Chapter XI: Injuries to the Abdomen (1)

Perhaps no chapter of military surgery was looked forward to with more eager interest than that dealing with wounds of the abdomen. In none was greater expectation indulged in with regard to probable advance in active surgical treatment, and in none did greater disappointment lie in store for us.

Wounds of the solid viscera, it is true, proved to be of minor importance when produced by bullets of small calibre; but wounds of the intestinal tract, although they showed themselves capable of spontaneous recovery in a certain proportion of the cases observed, afforded but slight opportunity for surgical skill, and results generally deviated but slightly from those of past experience. Such success as was met with depended rather on the mechanical genesis and nature of the wounds than upon the efforts of the surgeon, and operative surgery scored but few successes.

It is true that to the Civil Surgeon accustomed to surroundings replete with every modern appliance and convenience, and the possibility of exercising the most stringent precautions against the introduction of sepsis from without, abdominal operations presented difficulties only faintly appreciated in advance; but this alone scarcely accounted for the want of success attending the active treatment of wounds of the intestine when occasion demanded. Failure was rather to be referred to the severity of the local injury to be dealt with, or to the operations being necessarily undertaken at too late a date. Many fatalities, again, were due to the association of other injuries, a large proportion of the wound tracks involving other organs or parts beyond the boundaries of the abdominal cavity.

The frequent association of wounds of the thoracic cavity with those of the abdomen afforded many of the most striking examples of immunity from serious consequences as a result of wound of the pleura. It must be conceded that in a large number of such injuries only the extreme limits of the pleural sac were encroached upon, yet in some the tracks passed through the lungs, although without serious consequences. Under the heading of injury to the large intestine a somewhat special form of pleural septicæmia will be referred to.

It may at once be stated that such favourable results as occurred in abdominal injuries were practically limited to wounds caused by bullets of small calibre, and that, although in the short chapter dealing with shell injuries a few recoveries from visceral wounds will be mentioned, I never met with a penetrating visceral injury from a Martini-Henry or large sporting bullet which did not prove fatal.

_Wounds of the abdominal wall._--It is somewhat paradoxical to say that these injuries possessed special interest from their comparative rarity of occurrence, since they were not of intrinsic importance. Their infrequency depended on the difficulty of striking the body in such a plane as to implicate the belly wall alone, and their interest in the diagnostic difficulty which they gave rise to.

In many cases the position of the openings and the strongly oval or gutter character possessed by them were sufficient proof of the superficial passage of the bullet; in others we had to bear in mind that the position of the patient when struck was rarely that of rest in the supine position, in which the surgical examination was made, and considerable difficulty arose. Some superficial tracks crossing the belly wall have already been referred to in the chapter on wounds in general and in that dealing with injuries to the chest, in which the above characters sufficed to indicate that penetration of the abdominal cavity had not occurred. In other instances a definite subcutaneous gutter could be traced, and often in these a well-marked cord in the abdominal wall corresponding to the track could be felt at a later date. Again, limitation to the abdominal wall was sometimes proved by the position of the retained bullet, or sometimes by the presence in the track of foreign bodies carried in with the projectile. See case 160.

Fig. 84 illustrates an example where the limitation to the abdominal wall was evident on inspection. Here the division of the thick muscles of the abdominal wall had led to the formation of a swelling exactly similar to that seen after the subcutaneous rupture of a muscle, and two soft fluctuating tumours bounded by contracted muscle existed in the substance of the oblique and rectus muscles.

The cases which presented the most serious diagnostic difficulty in this relation were those in which the wound was situated in the thicker muscular portions of the lower part of the abdominal and pelvic walls. Such a case is illustrated in the chapter on fractures (see fig. 55, p. 191). I saw one or two such instances, in which only the exploration necessary for treatment of the fracture decided the point. In many of the wounds affecting the lateral portion of the abdominal wall the question of penetration could never be definitely cleared up, as wounds of the colon sometimes gave rise to absolutely no symptoms.

In a certain proportion of the injuries the peritoneal cavity was no doubt perforated without the infliction of any further visceral injury, and in these also the doubt as to the occurrence of penetration was never solved.

(158) _Wound of belly wall._--Wounded at Modder River. _Entry_
(Mauser), 2 inches below the centre of the left iliac crest;
_exit_, 1-1/2 inch above and internal to the left anterior
superior iliac spine. The patient was on horseback at the time
of the injury and did not fall; he got down, however, and lay
on the field an hour, whence he was removed to hospital.
Probably the track pierced the ilium, and remained confined to
the abdominal wall. There were no signs of visceral injury.

(159) Cape Boy. Wounded at Modder River. _Entry_ (Lee-Metford),
immediately above and outside right anterior superior spine;
_exit_, 1-1/2 inch below and to right of umbilicus. A
well-marked swelling corresponded with division of the fibres
of the oblique muscles and of the rectus, and on palpation a
hollow corresponding with the track was felt. The abdominal
muscles were exceptionally well developed (fig. 84).

(160) Wounded at Magersfontein while lying prone. _Entry_,
irregular, oblique, and somewhat contused, over the eighth left
rib, in the anterior axillary line; _exit_, a slit wound
immediately above and to the left of the umbilicus. The bullet
struck a small circular metal looking-glass before entering,
hence the irregularity of the wound. The patient developed a
hæmothorax, but no abdominal signs; the former was probably
parietal in origin, secondary to the fractured rib, and the
whole wound non-penetrating as far as the abdominal cavity was
concerned.

