valuable.[381]

Now, even a cursory examination of this continuous curve will yield the
following results:--

1. The discharges occur most frequently on the second lunar day.

2. The days of the next most frequent discharges are the 22d; the 13th;
the 7th, 20th, and 26th; the 11th and 16th; so that, if we regard only the
first six of these, we find that the discharges occur most frequently on
the 2d, 7th, 13th, 20th, 22d, and 26th lunar days--i.e., the discharges
occur most frequently on days separated, on the average, by four-day
intervals; but actually the period between the 20th and 22d days is that
characterized by the most frequent discharges.

3. The days of minimum of discharge are the 1st, 5th, 15th, 18th, and
21st.

4. The curve is characterized by a continual see-sawing; so that every
notable maximum is immediately followed by a notable minimum. Thus, the
curve is of an entirely different character from that representing the
monthly rhythm of the pulse,[382] and this is only what one might have
expected; for, whereas the _mean_ pulsations vary only very slightly from
day to day,--thus giving rise to a gradually rising or sinking curve,--a
discharge from the sexual system relieves the tension by exhausting the
stored-up secretion, and is necessarily followed by some days of rest and
inactivity. In the very nature of the case, therefore, a curve of this
kind could not possibly be otherwise than most irregular if the discharges
tended to occur most frequently upon definite days of the month; and thus
the very irregularity of the curve affords us proof that there is a
regular male periodicity, such that on certain days of the month there is
greater probability of a spontaneous discharge than on any other days.

5. Gratifying, however, though this irregularity of the curve may be, yet
it entails a corresponding disadvantage, for we are precluded thereby from
readily perceiving the characteristics of the monthly rhythm as a whole. I
thought that perhaps this aspect of the rhythm might be rendered plainer
if I calculated the data into two-day averages; and the result, as shown
in Chart 10, is extremely satisfactory. Here we can at once perceive the
wonderful and almost geometric symmetry of the monthly rhythm; indeed, if
the third maximum were one unit higher, if the first minimum were one unit
lower, and if the lines joining the second minimum and third maximum, and
the fourth maximum and fourth minimum, were s