In the rare cases in which the skin of the greater part of the body is affected, the nodules rarely appear below the level of the deltoid or the middle third of the thigh, the patient dying before the spread can reach the distal portions of the limbs.

Handley argues against the embolic origin of the metastases in the bones because of the rarity of these in the bones of the distal parts of the limbs, because of the fact that secondary cancer of the femur nearly always commences in the upper third of the shaft, which harmonises with the intimate connection of the deep fascia with the periosteum over the great trochanter, thus favouring invasion of the bone marrow when permeation has spread thus far.  He claims support for the permeation theory from the fact that the humerus is rarely involved below the insertion of the deltoid, and that spontaneous fracture of the femur is three times more common on the side on which the breast cancer is situated.

The tumour tissue may undergo necrosis, and when the overlying skin or mucous membrane gives way an ulcer is formed.  The margins of a cancerous ulcer (Fig. 57) are made up of tumour tissue which has not broken down.  Usually they are irregular, nodularly thickened or indurated; sometimes they are raised and crater-like.  The floor of the ulcer is smooth and glazed, or occupied by necrosed tissue, and the discharge is watery and blood-stained, and as a result of putrefactive changes may become offensive.  Haemorrhage is rarely a prominent feature, but discharge of blood may constitute a symptom of considerable diagnostic importance in cancer of internal organs such as the rectum, the bladder, or the uterus.

[Illustration:  FIG. 57.—­Carcinoma of Breast with Cancerous Ulcer.]

The Contagiousness of Cancer.—­A limited number of cases are on record in which a cancer appears to have been transferred by contact, as from the lower to the upper lip, from one labium majus to the other, from the tongue to the cheek, and from one vocal cord to the other; these being all examples of cancer involving surfaces which are constantly or frequently in contact.  The transference of cancer from one human being to another, whether by accident, as in the case of a surgeon wounding his finger while operating for cancer, or by the deliberate introduction of a portion of cancerous tumour into the tissues, has never been known to occur.  It is by no means infrequent, however, that when recurrence takes place after an operation for the removal of cancer, the recurrent nodules make their appearance in the main scar or in the scars of stitches in its neighbourhood.  In the lower animals the grafting of cancer only succeeds in animals of the same species; for example, a cancer taken from a mouse will not grow in the tissues of a rat, but only in a mouse of the same variety as that from which the graft was taken.