Sarcoma, on the other hand, whose clinical features correspond so closely with those of cancer, simulates, as shown by Virchow, the connective tissues. It is composed of cells and intercellular substance, both of which may be as varied as are those of the connective tissues. The shape of the cells is as diverse and their contents as various, while their possibilities of degeneration are alike. The cells of the sarcoma are not simply cemented together, as are epithelial cells, but they are separated from each other by an intercellular substance, which corresponds in its appearance and chemical properties with that of mucous, fibrous, cartilaginous, or osseous tissue. The structure of the sarcoma differs from that of these tissues in presenting a predominance of cells over intercellular substance, while the reverse is the characteristic of most varieties of connective tissue. In this predominant cell-formation lies its absence of type, {119} whereas the atypical character of the cancerous growth is manifested rather by the irregular grouping of the cellular masses than by an abundance of cells. As the original cancer is considered as possible only in parts where epithelium is a normal constituent, so the primitive sarcoma is possible only in parts where connective tissue is present. The apparent great frequency of sarcoma in recent times is thus obviously explained. With an agreement as to its histological characteristics, its possible place of origin is any of the connective tissues of the body, and their presence is universal. In the manner of its growth, its recurrence, and generalization it is subject to the same laws which determine similar events in the history of cancer. Its degenerations are often the same, and its symptoms are due to the action of like causes. The importance of distinguishing between these atypical tumors is real, in that it is only through the association of causes, symptoms, and results with defined and constant characteristics that a practical knowledge of tumors is to arise. The time-honored distinction between malignant or semi-malignant and benignant growths is always to be sought for, and can only be fully possessed when the natural history of the new formations is known. With an exact appreciation of the structure of a tumor it becomes possible to study its special pathology. From a knowledge of the latter are to be derived those features of importance in determining the relation of morbid growths to other deviations from normal and physiological processes. An immediately practical benefit arises from the Thiersch-Waldeyer modification of Virchow's theory of the origin of tumors, in that it permits with greater ease a more accurate clinical diagnosis. Lücke[87] has been prominent in calling attention to the suggestions thus presented. [Footnote 87: _Volkmann's Sammlung klinischer Vorträge_, 1876, xcvii.]