318 Robinson and Walkden.—A Critical Study of Crown Gall. 
ourselves and others in isolating Bacterium tumefaciens from the interior of 
sterilized galls. This distribution of the bacteria also throws light on the 
form of these galls and their general resemblance to ordinary growths of 
callus such as occur on the twigs of trees. The organisms present over the 
hemispherical outer surface of the gall provide the continued stimulus which 
accounts for the fairly uniform meristematic activity in a region of the gall 
comparatively close to the surface. This continuity of stimulus explains 
the differences in the degree of reaction obtained in the galls and in ordinary 
growths of callus. 
The presence of B. tumefaciens on the exterior of the galls also 
explains the extreme ease with which the soil in which diseased plants are 
growing becomes highly infectious, since the organisms must be washed into 
the soil whenever water falls on the plants. 
The bacteria often associated as zoogloea-like strands have, on the other 
hand, been traced for shorter or longer distances from the point of inocula¬ 
tion. The course of this bacterial extension is by way of intercellular 
spaces and protoxylem. The recognition that the formation of tumour- 
strands and secondary galls follows along the track of invading bacteria 
leads to a reconsideration of some of the more specific comparisons that 
have been made between crown gall and malignant tumours. 
Smith has held that the intrusive growth of tumour-tissue, which he 
has described for tumour-strands, is directly comparable to the migratory 
growth of the diseased tissues in certain forms of malignant disease. We 
have been unable to find any real intrusive growth in crown gall, but the 
demonstration of the bacteria advancing in the intercellular spaces and 
protoxylem fully explains the development of tumour-tissue in the neigh¬ 
bourhood of their path, and also the existence of strands of such tissue 
connecting secondary with primary galls and the threading of successive 
secondary galls in a linear series. In our experience the careful examina¬ 
tion, by staining methods, of serial microtome sections usually reveals the 
relatively close proximity of the causal bacteria to the proliferating tissues. 
There is, therefore, no necessity to assume the stimulation of cells at any 
considerable distance from the active bacteria. It is unnecessary also to 
adopt Jensens hypothesis that stimulated cells removed from the bacterial 
influence behave in a parasitic manner, as, according to him, do the cancer 
cells in the animal disease. 
It was mentioned in the introduction to this paper that Smith has 
utilized the appearance of a radial stem-like structure in secondary galls on 
the leaves of Chrysanthemum for pressing even more closely a further com¬ 
parison with cancer. He suggests that the stem-like structure occurs in the 
leaf because the tissues originally inoculated were those of the stem, and the 
comparison is directly made with those secondary malignant tumours which 
reproduce in distant organs the tissues of the organ originally diseased. 
