Sexuality of the Uredineae . 41 
Neither in this case nor in that of Uromyces Poae was any evidence 
obtained for the origin of the fertile cells from one or several special 
hyphae, as described for a number of forms by Richards (7). The latter 
author laid considerable stress on this point at the time, as tending to 
confirm a relationship between the aecidium and the ascocarp ; but the 
work of recent years, which has so clearly established the fertile cells as 
independent sexual elements, reduces the exact origin of these cells to 
a point of little significance. The aecidium we now know is in no sense 
a definite morphological entity and cannot be compared with the ascocarp. 
Melampsora Rostrupi, Wagn. 
The aecidial stage of this form occurs on Mercurialis perennis , and 
before its relation to the Melampsora on Populus was worked out by 
Rostrup it was known as Caeoma Mercurialis. The aecidium is of par- 
ticular interest, for, like that of the Melampsoras generally, it is without 
a pseudoperidium, and, therefore, of the same type as that of Phrag- 
midium. 
A definite row of uninucleate fertile cells with sterile cells above is 
developed close beneath the epidermis just as in Phragmidium. Some- 
times, however, it is, apparently, not a superficial cell of the mycelium 
which develops into the fertile and sterile cells, but a slightly lower cell ; 
so that, later, one or two layers of small crushed cells are found above the 
sterile cells. The fertile cells become binucleate and then develop and 
produce rows of aecidiospores and intercalary cells in the typical way. 
No evidence of nuclear migration could be observed in these cases, but 
the greater width of the binucleate cells giving origin to the aecidiospore 
rows as compared with that of the uninucleate fertile cells was noticeable ; 
as was also the fact that two sterile cells could sometimes be clearly 
distinguished above the young binucleate cells (Fig. 37). This indicated 
that the conjugate nuclear condition was brought about by the fusion of 
cells as described by Christman. Owing to the age of the material the 
earliest stages of fusion were not observed, but very clear evidence of this 
fusion was obtained from later stages. In these cases the remains of the 
partition wall in the lower part of the cell were to be seen exactly as 
described by Christman (Figs. 17 and 18). 
In Melampsora Rostrupi , then, we have the other type of reduced 
fertilization in which the fertile (female) cells fuse in pairs. 
The cells of the aecidium produced in this way can no longer be 
termed ‘ fertile cells,’ since they are the product of two such cells. A single 
term is, however, obviously convenient for the cells, whatever their origin, 
which give rise to the rows of aecidiospore and intercalary cells. As the 
