necessary to plug both anterior and posterior nares with Belloc's sound or an equivalent. Some recommend the injection of astringents, such as perchloride of iron zi. of the liq. fortior. to 3i. of water. Wherever possible nothing is so efficient and so little liable to be followed by sloughing and ulceration as simple compression. This is the method for arresting hæmorrhage after tooth extraction. The extraction may give the first sign of the existence of the diathesis.

Persistent and continuous bleeding is best averted by a powerful purgative, the patient being kept in the horizontal position. Nature staunches the more acute hæmorrhages by an approach to syncope.

Hæmorrhage from the rectum may be controlled by a smooth cylinder of ice inserted within the sphincter. Injections of iron and very hot water (105° F.) are also lauded.

Liquid extract of ergot in frequent mx. doses has been recommended for urgent bleeding. The hemorrhage is generally from capillary vessels, and ergotine contracts the arterioles; though whether the raised blood pressure would promote hæmorrhage is not answered.

Turpentine, rectified spirits of, in half-dram doses, have been employed. Hæmatemesis has ceased under its administration in my hands, but it frequently fails, not only in "bleeders," but in other cases of hæmorrhage.

Joints.-Large extravasations of blood under the skin and into the joint may be treated by cold wet compresses under oil-silk frequently renewed and associated with uniform careful pressure by bandaging. Blisters and iodine have been tried to reduce the arthritic swellings. I have not used them nor seen them used. The actual cautery should not be employed.

After much bleeding or extravasation of blood profound anæmia with great diminution of the red blood disks is seen.

Here rest in bed, careful nursing and feeding are called for (see also remedies for severe anæmia).

CHAPTER V.

RHEUMATISM.

THERE are great differences between children's rheumatism and that of adults. But the migratory tendency of the arthritic manifestations, the tendency to relapse, and the widespread lesions are some features in common. I believe there is a type of children's rheumatism. A typical case of rheumatism in a child may be pictured as one that presents in the course of a single attack of the malady the following phenomena: Erythemata, subcutaneous nodules, joint affection, endocarditis, and chorea. All these phenomena or links of the rheumatic chain have I seen in the course of a single attack of rheumatism in a child. The children's rheumatism is characterized by the frequency of erythemata, nodules, endocarditis, and chorea, and by the slightness of arthritis, pain, pyrexia, and sweating. Sometimes the swelling of the joint appears to be outside rather than inside the joint. Indeed, there is abundant reason for thinking that tenosynovitis with effusion occurring in the neighbourhood of a joint has been mistaken for actual arthritis.

The slightness of the pain and tenderness rather favours the view that the effusion is not great within the joint. The swellings in children's rheumatism are well seen as cushion-like elevations about the backs of the wrists and the dorsa of the ankles. The painless and inconspicuous subcutaneous nodules, composed of delicate fibrous tissue, are situated about the elbows, patella, and malleoli for the most part; but they may be found on the vertebral spines, the spine of the scapula,

the line of the clavicle, the crest of the ilium, about the extensor tendons of hands and feet, the pinna of the ear, the temporal ridge, the forehead, and the superior curved line of the occiput. They are usually painless, and not noticed by the patient or parents. They may be few in number, isolated, or grouped in large numbers. Their size varies from a pin's head to an almond. The skin is movable over them, and they are not firmly attached to the fibrous structures beneath. They have a remarkable tendency to disappear, but sometimes they persist for weeks and months together. So they may come out in successive crops. They may last only three and four days. I have seen these nodules about a wrist joint in which there was well-marked effusion. Generally, however, the joint effusion is absent where these nodules exist. I have no wish to deny that acute rheumatic fever, such as is seen in the young adult, may not have its exact counterpart in a child, though this in my experience is rare.

Erythema marginatum and papulatum are unquestionably of rheumatic origin, and of very frequent occurrence in children's rheumatism.

Erythema nodosum is doubtless allied to rheumatism, but heart complication is very rare, if it occurs at all, in immediate connection with it.

Urticaria and purpura are sometimes of rheumatic origin. Pericarditis is not uncommon in children's rheumatism. Serous pleurisy is at times of distinctly rheumatic source. Stiff neck and tonsillitis may also be due to rheumatism. Unquestionably chorea is sometimes, but not always, a rheumatic symptom. It is of prime importance to remember that any of the rheumatic phenomena may occur alone, and that any combination of them may be met with. Moreover, they may alternate with one another, and rarely the whole of them may co-exist. Thus we can hardly call any of them

sequele without allowing also that they may be complications of the rheumatic attack. Sometimes the attack is long drawn and straggling, being constructed of a curious succession and combination of symptoms. If one of the symptoms comes first we must be on the qui vive for others. Hence it would not be ridiculous for a doctor to keep a child in bed for urticaria or erythema, even though there were no other sign of illness; for I hold that the frequent endocarditis and pericarditis of children may be partly at least due to the mechanical irritation resulting from the want of rest. In adult rheumatism the joint trouble leads to immobility, and the stress on the heart must therefore be less.

The rheumatism that sometimes follows and sometimes accompanies scarlet fever may be true rheumatism, but, unlike true rheumatism, suppuration of serous effusions is frequent.

The basis of a valvular vegetation in endocarditis may be described as a nodule, and Dr. Barlow regards a subcutaneous nodule in rheumatism as homologous with the valvular nodule. Rheumatism is manifested anatomically in discrete lesions, often grouped together. An erythema may be but one minute spot, but usually many red spots develop in close contiguity. So it is with the valvular beadings and the subcutaneous nodules. Dr. Barlow has described a nodular form of rheumatic pericarditis, and I have seen several fibrous nodules on the wall of the heart in a case of rheumatic pericardial adhesions.

I have said the pyrexia of rheumatism is slight; the temperature seldom goes beyond 100°. Occasionally, however, a high fever is registered, but hyperpyrexia is very rare. And so are the cerebral phenomena headache (see chapter on Headache), delirium, sleeplessness, and unrest-uncommon in children's rheumatism. The heart is almost invariably the seat of a murmur during some part of the

rheumatic career. These murmurs frequently are heard a few days after the illness begins. They often disappear altogether. Sometimes the anæmia has caused them. But at others I believe that the valvular nodules (endocarditis) have subsided, and the murmur caused by them also. It is an error to say that the right side of the heart is not affected in children. Post-mortem examinations not unfrequently reveal beadings on the tricuspid valve. The pericardium and endocardium may inflame without any indications of rheumatism in the limbs. This fact is included in the statement made on p. 120.

The severity of children's rheumatism may be measured by the number of nodules. It does not follow that all the rheumatism will not disappear when it is of severe degree. Much of children's rheumatism is smouldering and lingering in its course, and then, I think, the persistence of the nodules. is, to a certain extent, a warning and a guide of this. But the rheumatic cause may pass away and still the nodules remain and the heart mischief increase. Here must we suppose that some other irritant than rheumatism is keeping up the nodules ?

The diagnosis of rheumatism in children may often be made when there is no true rheumatism. Infantile palsy may be called rheumatism owing to the pains and tenderness and fever which may announce its onset; even swelling about the ankle may occur and simulate rheumatism. The rapid development of paralysis with loss of reflex action, however, soon clears the case. The Germans have described as rheumatic a slight swelling about the back of the wrist and front of the ankle that may attend the onset of painful tonic spasm of the hands and feet (tetany). The immobility (due to pain) (pseudo-paralysis) of limbs, often symmetrical, that is seen in some cases of congenital syphilis about the elbows and wrists, or knees and ankles, may be mistaken for rheu