volved. The lymphatic glands serve as filters and contribute much toward retarding general dissemination. General infection is likely to occur at an early date if the lymphatic glands do not participate in the regional extension of the tumor. The malignant cell after it has become detached from the mother-tumor retains all the qualities inherited from it at birth, and consequently produces the same kind of tissue, whether it remain in the vicinity of the tumor in the same region. or whether it is transported to the most distant organs. The secondary tumors resemble the primary tumor histologically and clinically (Fig. 27). Regional dissemination of a sarcoma takes place by a continuous growth of the tumor, usually in the direction of fascia, blood-vessels, or nerve-sheaths: it is a local infection on a large scale. Occasionally

a sarcoma gives rise to regional infection in the same manner and through the same channel as carcinoma.

Another method of regional infection takes place by the diffusion of particles of tumortissue or free tumor-cells over serous surfaces-in the abdominal cavity by the peristaltic movements of the intestines and the stomach, and in the pleural cavity by the movements of the lung during respiration. This manner of regional infection is witnessed most frequently in sarcoma. of the peritoneum and the pleura, and in carcinoma of any of the abdominal organs or of the lung after the tumor has reached the serous cavity.

FIG. 27.-Secondary sarcoma of lymphatic vessels of omentum in the course of a medullary sarcoma (after Lücke).

General Infection.-General infection during the growth of a malignant tumor is called metastasis-that is, the reappearance of the same disease in a distant organ. When this stage is initiated the tumor is no longer local: the disease has become general. No modern pathologist regards—as was formerly and quite recently done-a primary malignant tumor as a local manifestation of a general disease or dyscrasia. A careful study of the pathology and histology of malignant tumors, as well as the results of accurate clinical observation, has demonstrated that malignant tumors are primarily purely local affections, amenable to successful surgical treatment, and that they become general only by the dissemination of tumor-cells through the systemic circulation. Metastasis may occur in one of three ways: 1. Tumor

cells reach the venous circulation directly by their entrance from the primary tumor or the regional glandular tumors into a vein; 2. By progressive extension of the disease through the lymphatic channels until the last filter, the last lymphatic gland, is passed, when the tumor-cells reach the general circulation; 3. By the passage of tumorcells through the chain of lymphatic glands into the pulmonary or systemic circulation without implicating the lymphatic glands. It is strange that the tumor-emboli are not more constantly arrested in the finer branches of the pulmonary artery. The result of post-mortem examinations of persons who died of malignant tumors would tend to show that such emboli readily pass the pulmonary filter, and may become arrested in any of the more distant vascular organs. The exemption of non-vascular tissues from metastatic carcinoma is one of the many proofs that malignant tumors are generalized by cellular elements, and not through the agency of a virus or of microbes. Metastasis always takes place through the arteries. Usually the emboli are small (Fig. 28). In some cases perhaps a single cell becomes implanted upon the wall of an arteriole, and later a thrombus is formed by tissue-proliferation from this cell. In other instances a vessel of considerable size is obstructed by a malignant thrombus.

Metastatic tumors frequently

FIG. 28.-Embolism of the right pulmonary artery from a pigmented sarcoma of the thigh (after Lücke).

extend in the direction of a blood-vessel of considerable size, the multiple tumors with the blood-vessels and its branches presenting the appearance of a bunch of grapes (Fig. 29).

The number of emboli varies greatly, from a single metastatic tumor to thousands of nodules. In some very malignant forms of carcinoma and sarcoma the nodules are so numerous that the appearance of the

internal organs resembles very closely that of miliary tuberculosis. Metastasis occasionally takes place in the aged who have been the subjects of latent carcinoma for years. In some instances the patients were not aware of the existence of the primary tumor until the presence of a large and destructive metastatic tumor gave occasion to consult a physician. Sarcoma gives rise to general infection more constantly and at an earlier date than does carcinoma. Small-celled sarcoma is more frequently followed by early and diffuse general dissemination than are large-celled tumors.

