ing the wall of the artery from the
outside; for example in sloughing sore-throat, the separation of a
slough may implicate the wall of an artery and be followed by serious
and it may be fatal haemorrhage. The mechanical pressure of a fragment of
bone or of a rubber drainage tube upon the vessel may aid the septic
process in causing erosion of the artery. In pre-Listerian days, the
silk ligature around the artery likewise favoured the changes that lead
to secondary haemorrhage, and the interesting observation was often made,
that when the collateral circulation was well established, the leakage
occurred on the _distal_ side of the ligature. While it may happen that
the initial haemorrhage is rapidly fatal, as for example when the
external carotid or one of its branches suddenly gives way, it is quite
common to have one, two or more _warning haemorrhages_ before the leakage
on a large scale, which is rapidly fatal.

The _appearances of the wound_ in cases complicated by secondary
haemorrhage are only characteristic in so far that while obviously
infected, there is an absence of all reaction; instead of frankly
suppurating, there is little or no discharge and the surrounding
cellular tissue and the limb beyond are oedematous and pit on pressure.

The _general symptoms_ of septic poisoning in cases of secondary
haemorrhage vary widely in severity: they may be so slight that the
general health is scarcely affected and the convalescence from an
operation, for example, may be apparently normal except that the wound
does not heal satisfactorily. For example, a patient may be recovering
from an operation such as the removal of an epithelioma of the mouth,
pharynx or larynx and the associated lymph glands in the neck, and be
able to be up and going about his room, when, suddenly, without warning
and without obvious cause, a rush of blood occurs from the mouth or the
incompletely healed wound in the neck, causing death within a few
minutes.

On the other hand, the toxaemia may be of a profound type associated with
marked pallor and progressive failure of strength, which, of itself,
even when the danger from haemorrhage has been overcome, may have a fatal
termination. The _prognosis_ therefore in cases of secondary haemorrhage
can never be other than uncertain and unfavourable; the danger from loss
of blood _per se_ is less when the artery concerned is amenable to
control by surgical measures.

_Treatment._--The treatment of s