306 Robinson and Wa/kden.—A Critical Study of Crown Gall. 
shows the maximum amount of callus we have observed. It may be 
mentioned that a greater callus development occurs in cuttings of 
Chrysanthemum frutescens prepared through the nodal region than through 
the internodal region. In all our experiments the inoculations were made 
at an internode. Roots do not develop behind the upper cut surfaces of 
aerial shoots except as rudiments in the late stages of gall development on 
the inoculated shoots. In such cases the root rudiments arise on the 
swollen region of the otherwise normal stem immediately below the gall. 
The region referred to is seen in PI. V, Fig. 2, although roots have not yet 
arisen. 
When the cut surfaces, either of the upper ends of shoots in air or of 
the lower ends of cuttings in the soil, are inoculated galls arise and grow 
rapidly. The growth is usually somewhat more rapid in the case of gads 
in the air than of those in the soil. The general appearance of such arti¬ 
ficially produced galls on aerial shoots is shown in PI. V, Fig. i. 
The development of such galls from the time of infection will now be 
dealt with and the changes observed contrasted with those seen in the 
controls. 
Three days after wounding, the shoot (whether control or inoculated) 
shows marked alterations in the vicinity of the cut surface. The cells 
actually cut through are dead, and the walls of these, as well as the walls 
of all the tissues to a depth of two or three cells below the surface, are 
altered in properties and substance. The walls are much more readily 
stained by Sudan III, Scharlach R., methylene blue, and haematoxylin 
than are those of the normal corresponding tissues or than the walls of 
similar cells immediately after cutting across the shoot. These altered 
cell-walls, unlike those of the normal parenchyma, are not swollen and dis¬ 
solved by strong sulphuric acid, but are merely stained brown with this 
reagent. The changes are doubtless necrosis changes similar to those 
previously described by investigators of wound reactions. The changes 
referred to are also shown by the walls of the sieve-tubes, pericycle fibres, 
and vessels, in these cases extending to a greater depth than in the 
parenchyma tissues. These cell-wall changes occur whether the wounded 
surface is in contact with soil or with air, irrespective of polarity, and of 
whether the surface is inoculated or not. In all cases, however, the maxi 
mum effect of the alterations described is attained in three days. 
For the purposes of detailed description of the development of the gall, 
attention will now be confined to the galls arising on the upper ends of 
inoculated shoots in air. In the inoculated shoots it is evident, even after 
three days, that the bacteria have entered the open ends of the vessels, of 
the sieve-tubes, of the young pericycle fibres, and to some extent also the 
intercellular spaces of the cortex. The walls of the cells, with which the 
bacteria come into contact, show similar changes to those described above. 