(161) Wounded at Magersfontein. _Entry_ (Mauser), 1-1/2 inch
external to and 1/2 inch below the left posterior superior
iliac spine; _exit_, 1 inch internal horizontally to the left
anterior superior spine.

No signs of intra-peritoneal injury were noted, but free
suppuration occurred in left loin; the ilium was tunnelled.

The same patient was wounded by a Jeffrey bullet in the hand;
the third metacarpal was pulverised, although the bullet, which
was longitudinally flanged, was retained.

(162) Wounded outside Heilbron. _Entry_, below the eighth right
costal cartilage; _exit_, below the eighth cartilage of the
left side. The wound of entry was slightly oval; that of exit
continued out as a 'flame'-like groove for 2 inches. A week
later the wound track could be palpated as an evident hard
continuous cord.

_Penetration of the intestinal area without definite evidence of visceral injury._--This accident occurred with a sufficient degree of frequency to obtain the greatest importance, both from the point of view of diagnosis and prognosis, and as affecting the question of operative interference. Amongst the cases reported below a number occurred in which it was impossible to settle the question whether injury to the bowel had occurred or not, and I will here shortly give what explanation I can for the apparent escape of the intestine from serious injury.

We may first recall the general question of the escape of structures lying to one or other side of the track of the bullet. I believe that there can be no doubt as to the accuracy of the remarks already made as to the escape of such structures as the nerves by means of displacement, and that the occurrence of such escapes is manifestly dependent on the degree of fixity of the nerve or the special segment of it implicated. The general tendency of the tissues around the tracks to escape extensive destruction from actual contusion has also been referred to, and is, I think, indisputable.

If these observations be accepted, I think there can be no difficulty in allowing that the small intestine is exceptionally well arranged to escape injury. First of all, it is very moveable; secondly, it is so arranged that in certain directions a bullet may pass almost parallel to the long axis of the coils; thirdly, it is elastic, capable of compression, and light, and hence offers but a small degree of resistance to the passage of the bullet across the abdominal cavity.

Certain evidence both clinical and pathological supports the contention that the small intestine may escape injury from the passing bullet.

First of all, the fact may be broadly stated that injuries to the small intestine were fatal in the great majority of certainly diagnosed cases, while, on the other hand, many tracks crossed the area occupied by the small intestine without serious symptoms of any kind resulting. Secondly, experience showed that when the bullet crossed the line of the fixed portions of the large intestine the gut rarely escaped, and that, although a considerable proportion of these cases recovered spontaneously, in a large number of them immediate symptoms, or secondary complications, clearly substantiated the nature of the original injury. As far as my experience went, however, I never saw any instance in which an undoubted injury of the small intestine was followed by the development of a local peritoneal suppuration and recovery, a sequence by no means uncommon in the case of wounds of the large intestine. Although, therefore, I am not prepared to deny the possibility of spontaneous recovery from an injury to the small intestine, under certain conditions which will be stated later, I believe that in the immense majority of cases in which a bullet crossed the small intestine area without the supervention of serious symptoms, the small intestine escaped perforating injury.

Beyond the clinical evidence offered above, certain pathological observations support the view that the intestine escapes perforation by displacement. Most of my knowledge on this subject was derived from the limited number of abdominal sections I performed on cases of injury to the small intestine, and may be summed up as follows.

The small intestine may present evidence of lateral contusion in the shape of elongated ecchymoses, either parallel, oblique, or transverse to its long axis. These ecchymoses resemble in extent and outline those which ordinarily surround a wound of the intestinal wall produced by a bullet (see fig. 87, p. 418).

The wall of the small intestine may be wounded to an extent short of perforation, either the peritoneal coat alone being split, or the wound implicating the muscular coat and producing an appearance similar to that seen when the intestine is dragged upon during an operation, but without so much gaping of the edges (see fig. 85, p. 416).

I met with these conditions in association with co-existing complete perforations of the small intestine, and in one case of intra-peritoneal hæmorrhage in which no complete perforation was discoverable (No. 169, p. 432).

The implication and perforation of the small intestine are to some extent influenced by the direction of the wound. A striking case is included below, No. 201, in which a bullet passed from the loin to the iliac fossa on each side of the body, approximately parallel to the course of the inner margin of the colon, and I also saw some other wounds in this direction in which no evidence of injury to the small intestine was detected, and which got well. Again wounds from flank to flank were, as a rule, very fatal; but I saw more than one instance where these wounds were situated immediately below the crest of the ilium, in which the intestine escaped injury (see case 171). A very striking observation was made by Mr. Cheatle in such a wound. The patient died as a result of a double perforation of both cæcum and sigmoid flexure; none the less the bullet had crossed the small intestine area without inflicting any injury.

The sum of my experience, in fact, was to encourage the belief that, unless the intestine was struck in such a direction as to render lateral displacement an impossibility, the gut often escaped perforation.

As a rule, the wounds of the abdomen which from their position proved the most dangerous to the intestine were--

1. Wounds passing from one flank to the other were very dangerous, as crossing complicated coils of the small intestine, and two fixed portions of the colon. This danger was most marked when the wounds were situated between the eighth rib in the mid axillary line and the crest of the ilium; above this level the liver, or possibly liver and stomach, were sometimes alone implicated, and the cases did well. Again, when the wounds crossed the false pelvis the patients sometimes escaped all injury to viscera.

2. Antero-posterior wounds in the small intestine area were very fatal if the course was direct; in such the small intestine seldom escaped injury.

