whenever more than a certain amount of food is taken-an attempt to take more than their capacity, being immediately followed by vomiting. There may be a history of a progressive diminution in this capacity.

In later stages of the disease, when the patient is reduced, the vomiting may be a process of gentle regurgitation rather than a forcible expulsion of the stomach contents.

In 74 cases the vomiting was frequent; in 37 fairly often, in 17 only occasional. Of the 74 cases with frequent vomiting 20 came to autopsy. Among these there was ulceration in 11. The growth was at the pylorus in 16, general in 3, and in 1 mural. Of the 27 cases with moderate vomiting, 11 came to autopsy. Ulceration was present in 10 of these. The growth was pyloric in 4, situated in the walls in 4, cardiac in 2, and in 1 it was general. Of the 17 cases in which vomiting was rare there were 3 autopsies. Ulceration was present in 2 cases; one of these was a pyloric growth, the other two were in the walls. In only 1 case was there continued peristent vomiting; this was a case of growth at the pylorus with stenosis.

From the autopsy and operation records an attempt was made to establish some connection between the situation of the growth and the character of the vomiting, especially in relation to early onset and to the influence of food. Among 27 cases of growth at the pylorus, vomiting was an early symptom in 15, of late appearance in 5, and absent in 5. In 2 cases the history was not clear. In the 3 cases of growth at the cardia vomiting was not by any means a prominent symptom. One case had marked difficulty in swallowing, with frequent regurgitation; one had occasional vomiting only as a late symptom, and in the third it was absent. Of the 6 cases of general involvement, vomiting was an early symptom in 3, came on late in 2 cases, and was absent in 1 instance. Among the 12 cases of growth situated on the walls, it was not marked as an early symptom. It was only present at an early stage in 3 cases, and in 2 of these vomiting only occurred at the onset and not afterward. In 3 cases it came on as a late symptom, and in 6 it was absent throughout. In no case of growth of the walls was it a very severe symptom.

The conclusion from these figures is that vomiting seems less marked, especially as an early symptom in cases of growth situated on the stomach walls.

As to food, it does not seem possible to draw any definite relationship between the situation of the tumour and the occurrence of vomiting at any given time after eating. In a fair proportion of cases with tumours at the pylorus it came on in from half an hour to an hour and a half after food. In the cases of general involvement vomiting usually took place at once after eating, and this sometimes only after more than a certain quantity was taken. But several cases of pyloric growth also gave a history of immediate vomiting. In some instances vomiting eased the pain. This was especially marked in cases in which the taking of food caused pain, or the pain increased gradually in severity until vomiting occurred or was induced, when immediate relief was experienced. Many of these patients were in the habit of inducing vomiting to gain relief. Of 21 cases in which vomiting eased the pain, 5 came to autopsy, and of these in 2 ulceration was present.

Cases without Vomiting.-Of 22 cases there were 11 that came to autopsy, and in 9 of these ulceration was present. This goes to prove that vomiting is not necessarily connected with the presence or absence of ulceration. In two other cases an exploratory operation was done and the situation of the growth determined.

The number of cases, with the situation of the growth and the number in which no vomiting occurred, are tabulated below :

The number of cases with growth in the walls is just one-quarter of the total number, but they contribute almost one-half of the total cases in which vomiting was absent.

These figures are of interest in reference to the positive statement sometimes made that vomiting is never absent when an orifice of the stomach is involved. Six cases out of thirty at the pylorus and cardia were free from vomiting.

The general figures regarding vomiting are as follows :—

Vomiting of Blood.-This occurred in 36 cases, a percentage of 28.1 of the number in which vomiting was present, and of 24 for the total number. The relative figures for the two sexes were males 30, females 6, which is fairly close to the relative total proportion. In two instances, as already noted, it was the first symptom. There is no special peculiarity associated with the vomiting of blood. The blood in 32 cases was dark and altered more or less, in 3 it was bright red, and in one case both occurred at different times. In 9 cases the amount was small, leaving 27 in which the hæmorrhage was fairly profuse. The occurrence of this symptom showed no regularity: in some it was only once and early in the course, in others and the majority, it was a comparatively late symptom and was often repeated. In some cases there was also the passage of altered blood by the bowels. In no case of the series was the hæmorrhage severe enough to cause death, although in several instances it caused fainting and collapse. In the case of some very emaciated patients it hastened the fatal termination. In a case recently in the hospital there was a history of sudden profuse hæmorrhage from the stomach some months before. The patient fainted and was found in an unconscious condition. During the next few days he was only partially conscious, and vomited blood several times. Blood was also passed by the bowels, and on regaining consciousness he found that he was blind. The loss of sight was permanent.

