PATHOLOGIC CONDITIONS IN AMBROSIA TRIFIDA 
37 
strands (fig. 4) also shows cross sections of two other strands. It should 
be noticed that the strand shown in figure 4 encloses relatively large num- 
bers of the sporanges of the fungus, and that where there are large numbers 
of sporanges the strand is broader. The line of separation between the 
strand and the surrounding tissue is sharp, as can be seen in the figure. 
Near the lower right side of figure 4, a single sporange is shown lying just 
outside the strand, partly surrounded by cambiform cells, similar to those 
which make up the strand. 
The strands are evidently formed in place near the time of infection and 
are not a product of later growth which has been pushed in from some other 
source. This is shown by the fact that in the parts of the gall where they 
occur all the tissue outside them is very abnormal in structure. That 
infection takes place while the stems are still very young is evident, for in 
every gall examined some portion or portions of it showed tissue abnormali- 
ties in or nearly to the pith, as well as sporanges of the fungus. Further- 
more, there is no indication of crushed cells, or of cells which have been 
partially absorbed along the course of the strands, such as would probably 
be the case if the strands represented an invasion from some other source. 
They are evidently caused by some stimulus emanating from the fungus 
which brings about active and rapid cell division, but a stimulus which is 
very local in its effects. It seems to be true in general for this gall that 
the diffusion zone of the gall stimulus, gall poison, or whatever name we 
wish to apply to the cause of the gall formation, is very limited in extent. 
Bundles which are very abnormal in structure and which contain the 
sporanges of the fungus often occur next to bundles which are entirely 
normal and free from the fungus. Individual bundles sometimes show 
abnormalities on one side and nothing but normal tissue on the other. It 
sometimes happens that strands are formed shortly after secondary thicken- 
ing has started, and in such cases the protoxylem portions of the bundles 
are often surrounded by a cambium-like zone of cells. 
As the strands occur both in the galls near the ground and in those higher 
up on the stem, one might expect to find them extending through the in- 
tervening normal portion of the stem connecting one gall with another — 
a condition similar to that described by Smith in the crown gall. Such is 
not the case, however; they are purely local in their occurrence and extend 
but little above or below where the infection first took place in the stem. 
(For a possible explanation of the origin of the higher galls see Stewart, 10.) 
The bundles show different degrees of disturbance at different levels. 
At some level or levels in the gall the tissues are abnormal into or beyond 
the protoxylem of the bundles and probably represent the places where the 
fungus first began its activities. Going above or below the level of greatest 
disturbance the xylem becomes less and less involved, until at the upper 
and lower limits of the gall only that portion which lies just inside the cam- 
bium shows structural abnormalities. Somewhat similar conditions are 
