = 3127. 
If the births are diagramatically represented, the ordinates being the number 
of births, and the abscisse the time of the year in months, a curve is produced. 
Now every year this curve may be seen recurring with striking regularity, 
there being always some months in the year which are fertile in births, and 
others which are comparatively barren. There must be a cause, or causes, 
working with equal regularity to produce it. | 
It is our task to investigate these, and, if possible, to ascertain their value. 
On thinking over the factors which might exert an influence upon the 
number of births, it is natural that the number of “marriages” (which likewise 
varies in the different months) should be first thought of. On reference (see Chart 
I.) the marriage curve is seen to be of quite a different character from that of 
the birth curve. It may here be remarked that all the charts are prepared from 
the averaged returns in the same way: the abscisse are taken as months. and 
ordinates as numbers of births, &c. The “birth curve” rises to a maximum 
and then falls once in every year. The “marriage curve” exhibits two striking 
and sharp maxima recurring every year. - Obviously at any rate the marriage 
curve is not the only factor. It is, as we shall afterwards see, by no means the 
main one. 
On seeking further to elucidate the problem, it becomes apparent that in all 
probability it is not the births themselves which could have this seasonal or 
other influence brought to bear upon them, but rather the conceptions from 
which the births result. If the birth curve be shifted back nie months (so 
that October numbers are in January), a corresponding curve—in character— 
is produced. This is now not the “birth curve,” but the “conception curve.” 
And now at the onset we can dismiss the birth curve from the question. 
It is the “conception curve,” which we shall afterwards prove is subject to 
seasonal and other influences ; whose effects, however, only become apparent 
by the increased number of births. | 
“ The Conception Curve” (Chart I.) 
From January to February the curve falls, there being not so many concep- 
