a month before his admission. It is, of course, possible that the spongy gums and decayed teeth might be looked upon by some as sufficient to allow of the entrance of blood-destroying poisons. The post-mortem examination also showed nothing that could be regarded as a portal of infection. There were no ulcers of the intestine, although it is quite probable that the gastro-intestinal mucous membrane was not quite healthy. We are therefore left pretty much to conjecture and theory in our efforts to unravel the essential nature of the disease. The general appearance of the patient, with his bleeding gums, reminded me very much of a case of hæmophilia which I once observed in the wards. In endeavouring to estimate the essential nature of the case, we must remember the inducer of activity in the lymphatic glands, in the thymus, and in the thyroid; and it seems to me that we might almost be entitled to regard the disease as of the nature of a neoplasm of the lymphatic glands and blood. But this is a mere conjecture. In a previous case of spleno-medullary leukæmia (International Clinics, 1898) careful attempts to cultivate organisms from the blood failed; and I am, on the whole, of opinion that leukæmia is not a bacterial disease.

From the pathological standpoint, perhaps the most astonishing feature was the green colour of the blood-clot in the large vessels and of the subcutaneous nodules and lymphatic glands. This feature at once raises the question of the relationship of acute lymphatic leukæmia to that rare form of greencoloured tumour which has been named chloroma. Certainly, if the colour of the tumour on section determines the name, then the subcutaneous nodules in this case were chloromata. That many of them were actual neoplasms Dr. Ferguson's careful histological report abundantly proves, and there can, I think, be little doubt that in a case such as the present the nodular developments, the altered state of the blood, and the hyperplasia of the lymph glands, thymus, thyroid, and bonemarrow are all part of the one general disorder.

PERFORATIVE PERITONITIS IN ENTERIC FEVER.

BY ANDREW LOVE, M.B., CH. B.,

Senior Resident Assistant Physician, Belvidere Fever Hospital, Glasgow.

DURING the two years and eight months extending from July, 1900, to March, 1903, 948 cases of enteric fever were admitted to Belvidere Fever Hospital, Glasgow, and of these 13 (or approximately 1 per cent) suffered from peritonitis; 10 of these were operated on, and 8 were found to be due to perforation of the bowel; of the remaining 2, 1 was due to a ruptured mesenteric gland, and the other was "idiopathic" peritonitis. Perforation of the bowel in enteric fever, while not necessarily fatal, is so frequently followed by death that sometimes one wonders if operation is advisable; but bearing in mind the number of recoveries--though small-which have been placed on record, and that practically no cases of general peritonitis due to perforation have ever been known to recover if left to nature, the surgeon must be summoned whenever a a perforation has been diagnosed.

The great factor in the non-success of this operation is the difficulty of early diagnosis. As the surgeon is constantly reminding us, his chances of success diminish rapidly with the duration of peritonitis, so that early diagnosis is extremely important, and yet this is very difficult. When perforation of the stomach or intestine occurs in persons who are otherwise in possession of their faculties, the onset is of so severe and striking a character as to compel the attention of even the most unobservant, but in enteric fever the exact opposite is often the case. The patient has been weakened both physically and mentally by a fever which has lasted for at least a fortnight. fortnight. His body no longer responds to the accident which has befallen him, and the damage may easily become so extensive as to be beyond remedy before attention has been drawn to the matter. The early diagnosis of perforation must depend on the accuracy of the nurse's observations, and this accuracy can only be gained by actual experience of previous cases; and as the opportunity of gaining experience is very limited, it will be seen that the art of the surgeon is severely handicapped. A typical case of perforation of the bowel or stomach is easily diagnosed, but how many of these are seen in enteric fever?-practically none; and even when typical signs are present there is

always the suspicion that the perforation has occurred some hours previously, and that the so-called classical symptoms are really manifested only when the perforation is enlarged and peritonitis has set in. It is in order to emphasise the necessity for the earliest possible diagnosis and operation in perforation in enteric fever that the following series of cases is placed on record. In this connection, it is worthy of note that it is stated by Dr. Harvey Cushing, of Baltimore, that not only can perforation be diagnosed immediately it has occurred, but that frequent leucocyte counts, revealing a marked and progressive leucocytosis, will enable one to detect the perforative stage. Unfortunately, counts were not made in all of the cases observed here, but the results in those in which counts were made do not seem to bear out Dr. Cushing's statement. It is but fair to state, however, that only isolated and not systematic counts were made in the cases here reported.

