Robinson and Walk den.—A Critical Study of Crown Gall. 313 
in most cases, but not in all, where separated galls appeared the phyllotaxy 
diagram indicated their connexion with the original needle-track. 
In a number of other cases, however, both in longitudinal and horizontal 
inoculations of apices, smooth galls of a different appearance were obtained 
at a distance from the larger rough galls that were always directly associated 
with actual wounds. Such galls, which are similar to many of Smith's 
secondary tumours, are seen in PI. V, Fig. 11 {s.g.), and PI. VI, Figs. 12 {s.g\ 
1, 2, and 3) and 13 {s.g.). Galls of this type only will be regarded as true 
secondaries. We have never found B. tumefaciens on the exterior of such 
smooth galls either by direct observation or by cultural methods, and, like 
Smith, we have had considerable difficulty in isolating the organism from the 
interior, but in this we were successful in two cases. We have, however, in such 
secondary galls directly demonstrated, by staining, the position of the bacteria 
in the protoxylem vessels and in the intercellular spaces adjoining these, 
and further we have found continuity of the organisms from such galls to 
the primary gall, which arises where the inoculation wound is made. The 
internal structure of such secondary galls may now be dealt with, and the 
relations of the infecting organisms to them will be made clearer after 
secondary galls on Nicotiana have been described. 
PI. VI, Fig. 12, shows a portion of a shoot which was inoculated by a single 
longitudinal needle-prick at the apex ; the rough elongated gall on the side 
of the shoot is a primary gall (using Smith’s terminology), and extending 
along the midrib of the leaf are three c secondary ’ galls in linear series. 
PI. VI, Fig. 14, is a longitudinal section of the middle secondary gall on the 
leaf shown in PI. VI, Fig. 12, s.g. 2. The swelling is originating from within 
by the subdivision of cells of the parenchyma in a manner exactly similar to 
that seen in the galls on the ends of shoots dealt with in the first section of 
this paper. There is, in addition, a more striking subdivision and proliferation 
of tissues immediately adjoining the protoxylem {px.) of the vascular bundle, 
resulting in the slight general displacement of the tissues of the parenchyma, 
but there is no evidence of any considerable intrusive growth of tumour- 
tissue outwards. The elements of the protoxylem {px.) stain deeply in 
the manner characteristic when organisms have been found in them, and the 
appearances generally are consistent with an influence diffusing out from the 
protoxylem. PL VI, Fig. 15, shows a transverse section of the much older 
elongated secondary gall {sg.) seen in the deformed leaf in PL VI, P'ig. 13. 
The structure is very similar to that frequently figured by Smith for secondary 
galls on the leaf of Chrysanthemum. It shows the modification of one of 
the vascular bundles into a radial structure, which was regarded by Smith 
as evidence of the stem origin of this by the intrusive growth of tumour- 
tissue from the infected stem out to the leaf in which the secondary gall has 
arisen. As will be seen below, particularly clearly for Nicotiana , we have 
shown that such galls arise around definite centres of bacteria in the interior 