3. Wounds with a certain degree of obliquity from anterior wall to flank, or from flank to loin, were on the other hand comparatively favourable, as the small intestine often escaped, and if any gut was wounded, it was often the colon.

4. Vertical wounds implicating the chest and abdomen, or the abdomen and pelvis, were on the whole not very unfavourable. For instance, when the bullet entered by the buttock and emerged below the umbilicus, a number of patients escaped fatal injury; this depended on the comparatively good prognosis in wounds of the rectum and bladder. A good many patients in whom the bullet entered by the upper part of the loin, and escaped 1-1/2 inch within the anterior superior spine of the ilium, also did well. The same holds good when the wounds either entered or emerged under the anterior costal margin of the thorax, either prior to or after traversing the thorax.

Wounds passing directly backward from the iliac regions were in my experience very unfavourable; but I believe mainly as a result of hæmorrhage from the iliac arteries.

_The occurrence of wounds of the abdomen of an 'explosive' character._--The vast majority of the abdominal wounds observed in the Stationary or Base hospitals were of the type dimensions. A certain number of the abdominal injuries which proved fatal on the field or shortly afterwards were described as explosive in character, and were referred by the observers to the employment of expanding bullets.

A few words on this subject seem necessary, because it seems doubtful whether such injuries could be produced by any of the forms of expanding bullet of small calibre in use, unless the track crossed one of the bones in the abdominal or pelvic wall. That this was sometimes the case there is no doubt: thus I saw two cases in which the splenic flexure of the colon was wounded, in which the external opening was large, and a comminuted fracture of the ribs of the left side existed. One can well believe that bullets passing through the pelvic bones might 'set up' to a considerable extent, and although I never happened to see such a case, an explanation of some of the wounds described by others might be found in this occurrence.

In instances in which the soft parts alone were perforated, I am disinclined to believe that bullets of small calibre, either regulation or soft-nosed, were responsible for the injuries. I had the opportunity of examining two Mauser bullets of the Jeffreys variety which crossed the abdomen and caused death. In the first (figured on page 94, fig. 40) very little alteration beyond slight shortening had occurred. In the second the deformity was almost the same, except that the side of the bullet was indented, probably from impact with some object prior to its entry into the body. In each case the bullet was of course travelling at a low rate of velocity; hence no very strong inference can be drawn from either. In the case of the second specimen, which was removed by Mr. Cheatle, a remarkable observation was made, which tends to throw some light on one possible mode of production of large exit apertures. This bullet crossed the cæcum, making two small type openings; but later, when it crossed the sigmoid flexure, it tore two large irregular openings in the gut. This might be explained on the ground that the velocity was so small as only just to allow of perforation, which therefore took the nature of a tear. I am inclined to suggest, as a more likely explanation, that the spent bullet turned head over heels in its course across the abdomen, and made lateral or irregular impact with the last piece of bowel it touched. A slightly greater degree of force would have allowed a similar large and irregular opening to be made in the abdominal wall also.

In this relation the question will naturally be raised as to how far the explosive appearances may have been due to high velocity alone on the part of the bullet. I am disinclined from my general experience to believe that explosive injuries of the soft parts were to be thus explained. On the other hand, I believe that the possession of a low degree of velocity very greatly increased the danger in abdominal wounds. I believe that the bowel was, under these circumstances, less likely to escape by displacement, and was more widely torn when wounded; again, that inexact impact led to increase of size in the external apertures, and the bullet was of course more often retained.

Mr. Watson Cheyne[19] published a very remarkable instance of one of the dangers of an injury from a spent bullet, in which, in spite of non-penetration of the abdominal cavity, the small intestine was ruptured in two places.

I believe the majority of the wounds designated as explosive were the result of the passage of large leaden bullets, either of the Martini-Henry or Express type. The small opportunity of observing such injuries in the hospitals of course depended on the fact that the majority were rapidly fatal.

_Nature of the anatomical lesion in wounds of the intestine._--The openings in the parietal peritoneum tended to assume the slit or star forms, probably on account of the elasticity of the membrane. A diagram of one of these forms is appended to fig. 89. In this instance the opening in the peritoneum was made from the abdominal aspect, prior to the escape of the bullet from the cavity, and on the impact of the tip, the long axis of the bullet was oblique to the surface of the abdominal wall.

In the intestinal wall the openings varied in character according to the mode of impact.

In some cases the gut was merely contused by lateral contact of the passing bullet. The result of this was evidenced later by the presence of localised oval patches of ecchymosis. These were identical in appearance with the patches shown surrounding the wounds in fig. 87.

More forcible lateral impact produced a split of the peritoneum, or of this together with the muscular coat. Such a lateral slit is shown in fig. 85, although the clearness of outline is somewhat impaired by the presence of a considerable amount of inflammatory lymph.

Fig. 86 exhibits a lateral injury of a more pronounced form. The bullet here struck the most prominent portion of the under surface of the bowel, and produced a circular perforation not very unlike one produced by rectangular impact, except in the lesser degree of eversion of the mucous membrane. Here again the appearance is somewhat altered by the presence of a considerable amount of lymph, but this is of less importance in this figure because the lymph is localised to the portion of the bowel in the immediate neighbourhood of the opening which had suffered contusion and erasion.

