lled by pressure, rest and an expectant
attitude are to be preferred.

A word must be added as to the objections to distant proximal ligature
for primary or recurrent haemorrhage. In some situations this may be
unavoidable, and it is sometimes successful, but none the less it is
opposed to all rules of good surgery and a most uncertain procedure. It
leaves the patient exposed to all the risks attendant on the employment
of simple pressure. In one case which I saw, the third part of the
subclavian artery had been ligatured for axillary bleeding; secondary
haemorrhage, as might have been expected, occurred, and that as late as
five weeks after the operation. In another case ligature of the femoral
artery for popliteal haemorrhage was followed by the development of a
traumatic aneurism in the ham.

_Secondary._--In secondary haemorrhage the treatment to be adopted
depends upon the nature of the case. When the wound is aseptic, and
bleeding the result of the separation of sloughs, local ligature is the
proper treatment, and this was often successfully adopted, especially in
the case of such arteries as the tibials. In septic cases, on the other
hand, it is usually far better if possible to amputate, unless the
general state of the patient and the local conditions are especially
favourable.

When neither amputation nor direct local ligature is practicable,
proximal ligature may be of use. Sometimes this may be obligatory in
consequence of the difficulties attendant on direct local treatment. I
saw a few cases successfully treated in this manner: in one the common
carotid was tied (Mr. Jameson) for haemorrhage from an arterial haematoma
in connection with the internal maxillary artery. Although ligature of
the external carotid would perhaps have been preferable, the result was
excellent. When even this expedient is impracticable, local pressure is
the only resort.

Lastly, as to the treatment of secondary interstitial blood effusions, I
believe the best initial treatment is the expectant. If interference is
needed, it is much more likely to be satisfactory the more chronic the
condition has become, since the source of the bleeding may be impossible
to discover. I never saw a patient's life endangered by the amount of
such haemorrhage, but if this should seem to be likely, local treatment
is of course unavoidable. In several cases quoted below, incision and
evacuation were followed by excellent results; in any such operati