Nov., 192 1 ] MAINS RUSTS ON NYSSA AND URTICASTRUM 
449 
presence of paraphyses seems no reason for excluding this rust from the- 
genus Cerotelium, the smaller number of teliospores in chains being hardly 
more than is to be expected in a simpler species of the genus. 
The Sydows^ (pp. 524, 525) consider Cerotelium as a synonym of 
Dietelia and transfer the type species Cerotelium Canavaliae Arth. of the 
former genus to the latter genus. According, however, to their description 
of the genus Dietelia, the presence of a peridium around the telium is the 
most important characteristic of this genus. The telia of. C. Canavaliae as 
such do not have a peridium, but when they arise in the old uredinial sorus, 
as they often do, they are of course surrounded by the peridium there 
present. Neither does the telial peridium of Dietelia resemble in structure 
the uredinial peridium of Cerotelium, and for these reasons the writer is 
of the opinion that Dietelia and Cerotelium should be considered as distinct 
genera. 
In the discovery of the Aecidium of Cerotelium Dicentrae the first clue 
to an aecial stage of rusts of this type has been obtained. The combination 
of subcuticular pycnia with the cupulate aecium surrounded by a peridium 
is such as was to be expected from the relationship of the genus Cerotelium 
to other genera of the Melampsoraceae. The systemic nature of the 
mycelium is probably specific and accounts for the survival of Cerotelium 
Dicentrae in a temperate climate. The infection of Bicuculla Cucullaria 
by the basidiospores of the rust must occur through the dormant buds of 
the corm, which are either exposed or immediately below the surface of 
the soil. Such a type of infection would apparently necessitate a close 
association of the two alternate hosts and probably accounts for the rather 
localized occurrence of the rust. 
Discussion of Relationships 
The character of early maturity and germination of the teliospores has 
been given considerable prominence in establishing the position of these 
two rusts. This is a character which in some groups of the rusts is of little 
or no significance. Here, however, on account of the evident grouping 
and relationship of rusts with this character it appears to take on con- 
siderable importance. Thus if we consider the rusts of the Melampsoraceae 
which have teliospores germinating without a resting period, they would 
appear to group themselves in certain definite lines of development. Start- 
ing with Aplopsora w^e have teliospores in a one-layered crust which tardily 
breaks through the epidermis and germinates at once. In Cerotelium 
Dicentrae we have another step in which a number of the sporogenous 
hyphae cut off two and three spores in succession to form a compact crust 
which, on account of its continued spore formation, rather readily ruptures 
the epidermis. The next type is represented by Cerotelium Canavaliae in 
which the number of spores produced by the sporogenous hyphae is greater 
and in consequence the epidermis is quickly ruptured and the telium is 
^ Sydow, P., and Sydow, H. Monographia Uredinearum 3: 1-726. 1915. 
