1905] CHRISTMAN—SEXUAL REPRODUCTION IN RUSTS 271 
Phragmidium, hyphae made up of large dense cells collect in layers 
, one to four cells deep just beneath the epidermis of the host (jig. 11). 
From these hyphae a series of large elongated cells is formed which 
raise the epidermis. A cell wall cuts each of these cells into a sterile 
cell and the gamete (fig. 12). 
The remainder of my material of Caeoma was so far advanced 
in development that each pair of gametes had already formed a con- 
siderable series of spores. Here, as in Phragmidium, a most con- 
spicuous structure through all the later stages of the development of 
the aecidium is the basal remnant of the walls which separate the 
two gametes (see jigs. 6-10). There may be more or less of it accord- 
ing to the completeness with which the adjacent walls were dissolved 
away when conjugation occurred, but in all cases it and the two dis- 
tinct bases remain throughout as an evidence of the double origin of 
each row of aecidiospores (jig. 10). 
Modifications of figs. 6-10 sometimes occur when two gametes 
lie, before fusion, with the entire length of their adjacent walls in 
contact. A very complete destruction of the walls in the formation 
of the conjugation pore now may cause the two separate bases of the 
gametes almost to disappear. A section of jig. 8, showing a lateral 
or edge view, would appear as a single large cell containing two 
nuclei. Either of these conditions gives us a figure which might 
readily be mistaken for a binucleated basidium. 
In Uromyces Caladii Pers. a further interesting modification was 
observed. In this form the fused portion of the gametes elongates 
Steatly and the two nuclei come to lie in the upper part of this region. 
Much of the base of this region, and the two basal parts of the gametes, 
are occupied by a large vacuole (fig. 13). By normal conjugate 
division of the nuclei followed by cell division, an aecidiospore 
mother-cell is formed. The fused portion of the cell is so long and 
the base is so difficult to trace on account of the vacuole that this also 
may be readily mistaken for an ordinary binucleated cell. 
BLACKMAN considers the process described by him for Phrag- 
midium violaceum as a ‘“vegetative fertilization,” in which an egg— 
the fertile cell—is fertilized by a vegetative nucleus, and his conclu- 
sions seem justified by the nature of the process in Phragmidium 
violaceum. In P. speciosum and Caeoma nitens, however, the cells 
