136 BOTANICAL GAZETTE [FEBRUARY 
The first method is in striking contrast with the usual pycnial 
and aecial relation of rusts on the Angiosperms; the second method 
appears to be intermediate; while the third method apparently is 
the one most prevalent with all species of rust. Perennial infec- 
tion on the Angiosperms is found to complete the gametophytic 
development in one season, regardless of being heteroecious or 
autoecious. 
Much confusion arises as to the determination of Peridermium 
cerebrum and P. Comptoniae on certain species of pines, owing to 
Fic. 3 Fic. 4 
IGS. 3, 4.—Fig. 3, cross-section of pycnial layer of Peridermium Comptoniae on 
Pinus virginiana; fig. 4, cross-section of pycnial layer of Peridermium cerebrum on 
Pinus rigida; photomicrographs taken at same magnifications 
the similarity of infection upon the host. The types of infection 
are represented by fusiform, globoid, and semigloboid swellings. 
While the peridium and dehiscence of the aecia are good characters 
for differentiating these two species, it often occurs that specimens 
are immature or past their maturity for these characters to be 
depended upon. Under such conditions material collected in the 
spring or late summer may be differentiated by the pycnial charac- 
ters. If the aecia are mature one can examine for the presence of 
pycnia in the overlying tissue. Providing pycnia are found, the 
form would agree with P. Comptoniae; if pycnia were absent, the 
form would agree with P. cerebrum. Exudation of pycniospores 
