With regard to vomit, the first thing to consider is the quantity vomited at one time. Nurses should be trained to estimate this carefully and also to preserve specimens on all occasions. In dilated stomach vomiting usually does not occur more than once daily, sometimes only every second or third day, and the quantity at any time is correspondingly large.

In ulcer of the stomach a considerable portion of the last meal may be brought up within an hour or two of its ingestion, and the pain it has caused be thereby relieved.

The smell should be considered, a yeasty smell being characteristic of dilation of the stomach, a habitually foetid odour of cancer of the stomach, and a feculent odour of intestinal obstruction.

Vomit is usually acid in reaction ; but it may be alkaline in some cases of chronic dyspepsia, or when there is much blood present.

The most important abnormal constituent of vomit is blood. In large quantities its nature is obvious, and the event is suggestive of simple ulcer; but in cirrhosis of the liver profuse haematemesis may occur owing to rupture of dilated veins. In smaller quantities the vomit has a characteristic dark appearance, resembling coffee grounds, and this may be due to cancer or simple ulcer. When the existence of blood in vomited matter is doubtful, the most reliable guide is the haemin test, which may be done in the following manner : evaporate a small quantity of the gastric contents to dryness, powder the residue and place some along with a crystal of common salt on a microscopic slide, add a drop of glacial acetic acid and boil over a spirit lamp, cover with a cover glass and examine under a high power for the small, dark-brown crystals of haemin. As a rule it is not necessary to add sodium chloride, since fresh blood contains sufficient of it ; but since excess of the salt does not interfere with the reaction it is well to use a crystal or two.

In cancer of the stomach blood is frequently present in the vomit, often in small, sometimes in considerable, only rarely in large quantity.

Pus is sometimes, but not often vomited. In considering both pus and blood in a fluid said to have been vomited, it must be remembered that when large quantities of fluid are expelled from the lungs-e. g. on the rupture of an empyema into the lung, or a profuse haemoptysis-the sensation to the patient is often as if vomiting had occurred. The presence of food and the general absence of frothi-ness will help to distinguish true vomit, while vomited blood is generally much darker than blood from the lungs. But the only reliable way to make a distinction is to inquire carefully into the facts of the occurrence. Pus in the vomit may arise from an empyema of the gall-bladder, or a pancreatic or other abscess bursting into the stomach or oesophagus.

Examination by the microscope of vomited material is usually of secondary importance, but it sometimes affords great assistance, as in the case of a subdiaphragmatic abscess under my care bursting into the lung*, where the presence of half-digested muscular fibres and the absence of elastic tissue distinctly proved the source of the pus to be from the stomach and not from an abscess of the lung or an empyema; and in some cases of cancer where portions of growths or groups of cells are occasionally obtained by means of lavage. In dilatation of the stomach the sarcina ventriculi is frequently to be seen together with yeast cells. In cancer, where maeros-copically there is no evidence of blood, red blood-corpuscles may often be found on microscopic examination.

An examination of the faeces is said to show the presence of blood by chemical tests almost constantly when there is cancer of the stomach, but only occasionally in case of ulcer. My personal experience of this diagnostic sign is not sufficiently extensive for me to speak positively as to its value, though I have found it of use in some cases.

An examination of the urine for nitrogen is said to be of use in the diagnosis of cancer, as in every case of malignant disease it is said to be considerably reduced from the normal.

A diagnosis of the position of a gastric tumour has been claimed by Glaessner by means of an examination of the stomach contents.

Having determined that the gastric mucous membrane can be divided into two physiologically distinct segments, the fundus, which has a large supply of glands, and the pylorus, which has but few glands, K. Glaessner (4) describes how the localisation of tumours can be made. Pepsin and rennet are both produced by the mucosa of the fundus, but only pepsin and no rennet is secreted by that of the pylorus. It therefore occurred to Glaessner that if one examined the contents of the stomach after a trial meal in case of gastric tumour one might be able to learn more of the situation of the growth. In the cases in which he was able to test this he estimated the total acidity by means of phenol-phthaleine HCl by Toepfer's reagent, pepsin by Mett's method, and rennet by its direct action on milk, within a given time. He considers that normally the pepsin should be present at 5 millimetres-that is, that 10 cubic centimetres of normal gastric juice should be able to completely digest a column of albumen in Mett's test tube measuring 5 millimetres in twenty-four hours; and that rennet should have the value of 1 in 100, that is, that 0.1 cubic centimetre of normal neutralised juice should be able to coagulate 10 cubic centimetres of milk at 30° to 40° C. within half-an-hour. In six cases of carcinoma of: the pylorus, confirmed at the operation or at the necropsy, the pepsin was between 1 and 3 millimetres, while the rennet was normal. In seven cases of carcinoma of the fundus he found that not only was the pepsin much diminished, as in the pyloric cases, but the rennet also was diminished, or was entirely absent. He looks upon this method of diagnosis as highly valuable.