36 
ALBAN STEWART 
through the action of the fungus, while those on the lower part and sides 
have remained normal in structure. In this instance only a sector of the 
stem has been involved in gall-formation. The xylem on the infected side 
has been greatly increased both radially and tangentially, while on the 
other sides it has only its normal growth. Both the leaf traces and the 
stem bundles can be recognized on the infected side, as the changes brought 
about by the fungus in this particular instance are less pronounced than 
in some other specimens examined. Furthermore, the inner extremities 
of all the infected bundles have their normal structure, showing that this 
particular part became infected some time after secondary thickening 
started. Broad intrafascicular rays have been formed, cutting up the 
bundles into narrow radial segments of xylem. Masses of parenchyma have 
formed in the xylem where normally there is no parenchyma. There has 
also, been a great reduction in the number of vessels per unit of area, and 
the size of the vessel has decreased. 
The amount of parenchyma, both ray and wood, and the number and 
size of the vessels vary greatly in different specimens. The specimen shown 
in figure 8 is about at a minimum in this respect. In extreme cases the 
xylem portions of the leaf traces are often suppressed or nearly so, resulting 
in broad ray-like masses of parenchyma. Where leaf traces fail to appear 
there are deep notches in the xylem ring, their position being shown ex- 
ternally by longitudinal furrows in the surface of the gall. 
The formation of depressed xylem segments is not unknown in normal 
stems of the Compositae. According to Solereder (7), Schenck has ob- 
served a furrowed xylem mass in an undetermined member of the order 
recalling the structure of the Bignoniaceae. "The furrows in this xylem 
mass are due to the reduced activity on the part of certain longitudinal 
strips of the cambium which leads to a smaller production of wood and a 
proportionally greater development of phloem externally." 
It is often the case that the individual bundles disappear entirely and 
in their place there are masses of parenchyma which show but little dis- 
tinction between wood and bark. Occasional vessels appear in these 
masses, and irregularly running strands of cambiform cells. 
In this connection it might be well to mention certain peculiar strands 
of tissue which occur under such conditions. Towards the center of many 
of the galls, in the region of the pith or in close proximity to the protoxylem 
portions of the bundles, these strands sometimes appear. They consist of 
whorl-like arrangements of cambiform cells (fig. 5) which enclose sporanges 
of the fungus, shown by the dark circles in the figure. Structurally they 
recall strikingly the tumor-strands described and figured by Smith (6) in 
stems infected with the crown-gall organism, Pseudomonas tumefaciens. 
Short tracheids sometimes accompany these strands, a condition similar to 
that described by Smith in some of his strands. They usually extend in a 
vertical direction but not always for the longitudinal section of one of these 
