Studies in the Psychology of Sex (Vol. 1-6). Havelock Ellis
Die periodischen Geistesstörungen, 1901); it is much commoner in women than in men. Periodicity has been observed in stammering (a six-weekly period in one case), and notably in hemicrania or migraine, by Harry Campbell, Osler, etc. (The periodicity of a case of hemicrania has been studied in detail by D. Fraser Harris, Edinburgh Medical Journal, July, 1902.) But the cycle in these cases is not always, or even usually, of a menstrual type.
It is now possible to turn to an investigation which, although of very limited extent, serves to place the question of a male menstrual cycle for the first time on a sound basis. If there is such a cycle analogous to menstruation in women, it must be a recurring period of nervous erethism, and it must be demonstrably accompanied by greater sexual activity. In the American Journal of Psychology for 1888, Mr. Julius Nelson, afterward Professor of Biology at the Rutgers College of Agriculture, New Brunswick, published a study of dreams in which he recorded the results of detailed observations of his dreams, and also of seminal emissions during sleep (by him termed "gonekbole" or "ecbole"), during a period of something over two years. Mr. Nelson found that both dreams and ecboles fell into a physiological cycle of 28 days. The climax of maximum dreaming (as determined by the number of words in the dream record) and the climax of maximum ecbole fell at the same point of the cycle, the ecbolic climax being more distinctly marked than the dream climax.
The question of cyclic physiological changes is considerably complicated by our uncertainty regarding the precise length of the cycle we may expect to find. Nelson finds a 28-day cycle satisfactory. Perry-Coste, as we shall see, accepts a strictly lunar cycle of 29½ days. Fliess has argued that in both women and men, many physiological facts fall into a cycle of 23 days, which he calls male, the 28-day cycle being female. (W. Fliess, Die Beziehungen zwischen Nase und weiblichen Geschlechts-Organen, 1897, pp. 113 et seq.) Although Fliess brings forward a number of minutely-observed cases, I cannot say that I am yet convinced of the reality of this 23-day cycle. It is somewhat curious, however, that at the same time as Fliess, though in apparent independence, and from a different point of view, another worker also suggested that there is a 23-day physiological cycle (John Beard, The Span of Gestation and the Cause of Birth, Jena, 1897). Beard approaches the question from the embryological standpoint, and argues that there is what he terms an "ovulation unit" of about 23½ days, in the interval from the end of one menstruation to the beginning of the next. Two "ovulation units" make up one "critical unit," and the length of pregnancy, according to Beard, is always a multiple of the "critical unit;" in man, the gestation period amounts to six critical units. These attempts to prove a new physiological cycle deserve careful study and further investigation. The possibility of such a cycle should be borne in mind, but at present we are scarcely entitled to accept it.
So far as I am aware, Professor Nelson's very interesting series of observations, which, for the first time, placed the question of a menstrual rhythm in men on a sound and workable basis, have not directly led to any further observations. I am, however, in possession of a much more extended series of ecbolic observations completed before Nelson's paper was published, although the results have only been calculated at a comparatively-recent date. I now propose to present a summary of these observations, and consider how far they confirm Nelson's conclusions. These observations cover no less a period than twelve years, between the ages of 17 and 29, the subject, W. K., being a student, and afterward schoolmaster, leading, on the whole, a chaste life. The records were faithfully made throughout the whole of this long period. Here, if anywhere, should be material for the construction of a menstrual rhythm on an ecbolic basis. While the results are in many respects instructive, it can scarcely, perhaps, be said that they absolutely demonstrate a monthly cycle. When summated in a somewhat similar manner to that adopted by Nelson in his ecbolic observations, it is not difficult to regard the maximum, which is reached on the 19th to 21st days of the summated physiological month, as a real menstrual ecbolic climax, for no other three consecutive days at all approach these in number of ecboles, while there is a marked depression occurring four days earlier, on the 16th day of the month. If, however, we split up the curve by dividing the period of twelve years into two nearly equal periods, the earlier of about seven years and the latter of about four years, and summate these separately, the two curves do not present any parallel as regards the menstrual cycle. It scarcely seems to me, therefore, that these curves present any convincing evidence in this case of a monthly ecbolic cycle (and, therefore, I refrain from reproducing them), although they seem to suggest such a cycle. Nor is there any reason to suppose that by adopting a different cycle of thirty days, or of twenty-three days, any more conclusive results would be obtained.
It seems, however, when we look at these curves more closely, that they are not wholly without significance. If I am justified in concluding that they scarcely demonstrate a monthly cycle, it may certainly be added that they show a rudimentary tendency for the ecboles to fall into a fortnightly rhythm, and a very marked and unmistakable tendency to a weekly rhythm. The fortnightly rhythm is shown in the curve for the earlier period, but is somewhat disguised in the curve for the total period, because the first climax is spread over two days, the 7th and 8th of the month. If we readjust the curve for the total period by presenting the days in pairs, the fortnightly tendency is more clearly brought out (Chart I).
A more pronounced tendency still is traceable to a weekly rhythm. This is, indeed, the most unquestionable fact brought out by these curves. All the maxima occur on Saturday or Sunday, with the minima on Tuesday, Wednesday, Thursday, or Friday. This very pronounced weekly rhythm will serve to swamp more or less completely any monthly rhythm on a 28-day basis. Although here probably seen in an exaggerated form, it is almost certainly a characteristic of the ecbolic curve generally.[123] I have been told by several young men and women, especially those who work hard during the week, that Saturday, and especially Sunday afternoon, are periods when the thoughts spontaneously go in an erotic direction, and at this time there is a special tendency to masturbation or to spontaneous sexual excitement. It is on Friday, Saturday, Sunday, and Monday, according to Guerry's tables,[124] that the fewest suicides are committed, Tuesday, Wednesday, and Thursday, with, however, a partial fall on Wednesday, those on which most suicides are committed, so that there would appear to be an antagonism between sexual activity and the desire to throw off life. It also appears (in the reports of the Bavarian factory inspectors) that accidents in factories have a tendency to occur chiefly at the beginning of the week, and toward the end rather than in the middle.[125] Even growth, as Fleischmann has shown in the case of children, tends to fall into weekly cycles. It is evident that the nervous system is profoundly affected by the social influences resulting from the weekly cycle.
The analysis of this series of ecbolic curves may thus be said to recall the suggestion of Laycock, that the menstrual cycle is really made up of four weekly cycles, the periodic unit, according to Laycock, being three and one-half days. I think it would, however, be more correct to say that the menstrual cycle, perhaps originally formed with reference to the influence of the moon on the sexual and social habits of men and other animals, tends to break up by a process of segmentation into fortnightly and weekly cycles. If we are justified in assuming that there is a male menstrual cycle, we must conclude that in such a case as that just analyzed, the weekly rhythm has become so marked as almost entirely to obliterate the larger monthly rhythm.
However constituted, there seems little doubt that a physiological weekly cycle really exists. This was, indeed, very clearly indicated many years ago by the observations of Edward Smith, who showed that there are weekly rhythms in pulse, respiration, temperature, carbonic acid evolution, urea, and body-weight, Sunday being the great day of repair and increase of weight.[126]
In an appendix to this volume I am able to present the results of another long series of observations of nocturnal ecbolic manifestations carried out by Mr. Perry-Coste, who has elaborately calculated the results, and has convinced himself that on the basis of a strictly lunar month, thus abolishing the disturbing influence of the weekly rhythm, which in his case also