520 Wisconsin Academy of Sciences, Arts , and Letters . 
flow of cytoplasm. The narrowness of the cell in this region 
necessitates that one nucleus precede the other, and we find the 
spindles thus arranged the one somewhat above the other. From 
analogy we may conclude with tolerable certainly that each cell 
in the division which follows will receive one daughter nucleus 
from each spindle. Still, with the spindles arranged as shown 
in Figure 7, it is possible that such is not to be the case here. 
Another interesting feature in this figure is the utter abandon¬ 
ment of the two nucleoles by the remainder of the karyokinetic 
figure. In all other nuclear divisions which I have observed in 
the rusts, the nucleole lies at no great distance from, and to 
one side of, the spindle. It would appear in this case that the 
nucleoles could never be of further use to the nuclei in the 
process of reconstruction. 
Two of the daughter nuclei from this simultaneous division 
come to lie in the hyphal bud, and two return to the basal cell 
(fig. 8). By processes similar to those described for the forma¬ 
tion of the first spore, this second outgrowth is formed into a 
uredospore and a stalk cell (fig. 9). 
How many times this process can be repeated it is difficult to 
say, since two branches are all that can well appear in 
one section. I have followed the process as far as to the for¬ 
mation of the third bud. The later stages are rendered diffi¬ 
cult of study by a lateral crowding produced by the pushing of 
the newer spores between those already formed. This distorts 
the basal structures and loosens the older spores, so that the 
original arrangement and connections of the cells are not easily 
to be made out. 
If we compare the spore formation as described here with the 
process in Phragmidium speciosum, we find that, in the forma¬ 
tion of the gametes and in their fusion, the two fungi follow a 
common method. This fact is, of course, of the greatest sig¬ 
nificance, since it shows that a process of conjugation, funda¬ 
mentally similar to that which occurs in the aecidium as an 
origin for each row of aecidiospores, may be transferred in the 
life cycle to the uredospore sorus, and lead to the formation of 
what appear to be ordinary uredospores. This condition of 