Fig. 87, A B, illustrates a symmetrical perforation of the small intestine; the aperture of entry (A) is roughly circular, and a ring of mucous membrane protrudes and partially closes the opening. The aperture of exit is a curved slit, again partially occluded by the mucous membrane. The same amount of difference between the two apertures did not always exist; in many cases both were circular, and apparently symmetrical. Beyond this I have seen three apertures in close proximity, two lying on the same aspect of the bowel, and the first of these was no doubt an opening due to lateral impact similar to that seen in fig. 86. In the recent condition little difference existed between the three apertures.

The localised ecchymosis surrounding the apertures is quite characteristic of this form of injury, and is a valuable aid to finding the openings during an operation.

Fig. 88 shows the interior of the same segment of bowel, as fig. 87. It shows the localised ecchymosis as seen from the inner surface, here rather more extensive from the fact that the blood spreads more readily in the submucous tissue.

It will be noted that the main feature of the form of injury is the regular outline and the small size of the wounds. Another feature not illustrated by the figures should also be mentioned. In the ruptures of intestine with which we are acquainted in civil practice the wound in the gut is almost without exception situated at the free border of the bowel, but in these injuries it was just as frequently at the mesenteric margin. The importance of this factor is considerable, since wounds near the mesenteric edge are much more likely to be accompanied by hæmorrhage, and thus the opportunity for diffusion of infection is considerably multiplied, to say nothing of the danger from loss of blood.

Beyond these more or less pure perforations, long slits or gutters were occasionally cut. I saw instances of these in the case of the ascending colon, and in the small curvature of the stomach. The comparative fixity of the portion of bowel struck is a matter of great importance in the production of this form of injury.

It may be well to add that, although the figures inserted are all taken from small-intestine wounds, the nature of the wounds of the peritoneum-clad part of the large intestine in no way differed from them, except in so far as fixity of the bowel exposed it to a more extensive wound when the bullet took a parallel course to its long axis.

A more important point in the injuries to the large intestine was the possibility of an extra-peritoneal wound. I saw several such lesions of the colon, every one of which ended fatally. I became still more fully convinced of the greater seriousness of extra- to intra-peritoneal rupture of this portion of the gut than I was when I expressed a similar opinion in a former paper.[20] It will be seen later that the results of intra- and extra-peritoneal wounds of the bladder fully confirm this view, as all extra-peritoneal injuries died, while many intra-peritoneal perforations recovered spontaneously.

_Wounds of the mesentery._--I had little experience of this injury; in fact, case 169, on which I operated, was my sole observation. It stands to reason, however, that injuries to the mesentery would be much more frequent proportionately to wounds of the gut than is the case in the ruptures seen in civil practice, since the whole area of the mesentery is equally open to injury. Viewing the extreme danger of hæmorrhage into the peritoneal cavity in these injuries, I should be inclined to expect that a considerable proportion of those deaths from abdominal wounds which took place on the field of battle were due to this source.

_Wounds of the omentum._--Here, again, I am unable to express any opinion, although the supposition that hæmorrhage from this source took place is natural.

Prolapse of omentum was comparatively rare, except in cases with large wounds; it was apparently seen with some frequency among patients who died rapidly on the field of battle. I only saw it twice, and on each occasion in shell wounds. The wounds from small-calibre bullets were as a rule too small to allow of external prolapse.

Fig. 89, however, illustrates a very interesting observation. A patient in the German Ambulance in Heilbron, under Dr. Flockemann, died as a result of suppuration and hæmorrhage secondary to an injury to the colon. At the autopsy a portion of the omentum was found adherent in the wound of exit, but it had not reached the external surface. The chief interest of the observation lies in the light it throws on the mechanism of these injuries. It is impossible to conceive that a small-calibre bullet coming into direct contact with the omentum could do anything but perforate it. It, therefore, appears clear that in a displacement like that figured, only lateral impact occurred with the omentum, which was carried along by the spin and rush of the bullet into the canal of exit, where it lodged.

_Results of injury to the intestine._ 1. _Escape of contents and infection of the peritoneal cavity._--I think there is little special to be said on this subject. The escape of contents into the peritoneal cavity was by no means free, unless the injury was multiple. Thus in one case of injury to the small intestine, No. 166, on which I operated, there was absolutely no gross escape until the bowel was removed from the abdominal cavity, when the contents spurted out freely. In one case of very oblique injury to the colon there was a considerable quantity of fæcal matter in a localised space, but as a rule the ordinary condition best described as 'peritoneal infection' from the wound was found. The bad effect of anything like free escape was well shown in multiple perforations; in these suppurative peritonitis rapidly developed and the patients died at the end of thirty-six hours or less. A typical case is quoted in No. 168.

2. _Peritoneal infection, and general septicæmia._--As is evident from the results quoted among the cases, the degree which this reached varied greatly. It may of course be assumed that in some measure it occurred in every case in which the bowel was perforated, but it was sometimes so slight as to be scarcely noticeable. This may be said to have been most common in injuries to the large intestine. Wounds of the cæcum, ascending and descending colon, the sigmoid flexure, or the rectum, were sometimes followed by no serious symptoms, either local or general. Again in these portions of the bowel the development of local signs, and the later formation of an abscess, were by no means uncommon.

In the case of the small intestine I never observed this sequence, and the same may be said of the transverse colon, which in its anatomical arrangement and position so nearly approximates to the small bowel. In suspected wounds of these portions of the bowel either the symptoms were so slight as to render it doubtful whether a perforation had occurred, or marked signs of general peritoneal septicæmia developed, and death resulted.

