308 Robinson and Walk den.—A Critical Study of Crown Gall. 
gall-cells arching over the greatly modified end of the stem; and parenchy¬ 
matous cells are developed externally to this active tumour-tissue. As the 
gall increases in size, the outermost layers of cells of this cortical region are 
continually torn apart by the expansion, the layer of dead cells on the 
exterior becoming very marked in the older galls (PL V, Fig. 3,c). There 
is, however, no production of cork on the exterior of the galls, and, as will 
be shown below, vast numbers of Bacterium tumefaciens are found upon the 
layers of dead cells on the exterior. The increase in size of the gall, in the 
later stages at least, seems to result from the presence of the Causal bacteria 
in preponderant numbers in this position. 
Apart from the direct effect of the presence of the organisms in leading 
to the production of tumour-tissue there is also, as Smith’s figures show, an 
extension backwards of the influence some distance from the actual gall. 
This effect is often manifested in a widening of the vascular ring by more 
than the ordinary cambial activity in the portion of the stem below the gall 
as in PI. V, Fig. 2 . A similar effect frequently may be seen in the petiole 
or midrib beneath a gall borne on a leaf. While dealing with our inocula¬ 
tions of shoots of Chrysanthemum, it may be mentioned that on several 
occasions we have inoculated axillary shoots growing from positions very 
near to galls produced by earlier inoculations. We have invariably obtained 
galls on such shoots, showing that there is no immunity acquired by plants 
against B. tumefaciens following an earlier infection by • the crown-gall 
organism, as has been suggested (PI. V, Fig. i,C). 
Galls of the general type described above were repeatedly produced 
by inoculating the cut surfaces of the shoots of healthy plants, and these 
galls resembled, in every particular, the naturally occurring galls on plants 
grown in the nurseries. A general distinction must be drawn between the 
galls described above, whether naturally occurring or experimentally pro¬ 
duced, and the secondary galls and tumour-strands to be described below. 
The former type of gall has a rough exterior, and develops quite differently 
from the latter type, which has a smooth exterior. The secondary galls 
and tumour-strands, described by Smith and Peklo and produced by us, are 
always the result of artificial inoculations and have only been met with in 
experiment. 
The evidence which we have obtained regarding the position of the 
bacteria in the large galls with rough exterior on the shoots of Chrysan¬ 
themum frutesceiis will now be dealt with. 
Distribution of B. tumefaciens in the Rough Galls on 
Che ysan the mum frutescens. 
The difficulties experienced by Walkden (35) and others in isolating 
B. tumefaciens from the interior of naturally occurring galls, after sterilizing 