Frequency of Recurrence after Extirpation.-It has been shown. that the recurrence of a benign tumor is always local, and is invariably the result of incomplete removal of the tumor. The recurrence after the removal of a malignant tumor is either local or metastatic-in the former instance caused by incomplete removal of the primary tumor, and in the latter instance a sad reminder that the operation was not performed early enough to protect the patient against general infection.

The most competent surgeons are willing to admit that so far the best results of operations for malignant disease have not yielded more than about 15 to 25 per cent. of permanent recoveries. If we recollect how a malignant tumor reaches out in all directions into tissue which to the naked eye presents every indication of being normal, we can readily understand why local relapse should follow so frequently even after what seemed a thorough operation. Again, every surgeon has reason to regret that in most cases he is called upon to operate for malignant tumors after the disease has advanced beyond the limits of a successful radical operation. In some instances no local recurrence takes place, but the operation was performed too late, and the patient succumbs sooner or later to metastatic carcinoma or sarcoma. In such cases general infection had taken place when the operation was performed. A local recurrence may take place from three to seven years after the operation for carcinoma of the breast, as happened in a number of the writer's cases, and it may be postponed, according to Billroth, twenty years from the time of operation in cases of sarcoma. Sarcoma usually returns in the scar; carcinoma, either in the scar or in the adjoining lymphatic glands.

Intrinsic Tendency of the Tumor to Destroy Life.-If we reflect upon the fact that with the best efforts of the surgeon only 15, and at best only 25, per cent. of all persons suffering from malignant tumors escape a painful and lingering death from their immediate and remote effects, we must admit that the intrinsic tendency of a malignant tumor is to destroy life. The average duration of life of all persons suffering from malignant tumors of all kinds and of all parts and organs of the body, without surgical intervention, is about three years. It is a source of satisfaction to the surgeon to know that life is prolonged by radical attempts to remove malignant tumors, and that in a fair proportion of cases the disease never returns. Life is destroyed by regional or general dissemination involving important organs, by the primary tumor interfering with the function of an important organ, by hemorrhage, or, lastly, by a progressive chronic sepsis or septico-pyæmia caused by an open ulcerating carcinoma or sarcoma. The so-called "cachexia" which appears so constantly some time before the fatal termination is the result of impaired nutrition and of the introduction into the circulation from the tumor of toxic substances.

VIII. TRANSFORMATION OF BENIGN TUMORS AND POST-NATAL EMBRYONIC TISSUE INTO

MALIGNANT TUMORS.

THE possibility of the transformation of a benign into a malignant tumor has been asserted by a few and denied by most of the older writers on surgical pathology. The subject is of great interest to the pathologist, and of equal practical importance to the surgeon. Accumulated clinical observations, since the diagnosis of tumors has been made more accurate by increased knowledge of their pathology and by a more frequent resort to the use of the microscope in the examination of tissue removed for diagnostic purposes and of fresh specimens after operation, have brought more convincing proof of the possibility of such an occurrence. As the result of his own observations the writer is convinced not only that such a transformation is possible, but also that it takes place much more frequently than has heretofore been supposed. The writer is equally certain that malignant tumors not infrequently originate from embryonic tissue of post-natal origin.

Transformation of Benign into Malignant Tumors.-The transformation of a benign into a malignant tumor implies a change in the histological structure of the cells of the benign tumor as well as a change in its environments. We have seen that the cells of which benign tumors are composed resemble the normal cells of the part or organ in which the tumor is located. In a myofibroma of the uterus the cells resemble the connective tissue and the unstriped muscle-cells in the uterine wall in which the tumor is located. The epithelial cells in an adenoma of the breast cannot be distinguished from the epithelium of the acini and tubules of the mammary gland. The transformation depends, therefore, upon influences which accomplish such a change from mature into embryonic cells. At the same time, and probably from the same causes, the physiological resistance of the adjoining tissues is diminished.

The liability of benign tumors to become malignant is of interest not only as a subject of pathological study, but also in relation to an opinion which is often made an argument for operations-namely, that if a tumor of any kind is left to pursue its own course, it is not unlikely to become malignant. This belief, which is entertained by the general