384 BOTANICAL GAZETTE [MAY 
No nuclei were noted in actual process of division, but in a few 
instances there is some evidence of abnormal division. - In Poten- 
Ailla (fig. 4) such is the case. The nucleus is somewhat elongated, 
and at one end is stretched out into a conical point. A comparison 
of fig. 4 with HAcKER’s (25a) figs. 9, 10, 14, and 16 will show the 
close resemblance between the nuclei supposed to have divided 
recently (fig. 4) and the actual process as reported by HACKER. 
A case which is less clearly related is seen in fig. 1, where the nucleus 
is pear-shaped. Whether or not these are cases of true amitosis 
or of pseudoamitosis could not be determined, as only the end 
products were seen. 
The composition of the cell walls of the host may be changed. 
Two evidences of this are forthcoming. In Panicum, especially, 
it was noted that the walls of the parasitized cells are more deeply 
stained than those of the ordinary cells. Since the walls are not 
noticeably thickened, the only remaining explanation is that some 
change in the chemical or physical composition of the walls has 
taken place, in such a way that they have a greater affinity for 
the stain, anilin blue. In most of the diseased tissues the walls 
become brown. This color is probably due to the formation of 
tannins. There is always more or less of these substances in the 
walls, and when the cells begin to die, it is known that this brown- 
ing often takes place through the work of the tannins. In several 
cases, also, granular deposits were found in various cells in the 
diseased regions only (of Castanea, Pyrus, etc.), similar to those 
described by ZIMMERMAN (98, § 207). Just what would cause 
this precipitation is not clear. 
It was not always possible to determine which constituents of 
the host cell first showed the effect of fungous invasion. It is 
evident, however, that there is no general rule to be laid down. In 
Smilacina, for example, the chloroplasts seem to disappear before 
any of the other cell contents, while in Pyrus the nuclei of the 
sponge cells become enlarged before there are any other signs of 
change. It has already been pointed out that in some cases the 
cuticle and epidermis are affected first, as the fungus spreads from 
the center of infection, while in other cases no such result was 
observed. 
