890 
Journal of Agricultural Research 
Vol. XXIV, No. n 
less interfered with, and uredospores often appear to be formed from 
almost any cell of the vertically oriented hyphae. The dark area in the 
sorus shown in Plate i, G, indicates where disorganization of abnormal 
spores or of a portion of the primordium tissue is occurring, so that space 
will be provided and the host tissue above ruptured. More detailed 
studies of the process involved in spore production in such atypical sori 
should be made in order to determine to what extent these apparent 
variations from true form are really the results of abnormal conditions. 
The “hyphal plate’' layer mentioned by Ludwig and Rees is com¬ 
monly present along the base of the sorus in P. americanum and seems 
to be almost continuous with the peridium, as shown in Plate i, B. 
The origin of this tissue is perfectly clear. The columns or chains of 
cells constituting the plectenchymatous primordium are the much en- 
larged, vertically oriented branches of the ordinary hyphae which orig¬ 
inally formed a small tangled mass at the commencement of the uredin- 
ium. These hyphae can not be distinguished from the purely vegetative 
hyphae by the size, form, or staining properties of their cells. Now, 
the sorus increases in breadth simply by the lateral extension of branches 
from this undifferentiated tissue. Certain branches grow out vertically 
and parallel the other columns of cells and form peridial, intercalary, 
and basal cells in re^lar order, so that there are bound to be platelike 
layers of cells, especially along the base at the margins, which connect 
up with the peridium. 
In order to secure further proof of the homology of intercalary cells 
and uredospore stalks, the author gathered and fixed material showing 
various stages in the development of uredinia of P. agrimoniae and P. 
hydrangeae on November 2 at Occoquan, Va. A study of sections of P. 
agrimoniae showed that Ludwig and Rees have given a good figure of a 
young sorus. There is ample evidence, on the other hand, proving 
that intercalary cells which degenerate are cut off below the peridium 
(PI. 5, D, E) and that such cells are also formed with at least the first 
uredospores (PI. 5, F). Many cases were observed where there could 
be no doubt that the basal cell forms one or more buds, spore initials 
(PI. 5, G), each of which in turn divides to form a spore supported by a 
stalk. The stalk may disappear without much or any elongation. A 
few other stages in spore formation are figured here showing that P. 
agrimoniae does not differ materially in the organization of its uredinium 
from P. americanum. 
Much of the material of P. hydrangeae was probably gathered too late 
in the season, as most of the sori sectioned had aborted or undergone 
degeneration after having developed the chains of cells constituting the 
primordium. Such stages are, in this condition (PI. 5, J), apt to be 
misleading. In a few cases the uredinia appeared to be healthy and 
normal. Some evidence was found showing that the intercalary cells 
are cut off as the peridium is formed (PI. 5, K). Certain irregularities 
in the structure of this tissue need further study. One could find stages 
where the first spores were being formed, showing that the uredospores 
originate as a result of the cutting off of a spore initial from the basal 
cell as a whole and not as a bud. In this species the cell supporting 
the spore must be considered the homolog of a true stalk cell or pedicel. 
Young spores supported by intercalary cells are shown in Plate 5, H. 
Plate 3, A, B, shows more clearly that the peridium is composed of 
the terminal cells of the sorus primordium. The subterminals take the 
