It is known that an experimental tumor on transplantation invariably reproduces the same kind of tumor as to cell type. For instance, the Rous sarcoma always produces a fibrosarcoma whether it is injected into muscle, bone, connective tissue or an internal organ, such as the liver. When, therefore, an opportunity came to work with an endothelioma, I seized the chance to observe its behavior when injected into the bonemarrow of chickens. The purpose of such a procedure was several fold. First, it was desired to see to what extent it might produce a tumor resembling the much disputed tumor described by Ewing as an endothelioma of bone. Second, it seemed important to watch the development of tumor of the bone-marrow from its inception, a thing impossible to do in man, since in the latter it usually has advanced considerably before it is discovered. Third, the reaction of the bone itself