The condition of the peritoneum in fatal cases varied much. In some a dry peritonitis, or one in which a considerable quantity of slightly turbid fluid was effused, was found. In others a rapid suppurative process, accompanied by the effusion of large quantities of plastic lymph, was met with. My experience suggested that the latter condition was the result of free infection from multiple wounds of the gut, the former the accompaniment of single wounds. Hence I should ascribe the difference mainly to the extent of the primary infection.

This is perhaps a suitable place to further discuss the explanation of the escape of a considerable number of the patients who received wounds of the abdomen, possibly implicating the bowel. Although this was not, I think, so common an occurrence as has been sometimes assumed, yet many examples were met with. Several reasons have been advanced.

(1) Great importance has been given to the fact that many of the men were wounded while in a state of hunger, no food having been taken for twelve or more hours before the reception of the injury. In view of the well-proved fact in these, as in other intestinal injuries, that free intestinal escape does not occur, and that it is usually a mere question of infection, this explanation, in my opinion, is of small importance. It might with far more justice be pointed out that many of these wounded men were for them in the happy position of not having friends freely dosing them with brandy and water after the reception of the injury, and this was possibly an element of some importance.

Some of the men did, however, drink freely, and in one case which terminated fatally a comrade gave a man wounded through the belly an immediate dose of Beecham's pills.

(2) Mr. Treves has suggested that the effect of the severe trauma on the muscular coat of the bowel is to cause a cessation of peristaltic movement. This, as in the case of 'local shock' elsewhere, may no doubt be of importance, and to it should be added the simultaneous cessation of abdominal respiratory movements in the segment of the belly wall covering the injured part. The occurrence of general cessation of peristaltic movement is, however, to some extent opposed by the fact that in a certain number of the cases early passage of motions was seen just as happens in the intestinal ruptures seen in civil practice.

I should be inclined to ascribe the escape from serious infection in these injuries to the same cause which accounts for their comparative insignificance in other regions--namely, the small calibre of the bullet and consequent small size of the lesion: in point of fact to the minimal nature of the primary infection. I very much doubt if any patient who had more than one complete perforation of the small intestine got well during the whole campaign. This opinion is, moreover, supported by the fact that the prognosis was so far better in cases of injury to the large than to the small intestine, in which former segment of the bowel we have the advantages of a position beyond the region in which intestinal movement is most free, the unlikelihood of multiple injury, and a drier and more solid type of fæcal contents.

In the instances in which recovery followed perforating injuries without any bad signs we can only assume a minimal infection, and sufficient irritation and reaction on the part of the bowel to produce rapid adhesion between contiguous coils, and thus provisional closure.

The other mode of spontaneous recovery which I saw several times take place in the injuries to the large bowel consisted in the limitation of the spread of infection by early adhesions and the development of a local abscess. The non-observance of this process in any case of injury to the small intestine raises very great doubts in my mind as to the frequent recovery of patients in whom the small intestine was perforated.

INJURIES TO THE INTESTINAL TRACT

1. _Wounds of the stomach._--A considerable number of wounds in such a situation as to have possibly implicated the stomach were observed, and of these a certain number recovered spontaneously. The only two instances that came under my own observation are recorded below. It will be noted that in each the special symptoms were the classic ones of vomiting and hæmatemesis. In the first case blood was also passed per anum, and in the second the diagnosis was reinforced by the escape of stomach contents from the external wound.

The second case was a surgical disappointment. No doubt the fatal issue was mainly dependent on the fact that the external wound had to be kept open to allow of the escape of the abundant discharge from the wounded liver. In the absence of the hepatic wound, however, I believe it would have been possible for this patient to have got well spontaneously, in view of the firm adhesions which had formed around the opening in the stomach, and the consequent localisation which had been effected. Another unfortunate element in this case was the comminuted fracture of the seventh costal cartilage, which maintained the patency of the aperture of exit. The latter point, however, was of doubtful importance from this aspect, as the vent provided for the gastric and biliary secretions may have been the safety-valve that had allowed localisation to develop.

I believe that the secondary hæmorrhage was the main element in robbing us of a success in this case, and that this depended on the digestion of the wound by the gastric secretion. The early troubles which arose in the treatment of this patient well illustrate the difficulties by which the military surgeon is at times met; but the patient was admirably attended to and nursed by my friend Mr. Pershouse, and an orderly who was specially put on duty for the purpose.

(163) Wounded at Rensburg. _Entry_ (Mauser), in ninth left
intercostal space in posterior axillary line; _exit_, a
transverse slit 1/2 an inch in length to left of xiphoid
appendage. Patient was retiring when struck; he did not fall,
but ran for about 1,000 yards, whence he was conveyed to
hospital. He vomited half an hour after the injury (last meal
bread and 'bully beef,' taken two hours previously), and during
the evening three times again, the vomit consisting mainly 'of
dark thick blood.' He was put on milk diet, and not completely
starved; on the third day a large quantity of dark clotted
blood was passed per rectum with the stool, and this continued
for two days.

Ten days after the injury the temperature was still rising to
100°, and did not become normal till the fourteenth day. The
pulse averaged 80. The abdomen, meanwhile, moved fairly well,
respirations 18 to 20. Some tenderness was present in the
epigastrium and towards the spleen. Resonance throughout.
Ordinary diet was now resumed, and beyond slight epigastric
pain on deep inspiration, no further symptoms were observed,
and the patient left for England at the end of the month. The
spleen may have been traversed in this patient, as well as the
lower margin of the right lung.

