Control of primary haemorrhage from a wounded vessel by the impaction of a foreign body was of much less frequent occurrence than appears to have been the case with the older bullets. I saw a case in which, on removal of a fragment of shell (Mr. S. W. F. Richardson), very free haemorrhage occurred from a wound of one of the circumflex arteries of the thigh, but not a single one in which a similar result followed the extraction of a bullet of small calibre. The comparative infrequency of retention of modern bullets is probably one of the main elements in this relation. A very curious instance of provisional plugging of a wound in the upper part of the brachial artery by an inserted loop of the musculo-spiral nerve was related to me by Mr. Clinton Dent. This instance must, I think, be regarded as an accident definitely dependent on the size and outline of the bullet and on the nature of the force transmitted by it to neighbouring structures. While, however, deaths from external primary haemorrhage were rare, a considerable number resulted from primary internal haemorrhage. In some of these, injury to the largest trunks in the thorax or abdomen led to an immediately fatal issue; in others wounds of the large visceral arteries, as of the lungs, liver, or mesentery, were scarcely less rapid in their results. In such cases the potential space offered by the peritoneal or pleural cavities favours the ready escape of blood from the wounded vessel, while the tendency of the blood effused into serous cavities to rapid coagulation is notably slight. Beyond this the comparative deficiency in direct support afforded by surrounding structures to vessels running in the large body cavities is also an important element in their behaviour when wounded. These remarks receive support from the observation that few, if any, patients survived an injury to the external iliac vessels within the abdomen, while the remarkable instances of escape from fatal haemorrhage from large vessels recorded below (cases 1-19) indicate that the mere size of a wounded vessel is not to be regarded as the sole factor in prognosis.