Fifteen years ago some sections were cut for me from a tumor which I had produced in the cortex of a young tobacco plant by means of a single needle-prick introducing the Paris daisy strain of Bacterium tumefaciens Sm. and T. This was a young tumor, that is, one only three weeks old, and the sections were cut and stained in series. In the rush of other work these sections were overlooked and as they had been stained with a fugitive stain they faded so that when I came to examine them some years later there was a shoal of pale tissue surrounding the tumor which I could not in the least interpret. The cover-slips were removed by soaking in xylol and the sections restained, whereupon I was face to face with a new phenomenon (new to me). I had before me what cancer specialists have called conversion of normal cells into tumor-cells by apposition , that is by contact of the diseased with the normal, the “shoal” proving to be a 0.5 mm. wide layer of cells intermediate in character between the normal cortex-cells and the tumor-cells. By intermediate I mean midway in size and affinity for tumor-stains, and showing various transition stages including here and there a surrounded unchanged cortex-cell. (See this Journal, vol. i, 1916, no. 2, fig. 78). I was much impressed by the phenomenon and in searching through other boxes of sections I found evidence of it in various tumors due to Bacterium tumefaciens including some produced in the cortex of the Paris daisy. The phenomenon is also shown on plate 29 of Bulletin 213 (l.c.), published in 1911, but the evidence in this figure had not then specially attracted my attention and is not mentioned in the text. That there is here conversion of normal cells by apposition rather than invasion of normal tissue by cells growing out of the originally infected cells, or than simply an irritation-response of the host-cells which never passes over into tumor-tissue, there can be no question whatever for the cells have not changed places but the change has occurred in situ by the conversion of large cells wholly normal into congeries of small cells having all the characteristics of tumor-cells and visibly surrounded in many cases by the stretched wall of the original cell. An irritation-response that does not pass over into tumor tissue proper also occurs in the vicinity of many crown galls, viz., an overgrowth of wood and bark, due to the stimulus of an extra amount of food moving in the direction of the tumor in greater quantity than it can use, so that some part of this excess either never actually reaches the tumor or oozes backward from it into the adjacent tissues, which are thereby incited to excessive growth; but this is something quite different from the response we are here dealing with because the cells of the hyperplasia in the one case have normal arrangement, normal staining properties, and function more or less normally whereas the cells in the other (the appositional growth) are smaller more or less disoriented and stain and behave like tumor-cells. For figures showing thickening of the wood entirely outside of the tumor but influenced by it see Bulletin 255 (l.c.), plates 25, 62A and 63, or An introduction to bacterial diseases of plants, fig. 319, subs 5 and 6.2 These proofs should convince any one that the phenomena here described are not the same although both kinds of growth are brought about by the presence of the tumor.