(164*) Wounded at Enslin. _Entry_ (Mauser), 3/4 of an inch from
the spine, opposite the eighth intercostal space; _exit_,
through the seventh left costal cartilage, 1 inch from the
median line. The patient was lying in the prone position when
shot: he vomited blood freely, and the bowels acted three times
before he was seen forty hours after the accident, each motion
containing dark blood.

On the commencement of the third day the patient's expression
was extremely anxious, and he was suffering great pain. Pulse
96, temperature 100°. Tongue moist, occasional vomiting, bowels
open yesterday. Has taken fluid nourishment since injury. The
abdomen moved with respiration, but was moderately distended,
especially in the line of the transverse colon; it was
tympanitic on percussion, there was no dulness in the flanks,
and only moderate rigidity of the wall on palpation. Frothy
fluid stained with bile and fæcal in odour was escaping from
the wound of exit, and the everted margins of the latter were
bile-stained.

A vertical incision was carried downwards from the wound for 4
inches. A rugged furrow was found on the under surface of the
left lobe of the liver; the stomach was contracted and firmly
adherent by recent lymph to the under surface of the liver and
the diaphragm. The transverse colon was much distended. On
separating the stomach a slit wound was found at the lesser
curvature, immediately to the right of the oesophagus. This
wound was closed with some difficulty with two tiers of
sutures; the cavity was mopped out, and then irrigated with
boiled water; a plug was introduced along the line of the
furrow in the liver, and the lower part of the abdominal
incision closed.

The patient stood the operation well, and was removed to his
tent; during the day, however, two thunder showers occurred
during each of which water, several inches if not a foot deep,
rushed through the camp. After the second flood he was removed
to the operating room, the only house we had, and slept there.
The pulse rose to 120, and respiration to 26, and there was
pain, which was subdued by 1/3 grain of morphia, administered
subcutaneously. A fair amount of urine was passed, and the
bowels acted once, the motion containing blood.

On the second day after operation there was some improvement;
the pulse still numbered 116, and the temperature was raised to
100°, but the belly moved fairly, and pain was moderate.
Abundant foul-smelling, bile-stained discharge came from the
wound when the plug was removed. Rectal feeding was
supplemented by small quantities of milk and soda by the mouth.

The condition did not materially change, but on the fourth day
it was evident that the suturing of the stomach wound had given
way, and liquid food escaped readily when taken. The discharge
remained bile-stained and very foul. No extension of
inflammation to the general peritoneal cavity occurred, but it
was evident that the patient was suffering from constitutional
infection from the foul wound, the lower part of which opened
up somewhat after the removal of the stitches on the seventh
day. The wound was irrigated three times daily with 1-300
creolin lotion, but remained very foul. The man slowly lost
strength, although escape from the stomach considerably
decreased. On the tenth day a sudden severe hæmorrhage
occurred, presumably from a large branch of the coeliac axis.
The bleeding was readily controlled by a plug, and did not
recur; but the patient rapidly sank, and died on the twelfth
day after the operation, and fourteen days after reception of
the injury. No _post-mortem_ examination was made.

2. _Wounds of the small intestine._--These were comparatively common, but offered little that was special either in their symptoms or the results attending them. Wounds were met with in every part of the small gut; but I saw no case in which an injury to the duodenum could be specially diagnosed.

As to the symptoms which attended these injuries, it is somewhat difficult to speak with precision, and it must be left to my readers to form an opinion as to how many of the cases recounted below were really instances of perforating wounds. My own view is that in the majority of the cases that got well spontaneously, the injury was not of a perforating nature, and that for reasons which have been already set forth. It will, however, be at once noted that in all the five cases in which the injury was certainly diagnosed in hospital death occurred.

The cases of injury to the small intestine are perhaps best arranged in three classes.

1. Those who died upon the field, or shortly after removal from it. In these the external wounds were often large, the omentum was not rarely prolapsed, and escape of fæces sometimes occurred early. Shock from the severity of the lesion, and hæmorrhage, were no doubt important factors in the early lethal issue in this class. Many of the injuries were no doubt produced by bullets striking irregularly, by ricochets, by bullets of the expanding forms, or by bullets of large calibre. As being beyond the bounds of surgical aid, this class possessed the least interest.

2. Cases brought into the Field, or even the Stationary hospitals, with symptoms of moderate severity, or even of an insignificant character, in which evidence of septic peritonitis suddenly developed and death ensued.

3. Cases in which the position of the wounds raised the possibility of injury to the intestine, but in which the symptoms were slight or of moderate severity, and which recovered spontaneously.

The whole crux in diagnosis lay in the attempt to separate the two latter classes, and, personally, I must own to having been no nearer a position of being able to form an opinion on this point, in the late than in the early stage of my stay in South Africa. The advent of peritoneal septicæmia was in many instances the only determining moment. On this matter I can only add that, in civil practice, an exploratory abdominal section is often the only means of determination of a rupture of the bowel wall.

With regard to the cases of suspected injury to the bowel which recovered spontaneously, the symptoms were somewhat special in their comparative slightness, and in the limited nature of the local signs. Thus the pulse seldom rose to as much as 100 in rate, 80 was a common average. Respiration was never greatly quickened, 24 was a common rate. The temperature rarely exceeded 100°. Vomiting was occasionally severe, but usually not persistent, ceasing on the second day. A good quantity of urine was passed. As to the local signs, these again were of a limited nature; distension did not occur, or was slight; movement of the abdominal wall was only restricted in the neighbourhood of the wound, the affected area amounted to a quarter, or at most half, the abdominal wall, and rigidity was localised to a similar segment. Local tenderness usually existed; but, as a rule, there was little or no dulness to point to the occurrence either of fluid effusion or a considerable deposition of lymph.

