the accident occurs a few days after conception, bleeding may be its
only evidence, which will almost certainly be misinterpreted as an
irregularity of menstruation; and professional advice will not often
be thought necessary. Moreover, in other cases in which the true
situation is appreciated the patient does not feel sick enough to
seek medical assistance. If it were possible to include in the
statistics all these cases as well as those which are concealed
because intentionally provoked, the frequency with which pregnancy is
interrupted during the early months would be found somewhat greater
than is usually supposed.

If we omit the miscarriages which occur within the first few weeks of
pregnancy, and which consequently often escape detection, the
majority of cases fall within the second and third months. After the
fourth month has passed, the probability of such an accident, though
not excluded, is greatly diminished. Some statistics recently
published by Taussig make this clear. In a series of several hundred
cases of miscarriage, one hundred and fifty-seven instances occurred
in the second month, two hundred and twenty-two in the third month,
seventy-three in the fourth month, thirty-seven in the fifth month,
and five in the sixth month. This order of frequency might be
anticipated from the anatomical conditions which prevail during the
early months of pregnancy, since the attachment of the embryo to the
mother is at first relatively insecure, but gradually grows firmer,
and becomes as secure as it ever will be by about the fifth month.

It is noteworthy that miscarriage occurs much less commonly in the
first than in subsequent pregnancies. Indeed, a somewhat greater
liability to the accident with each succeeding pregnancy goes far
toward explaining the greater frequency of miscarriage among women