and remains discrete and ununited while other normally distributed cells proceed to form the normal blood vascular system. The origin, therefore, of hemangiomas can be traced to the failure of these cell groups to establish contact with the developing network of blood vessels, in consequence of which and in their dis-jointed state, the early angioma does not evolute usually beyond the stage of a mesh of minute, irregular, blind channels quite removed from the main, general circulation. The anlage of the angioma is a cluster of solid endothelial cords and islands, which later acquire lumens and form stunted blind capillaries, which in turn later intercommunicate to complete the well-developed hemangioma.

Ribbert strongly opposed the view, prevalent in his time, that the tumors arose from pre-existing blood vessels which had functioned as such. Ribbert found by interstitial injection that the angiomatous vessels had few or no lateral anastomoses and that the injection mass traversed the afferent artery and the efferent vein, but not in the peri-tumoral tissue. He believed that the heman-giomatous varicosities were due to elongation of the vessels comprising the tumor within a more or less closed territory or capsule and without the involvement of the surrounding vessels. Ribbert’s theory that circumscribed hemangiomas, with the exception of the cirsoid or racemose type, possessed only one afferent and one efferent blood vessel and were without anastomotic connections to surround-