Again many of the patients suffered with very slight symptoms of constitutional shock, although there was considerable variation in this particular.

(165*) Wounded at Graspan, sustaining a compound fracture of
the fibula. While being carried off the field, a second bullet
(Lee-Metford) entered immediately outside the left posterior
superior iliac spine, perforated the pelvis, and emerged 1-1/2
inch within the left anterior superior spine. The patient was
then put down and left on the field ten hours; later he was
carried to shelter for the night, and arrived at Orange River
on the second day. He suffered with some pain in the abdomen,
especially during the journey in the train, but was not sick;
the bowels were confined.

When seen on the third day at 6 P.M., some pain was complained
of in the abdomen, which moved freely in the upper part, but
was motionless below the umbilicus. No distension. Tenderness
around wound of exit and some rigidity. The bowels had acted
four times during the day; motions loose, dark brown, and
containing no blood. Face not anxious, eyes bright, temperature
102°. Pulse 96, regular, and of good strength. Tongue moist and
little furred.

The abdomen was opened at 5 A.M. on the fourth day, as the
local signs had become more pronounced, and the patient had
passed a restless night in great abdominal pain. A local
incision was chosen, as the wound was presumably in the sigmoid
flexure. The sigmoid flexure was adherent to the abdominal wall
opposite the wound of exit, and a dark ecchymosed patch was
found, but no perforation could be detected. Foul pus and gas
escaped freely from the pelvis, but no wound of the large bowel
could be discovered here. On enlarging the incision upwards
three openings were found in a coil of jejunum, probably that
about five feet from the duodenal junction usually provided
with the longest mesentery. No fourth opening could be found.
The openings were circular, about 1/3 inch in diameter, clean
cut, with a ring of everted mucous membrane, and the wall of
the bowel in the neighbourhood was thickened. All three
openings were included within a length of 2-1/2 inches. There
was no surrounding ecchymosis of the bowel wall. Very little
escaped intestinal contents were found in the situation of the
bowel. The latter had apparently been retracted upwards, and
lay to the left of the lumbar spine. The wounds were readily
closed by five Lembert's sutures, three crossing the openings,
and one at each end. The belly was then washed out with boiled
water and closed. The delay in finding the wounds due to the
mistaken impression that they would be found in the pelvis
materially prolonged the operation, which lasted an hour and a
half. The patient never rallied, and died seventeen hours
later. It is possible that a wound in the sigmoid flexure was
present which had already closed at the time of operation.

(166*) Wounded at Magersfontein. _Entry_ (Mauser), opposite
central point of left ilium; _exit_, 1-1/2 inch above the
centre of the right Poupart's ligament. Vomiting commenced soon
after the injury, and this was continuous until the patient's
arrival in the Stationary hospital on the fourth day, when the
condition was as follows:--

Face extremely anxious in expression. Temperature 101°,
sweating freely. Pulse 110, fair strength. Tongue moist.
Abdomen much distended, rigid, motionless, tympanitic
throughout. Bowels confined. No urine had been passed for
twenty-four hours, [Symbol: ounce]ij in bladder on
catheterisation, clear, and containing no blood.

Abdominal section. Median incision. A considerable quantity of
bloody effusion was evacuated. Intestine generally congested
and distended. No lymph. Two wounds were found in the ileum on
the opposite sides of one coil; the openings were circular,
with the mucous membrane everted. No escape of fæcal matter was
visible until the intestine was delivered, when intestinal
contents spurted freely across the room. The openings were
sutured with five Lembert's stitches. The bowel was punctured
in two places to relieve distension, and then returned into the
belly, after washing with boiled water.

Four pints of saline solution were infused into the median
basilic vein, and 1/30 grain strychnine sulph. was injected
hypodermically.

The patient did not rally, and died twelve hours after the
operation.

(167*) Wounded at Graspan. _Entry_ (Lee-Metford), midway
between the umbilicus and pubes; _exit_, 1 inch to the left of
the fifth lumbar spine. The patient was seen on the third day
in the following condition: in great pain, expression extremely
anxious, vomiting constantly. Pulse 150 running, respirations
48. Temperature 100°, sweating freely. Great distension,
rigidity, and general tenderness of immobile abdomen. No
improvement followed the administration of brandy and
hypodermic injection of strychnine 1/30 grain, and operation
was deemed hopeless.

In the evening the patient was apparently dying. Face blue and
sunken and covered with sweat, eyes dull, speechless, pulse
imperceptible, restlessness extreme, bowels acting
involuntarily, no urine in bladder.

The man was placed in a tent by himself, and to my surprise was
alive and better the next morning; the expression was still
anxious, but the face brighter and not sweating; the pulse
only numbered 100, but was very weak, and the hands and feet
were cold. The condition of the abdomen was unaltered, but the
thoracic respiration had decreased in rapidity from 48 to 28.

His condition still seemed to preclude any chance of successful
intervention, but none the less life was retained until the
morning of the seventh day, the state alternating between a
moribund one and one of slight improvement. He was lucid at
times, although for the most part wandering, and was so
restless that no covering could be kept upon him. Vomiting was
continuous, so that no nourishment could be retained; the
bowels acted frequently involuntarily, and little or no urine
was passed. Meanwhile, the abdomen became flat, then sunken, an
area of induration and tenderness about 6 inches in diameter
developing around the wound of entry. Slight variations in the
pulse, and from normal to subnormal in the temperature, were
noted, and death eventually occurred from septicæmia and
inanition.

