Robinson and Walkden.—A Critical Study of Crown Gall. 31 7 
Secondary galls and tumour-strands essentially similar in all respects 
to those we have described and figured for Chrysanthemum fmUescens have, 
however, also been obtained on the leaves of Nicotiana affinis. Numerous 
needle-prick inoculations of the apices of young flowering shoots of this 
species of tobacco were made and results similar to those described by Smith 
have been obtained ; but wherever the galls have a rough exterior the 
organisms are present in abundance on this surface, and when completely 
smooth galls occur they, like the secondary galls we have described above, 
have the organisms present in the intercellular spaces and vessels within. 
For such inoculations, both the enormous stretching by growth of the 
inoculated meristem and the actual migration and multiplication of B. tume - 
faciens within the vessels and cavities of the stem must be taken into account 
in explaining the resulting secondary galls and tumour-strands. In the 
tobacco, as Smith has shown, these frequently burst out to the exterior 
either from the pith or cortex, and the rough surface, which they then 
acquire, is richly populated with B. tumefacieiis. 
We have obtained no evidence of the migration of strands of tumour- 
tissue to any distance at all comparable with that postulated by Smith, but 
we have shown earlier in this paper that the effect of the bacterial stimulus 
proceeding from definite centres invariably results in the subdivision and 
enlargement of cells. There is no intrusive growth of these dividing cells 
in Smith’s sense, although there is often a slight displacement of the divid¬ 
ing cells due to the inequality of pressures developed by the abnormal cell- 
divisions and enlargements. PI, VI, Fig. 23, d , from the edge of one of 
the secondary tumours seen in PL VI, Figs. 16 and 17, shows one of the most 
extreme cases of cell-displacement which we observed. Even in this case, 
however, at e it is obvious that the tumour-cells are arising by subdivision 
of ordinary cortical cells. On the other hand, as we have already pointed 
out, PL VI, Figs. 19, 20, 21, make it clear that the tumour-strands are in no 
real sense intrusive growths of tissue, but the result of intruding masses 
of bacteria forming centres of disturbance. 
Discussion of Results. 
The results of our work do not require extended discussion, but the 
more general bearings of the new facts we have established must be briefly 
considered. The study of the mode of occurrence and of the distribution of 
the bacteria in the galls is a necessary preliminary to the understanding of 
the manner of growth and development of the latter. Much of the work 
described in this paper is therefore concerned with obtaining exact know¬ 
ledge regarding the position of the bacteria. 
The demonstration of the bacteria, in large quantities, on the exterior 
of naturally occurring galls, and also upon those produced in the immediate 
neighbourhood of inoculation wounds, has explained the earlier difficulties of 
