The first, most important, modification of Virchow's views, which has led to a more rational appreciation of the relation of the various tumors, especially of the epithelial group, to each other, arose in consequence of the investigations of Thiersch and others with regard to the origin of certain cancers. This observer[83] claimed that the epithelioid element of cutaneous cancers arose in all instances from pre-existing epithelium, either of the rete mucosum or cutaneous glands. Similar views were suggested, with various degrees of precision, by other authors concerning certain cancerous tumors elsewhere, but were first applied to all cancers with a more exact formulation by Waldeyer,[84] to whom the prevailing views with regard to the histogenesis of morbid growths are due. According to him, the essential (epithelioid) element of all primitive cancers arises from pre-existing epithelium; consequently, no cancer-cell can arise except in organs where epithelium is normally present. [Footnote 83: _Der Epithelial Krebs, namentlich der Haut, etc._, 1865.] [Footnote 84: _Virchow's Archiv_, 1867, xli. 470; 1872, lv. 67; _Volkmann's Sammlung klinischer Vorträge_, 1871, xxxiii.] This comprehensive statement was rendered possible by the embryological researches of Remak at the outset, and afterward by those of His and Waldeyer. Remak showed that after differentiation of the cells of the ovum into the several germinal layers, those from one layer could not serve to originate the cells belonging to another layer. The development of normal tissues takes place within the limits defined by this differentiation. Epithelium thus is not derived from connective tissue, nerves, or muscles, nor was the reverse known to occur. To His is due the exact appreciation of the superficial cells of serous membranes, which had been previously called epithelium, and had thus been confounded with the epithelial cells of mucous or cutaneous membranes and of secretory glands. He showed that these cells had a wholly different origin from epithelium, and were simply scale-like cells of fibrous tissue, to which he applied the name endothelium. The latter is now used as the term for the thin, squamous cells of fibrous tissue, whether they are found lining the walls of the great serous cavities or the smaller lymph-spaces, the endocardium, or the inner coat of blood-vessels and lymphatics. The importance of this distinction is obvious when the occurrence of tumors, called cancers, is observed in parts which contain no epithelium. Aside from the vagueness of the term cancer, as applied clinically, tumors are sometimes met with, even in parts where epithelium normally does not exist, whose structure resembles more or less closely that of cancer as usually recognized. Such tumors are to be regarded as of an endothelial rather than epithelial character, and as such their histogenesis falls under the general laws of the development of tissues.