83 
enormously swollen cells shown in Plate xv, Fig. 6% in which there is no 
trace of the starch, i. c., so far as could be determined by the use of 
iodine. Testing with Schulze’s solution, or hydrochloric acid and phlo- 
roglucine, a lignified condition of the cell walls is found, though to a 
less degree than in the regions where the cells are entirely filled with 
spores. Often the cells of the cambium are so pressed out of shape 
that the tangential walls alone are distinguishable. 
If a bit of the tissue, in which spores are found so abundantly, be 
placed under a cover glass and lightly tapped the spores are set free. 
If the slide be now placed in a moist chamber and allowed to remain 
from four or five to twenty-four hours, the swarm cells are distinguished 
either on the slide (Plate xvi, Fig. 11), or escaping from the spores (Fig. 
12). The appearance of the swarm cell after escaping is that of an irreg¬ 
ular protoplasmic mass which undergoes greater or less changes in con¬ 
tour. Plate xvi, Fig. 11 u, b , c, d, c, r/, represent the changes of outline 
through which one of the swarm cells passed in about fifteen minutes. 
A very much elongated process, ciliuin, is often observed (Fig. 1 In). 
The nucleus is often plainly visible (Fig. lie, /). Nothing could be 
determined as to nuclear changes. It seems fair to suppose, however, 
that these correspond to what has been observed mother Myxomycetes. 
It is in this stage of its existence that the organism is supposed to 
penetrate the root hairs, and thus gain access to the deeper parts of 
the cortex. Repeated endeavors were made by means of slide cul¬ 
tures to observe the penetration of these swarm cells, but without suc¬ 
cess. This is a point that needs further observation. The penetration 
has never been observed, and it is possible that it is through ruptures 
in the tissue, caused by insects, worms, or other forms, which are con¬ 
stantly present in the soil. If the slide be kept in a moist chamber for 
four or five days, other and larger forms are present (Plate xvi, Fig. 13), 
while the swarm cells h ave almost entirely disappeared. It is quite proba¬ 
ble that the larger forms resultfrom a fusion of the swarm cells, but direct 
proof is wanting. These forms undergo the same changes of outline as 
described for the swarm cell (Plate xvi, Fig. 13a, />, c, d). A nucleus is 
plainly visible and a pulsating vacuole is present. It is w^orthy of note 
that in the tissues these forms are never observed, while in slide cul¬ 
tures they are very abundant. Another condition observed is repre¬ 
sented in Plate xvi, Fig. 14, and may be designated as an early plasmodial 
stage. In most cases it does not at first fill the entire lumen, and more 
or less branching filaments extend to the walls of the cell, often appar¬ 
ently continuous with the plasmodium of the adjoining cell. It often 
presents a somewhat aggregated appearance (Plate xvi, Fig. 14). Vac¬ 
uoles are always present. They are, however, of a decidedly different 
nature from those found in the forms represented in Plate xvi, Fig. 13. 
In that, no pulsation is observed. Nuclei can not be observed by the 
use of ordinary nuclear stains; acetic methyl green; picric aniline blue; 
acetic carmine, etc. The absence of starch in all cells occupied by plas- 