(168*) Wounded at Driefontein. _Entry_ (Mauser), above the
posterior third of the left iliac crest, at the margin of the
last lumbar transverse process (probably through ilio-lumbar
ligament); _exit_, 1 inch below and to the left of the
umbilicus.

The patient was wounded at 3 P.M., but not brought into the
Field hospital until 9 P.M., when the temperature of the tents
was below 28°F. He was considerably collapsed, suffering much
pain, and vomited freely. The abdomen was flat, but very
tender. Bowels confined. The column had to move at 5 A.M. the
next morning, when the temperature was still near freezing, and
during the day continuous fighting prevented any chance of
operation. The man steadily sank during the day, and died
thirty-six hours after the reception of the injury.

_Post-mortem condition._--Belly not distended, dull anteriorly
in patches, and right flank dull throughout. When the belly was
opened, extensive adhesion of omentum and intestine enclosing
numerous collections of pus were disclosed, and on disturbing
the adhesions a large collection of turbid blood-stained fluid
was set free from the right loin. The great omentum was much
thickened and matted, with deposition of thick patches of
lymph; very firm recent adhesions also united numerous coils of
small intestine. The pus was foetid, but no appreciable
quantity of intestinal contents was detected in it. The lower
half or more of the small intestine was injected, reddened, and
thickened. The wounds which were situated in the lower part of
the jejunum and ileum were multiple, and seven perforations
were detected; besides these the intestine was marked by
bruises, and some gutter slits affecting the serous and
muscular coats only. Considerable ecchymosis surrounded these
latter. The clean perforations were circular, less than 1/4
inch in diameter, and for the most part closed by eversion of
the mucous membrane. Intestinal contents were not apparent, but
escaped freely on manipulation of the bowel.

(169*) Wounded at Magersfontein. _Entry_ (Mauser), over the
eighth rib in the anterior axillary line; _exit_, 1 inch to the
left of second lumbar spinous process, just below the last rib.
Vomiting commenced almost immediately after reception of the
injury, and the bowels acted frequently. This condition
persisted until the fourth day, when the patient was brought
down to Orange River, and the signs were as follows.
Considerable pain in left half of abdomen, pulse 110, fair
strength, temperature 101°. Some general distension of abdomen
with complete disappearance of hepatic dulness. Some movement
of right half of abdomen, left half immobile, dulness extending
from the flank as far forwards as linea semilunaris. An
incision was made in left linea semilunaris, and Oj blood
evacuated from the left loin. There was no lymph on the
intestines nor sign of inflammation. No perforation was
discovered in either stomach or intestine, but on two coils of
jejunum there were deep slits 3/4 inch long, extending through
both peritoneal and muscular coats. Beyond these wounds, on
other coils oval patches of ecchymosis, due to direct bruising,
were present. The peritoneal cavity was sponged free of all
blood and irrigated with boiled water; no bleeding point was
discovered, and the abdomen was closed.

The next morning the patient was comfortable; temperature
100.2°, pulse 100. Tongue clean and moist; he vomited once
during the night.

Some bloody discharge had collected in the dressing, and at the
lower angle of wound there was a local swelling, apparently in
the abdominal wall. The flank was resonant.

During the afternoon the patient became faint, and when seen at
6 P.M. was in a state of collapse, in which he shortly died.

Death was apparently due to renewal of the previous hæmorrhage.
No _post-mortem_ examination was made.

(170*) Wounded at Magersfontein. _Entry_ (Mauser), 1/2 inch to
the left of the second sacral spine; _exit_, immediately below
the left anterior superior iliac spine; the patient was
kneeling at the time, and the same bullet traversed his left
thigh in the lower third. When seen on the third day, the
lower part of the abdomen was motionless, tumid, and tender.
The bowels had been confined for three days; there had been no
sickness, and the tongue was moist and clean. Temperature 100°,
pulse 90, fair strength, respirations 38. The patient had once
had an attack of acute appendicitis, and he himself said he was
sure he now had 'peritonitis,' as he had pain exactly similar
in the belly to that he had suffered in his previous illness.

No further signs, however, developed under an expectant
treatment, and he remained some two months in hospital, while
the wound in the thigh and a third injury to the elbow-joint
were healing.

(171) _Entry_ (Mauser), at the highest point of the left crista
ilii; _exit_, through the right ilium, 2 inches horizontally
anterior to the posterior superior spine. Absolutely no
abdominal symptoms followed. The bowels were confined five
days, and then opened by enema. The patient complained of some
stiffness in the lumbo-sacral region, but the right
synchondrosis was no doubt implicated in the track.

(172) Wounded at Paardeberg (range 800 yards). _Entry_
(Mauser), 2 inches diagonally below and to the right of the
umbilicus; _exit_, not discoverable. For the first two days the
patient had to lie out with the regiment; on the fourth he was
removed to the Field hospital. During the first three days the
patient vomited (green matter) frequently, and the belly was
hard and painful; as biscuit was the only available food, no
nourishment was taken. The bowels acted on the second night. At
the end of a week the patient was sent by bullock wagon (three
days and nights) to Modder River, and then down to Capetown,
where he walked into the hospital on the thirteenth day,
apparently well.

Two days later the temperature rose to 104°, and enteric fever
was diagnosed, no local signs pointing to the injury existing.
The patient made a good recovery.