Some cases may be referred to.

No. 99. Large haemorrhages at onset and before death. Case has been referred to as one of sudden onset with hemorrhage; did not vomit in the hospital until two days before death, when she vomited three times. The estimated quantity of blood was 750 c.c.

No. 104. Hæmorrhages of large amount.-J. K., aged 65 years. He gave a history of slight stomach trouble for two years, which had

been much aggravated in the last nine months. Three months before admission, while walking in the street, he vomited a large amount of blood, and had to be assisted home. Three days before admission, while at work, he had another hæmorrhage. He first felt weak and dizzy, and then suddenly vomited a large amount of fresh and clotted blood. While in the hospital there was no further vomiting of blood, but there was always a small amount in the test meals. At autopsy an ulcerated growth was found at the pylorus.

Only three of these cases with hæmatemesis came to autopsy. All showed ulceration. The growth was situated at the pylorus, cardia, and on the posterior wall in these cases.

Vomitus with offensive smell.-The vomiting of foul-smelling material with a peculiar, intense, indescribable odour occasionally occurs. It was present in one case of the series.

No. 147.-C. H. (hosp. no. 21,664), aged 58 years, coloured. The patient was brought to the hospital in almost a dying condition; he was much emaciated. The abdomen was hard and board-like, and nothing could be made out by palpation. He vomited frequently. The vomitus had an intense, horrible, penetrating odour, which filled the whole ward. The probability of cancer of the stomach with perforation into the colon was considered, on account of the character of the vomitus. The patient rapidly sank. At autopsy a sloughing carcinoma of the lesser curvature which extended to the omentum was found. On opening the stomach the same odour was present as the vomitus had during life.

This case shows that the vomiting of offensive material does not mean always communication with the bowels. Dr. Osler has reported two cases with this horrible stench; in one, a patient with a diffuse large cancer of wall, the odour was foul and fæcal, and of astonishing power of diffusion; there was no autopsy. In the other the autopsy showed a sloughing necrotic cancer of the wall.

Lectures on Abdominal Tumours, and University Medical Magazine, 1895.

CHAPTER IV.

SYMPTOMS-continued.

FEVER

APPETITE AND DIGESTION-LOSS OF WEIGHT AND STRENGTH-EXAMINATION
OF STOMACH CONTENTS MICROSCOPICAL EXAMINATION
ENLARGED GLANDS-BOWELS URINE.

Symptoms relating to Appetite and Digestion. In rather a surprising number the appetite was practically normal, though many of these complained that while the desire for food was present, the fear and knowledge of subsequent pain and distress restrained them. There were 26 cases in which the appetite was normal. Autopsies in 7 of these showed ulceration only in one instance-a growth at the cardia. The growths were situated as follows: 4 at the pylorus, of which 2 caused stenosis; 1 at the cardia, 1 general, and in 2 on the walls. There were 17 cases in which the appetite was moderate. Only 2 of these came to autopsy, both of which presented ulcerated mural growths. In two patients the appetite was increased, and they could-to quote their own expressions-"eat a dozen times a day"; this also in spite of the fact that both vomited frequently. A total of 68 cases had a decided anorexia. The special distaste for meat, so often referred to, was present only in a comparatively few cases.

Under the term "dyspepsia" may be grouped a large array of symptoms which are more or less common to all stomach disorders. Exception is taken to the loose use of this term when it may be merely a means of covering ignorance. But it is a convenient term for a number of symptoms, such as discomfort in the epigastrium, feelings of fulness and weight, belching of gas, etc. These symptoms in cancer are probably due to many factors the disturbing effect of the growth itself, the fermentative changes made possible by stenosis and stagnation, ulceration,