300 Robinson and Walk den.—A Critical Study of Crown Gall. 
been carried out from this point of view. He regards most of his results as 
supporting the comparisons which he has drawn. It will be necessary to 
refer to these in some detail. 
Bacterium tumefaciens is regarded by Smith as a feeble wound-parasite 
present in small numbers within the cells of the plants attacked, although 
not demonstrable by direct methods. 1 The organism produces substances 
which stimulate the tissues to give rise to larger or smaller overgrowths 
according to the nature and age of the tissues affected. Along with this 
there is a stunting of growth and ultimately a slow killing of the plant 
attacked. The main comparisons with cancer and other malignant tumours 
have been drawn from the experimental production of secondary tumours 
and of ‘ teratomas ’, or crown galls bearing leafy shoots having no relation 
to preformed buds. The secondary tumours are described as ‘ growths 
from tumour-strands bedded deep in normal tissues derived by growth (cell- 
division), in the form of a continuous chain of cells, from the primary 
tumour’. 2 Smith’s photographs usually show secondary tumours at some 
considerable distance from the point of inoculation, and he infers that the 
proliferating tumour-strands have grown intrusively from the point of 
inoculation to the positions where secondary tumours appear. 
In further support of the supposed similarity with malignant tumours 
he holds that ‘secondary tumours reproduce the structure of the tissues in 
which the primary tumour has developed even when they appear in other 
organs ; thus if the primary growth is in the stem and the secondary growth 
is in the leaf, the attacked part of the leaf will be converted into a pseudo¬ 
stem ’. 3 This, and the origin and nature of the secondary galls and 
tumour-strands, obviously require further investigation before E. F. Smith’s 
interpretations of these aspects of crown gall can be accepted. 
The need for re-investigation of these and other comparative aspects of 
the tumour formation in crown gall was discussed in a paper by Professor 
W. H. Lang ( 12 ) at the meeting of the British Medical Association in Glasgow 
in July, 1922. We are in complete agreement with Professor Lang’s 
general formulation of these problems, which, as he states, was made in the 
knowledge and light of this work then in progress. We gratefully acknow¬ 
ledge our indebtedness to Professor Lang for his continual interest and 
helpful criticism during the course of the work. 
The work of some other investigators on crown gall may be briefly 
referred to before the scope of the present paper is indicated. Jensen ( 9 , 
10) has made observations on crown galls produced on beet by inoculation 
with B. tumefaciens and has compared these galls with malignant tumours. 
1 In earlier papers Smith figured the bacteria in small numbers in the cytoplasm of the tumour 
cells. More recently he has stated that the bodies he previously demonstrated by staining with gold 
chloride are not bacteria. 
2 Smith ( 31 ), p. 417. 
3 loc. cit., p. 417. 
