in autumn, the spring rise being in February, March, and April; the autumn rise in October or November.  There are, however, certain fallacies in dealing with institutions like Normal schools, where the conditions are not perfectly regular throughout the year, owing to vacations, etc.  It is, therefore, instructive to find that under the monotonous conditions of prison-life precisely the same spring and autumn rises are found.  Binet takes the consumption of bread in the women’s prison at Clermont, where some four hundred prisoners, chiefly between the ages of thirty and forty, are confined, and he presents two curves for the years 1895 and 1896.  The curves for these two years show certain marked disagreements with each other, but both unite in presenting a distinct rise in April, preceded and followed by a fall, and both present a still more marked autumn rise, in one case in September and November, in the other case in October.[169]

Some years ago, Sir J. Crichton-Browne stated that a manifestation of the sexual stimulus of spring is to be found in the large number of novels read during the month of March ("Address in Psychology” at the annual meeting of the British Medical Association, Leeds, 1889; Lancet, August 14, 1889).  The statement was supported by figures furnished by lending libraries, and has since been widely copied.  It would certainly be interesting if we could so simply show the connection between love and season, by proving that when the birds began to sing their notes, the young person’s fancy naturally turns to brood over the pictures of mating in novels.  I accordingly applied to Mr. Capel Shaw, Chief Librarian of the Birmingham Free Libraries (specially referred to by Sir J. Crichton-Browne), who furnished me with the Reports for 1896 and 1897-98 (this latter report is carried on to the end of March, 1898).

The readers who use the Birmingham Free Lending Libraries are about 30,000 in number; they consist very largely of young people between the ages of 14 and 25; somewhat less than half are women.  Certainly we seem to have here a good field for the determination of this question.  The monthly figures for each of the ten Birmingham libraries are given separately, and it is clear at a glance that without exception the maximum number of readers of prose-fiction at all the libraries during 1897-98 is found in the month of March. (I have chiefly taken into consideration the figures for 1897-98; the figures for 1896 are somewhat abnormal and irregular, probably owing to a decrease in readers, attributed to increased activity in trade, and partly to a disturbing influence caused by the opening of a large new library in the course of the year, suddenly increasing the number of readers, and drafting off borrowers from some of the other libraries.) Not only so, but there is a second, or autumnal climax, almost equaling the spring climax, and occuring with equal certainty, appearing