It is important that this leucocytosis question should be thoroughly worked out, because if it can be ascertained by this process when perforation is actually occurring, there is a much better chance of operation proving more. successful than hitherto. Of the thirteen cases here recorded, four were not operated on. Three of these (Cases VI, XII, and XIII) were so enfeebled and so evidently dying that operation was considered useless. In the other (Case XI) the surgeon decided not to operate, as the symptoms had subsided by the time he saw the case, and the patient presented then but little appearance of perforation. In all these cases a perforation was found post-mortem. Of the remaining ten cases which were operated on, two were not perforations of the bowel, one being a ruptured mesenteric gland (Case II), and the other (Case IX) a severe case of peritonitis which simulated perforation, and which is therefore especially worthy of attention. In the table presented at the end the cases are summarised.

The number of cases observed is too small to have any value as an observation on the influence of age in the occurrence of perforation. It is interesting, however, to note that nine of the cases were under 20 years of age, while only two were over 25.

Murchison points out that perforation is more frequent in males, even though the total death-rate in females is somewhat in excess. In the limited number of cases considered here, this observation is completely borne out, eight being males and five females.

A glance at the table will demonstrate the interesting fact that the average time at which perforation occurred was the fourth week, this, again, being, according to Murchison, the most probable time for that complication.

The earliest day on which perforation occurred was the fifteenth day, and the latest the thirty-seventh day of illness. In all the cases the disease was still in the acute stage at the time of perforation.

According to Treves, the duration of life in acute septic peritonitis, apart from perforation in enteric fever, is from about thirty-six hours to seven days. In enteric fever, as one would expect, the course is more rapid, for although one of the cases under consideration lived ninety-two hours after the onset of peritonitis, the average duration of life was only thirty-six hours. Six of the cases died in less than twentyfour hours, viz., in four, twelve, fourteen, fifteen, and sixteen hours respectively; one case lived thirty-six hours, another forty-eight hours, another fifty-three hours, and a fourth fifty-six hours.

Hæmorrhage was a marked precursor of perforation in three cases, and diarrhoea was present in all but two.

Attacks of pain occurring from one to three days before actual perforation were common. They were often very severe, but not usually prolonged, and were referred mostly to the right iliac fossa. In eight of the cases sudden acute abdominal pain was complained of; in three the pain was slight; and in two there was no pain whatever. Whether severe or slight, the pain was invariably confined at first to the right iliac fossa, spreading over the whole abdomen, however, later on. In five cases, therefore (or 38 per cent), the onset was gradual, and more or less obscure.

Abdominal distension was noted in four of the cases only. In three the tympanites was very marked, the liver dulness being completely obscured, whilst in the other-the case of peritonitis without perforation-there was only a very small area of liver dulness.

Rigidity of the abdominal wall, partial or complete, was present in every case. This is the most important objective symptom there is, and the one upon which most reliance can be placed, occurring as it does-at least at the onset-in practically all cases of perforation. True, it is got also in peritonitis without perforation, but this occurrence, as said before, is so uncommon that it may be almost disregarded. It is specially to be noted that the rigidity is usually local at first, and mainly confined to the right iliac region, that

it is sometimes general, and that in the severest cases it occasionally passes off altogether after the lapse of a few hours.

Fluctuation was obtained in four of the cases. It is somewhat uncommon to get this symptom early, but in a few hours after perforation occurs, a fluid wave is easily obtained. Frequently the fluid is detected first in the right iliac fossa and in the right kidney pouch.

The temperature at the time of perforation gives no indication whatever of the abdominal condition, perforation occurring apparently independently of a rising or falling temperature, but within a couple of hours a fall of several degrees was noted in all the cases here reported. Neither of the cases in which a rise occurred proved on operation to be peritonitis due to perforation; in the one case the peritonitis was due to a ruptured mesenteric gland, and in the other probably to infection through the damaged bowel wall. In the latter case, also, the increase of fever was accompanied by rigor.

The pulse, though invariably quickened at the moment of perforation, often remains otherwise unaltered; gradually, however, it becomes small and rapid, the slightest pressure obliterating it, so that difficulty is experienced in counting it accurately. In a few cases it was observed that within an hour the pulse-rate had fallen, and then begun to rise slowly and diminish in tension.

The behaviour of the respiration is especially important, as in all cases it is quickened and becomes thoracic, the abdominal wall scarcely moving at all.

Vomiting was noted in only one case at the onset of the symptoms, but frequently occurred along with hiccough just before death.

The facies of the patient when perforation occurs is very suggestive, even when marked pain is absent. There is a look of deep anxiety; the eyes are sunken and, as in typhoid cases generally, are brilliantly clear, with the pupils dilated. Sometimes a hectic flush is noticed in the malar region, but as a rule the face is grey, pinched-looking, and hollow.

CASE I. The first case was that of J. A., æt. 17 years, who was admitted to hospital on 5th November, 1900, having become ill fourteen days previous to that date. On admission, patient looked sharply ill. His abdomen was distended, and numerous rose spots were scattered over it. His tongue was moist. Temperature on admission was 1014° F.; pulse, 84;