PHRAGMIDIUM 289 
takes place in arctic and alpine Uredinales. As Plowright remarks, the 
teleutospores can be found in spring on last year’s leaves, and germinate 
readily when placed in water. 
The form on §. Filipendula, which is very uncommon in this country, 
differs from that on S. Ulmaria in having the teleutospores usually smooth, 
and only occasionally provided with a few warts round the germ-pores ; 
but similar spores may be found on both hosts. In both, some of the 
spores have three super-imposed cells, as in Phragmidium, also two cells 
placed’as in Puccinia or laterally as in Diorchidium. The uredospores on 
S. Filipendula are often pyriform and as much as 35 p long; it may bea 
distinct species. 
Arthur, who records 7. Ulmariae in the North American Flora on 
Filipendula rubra, says that it is confined to one locality “ of less than 
a half hectare extent,” at Lafayette, Indiana. 
DISTRIBUTION : Europe, Siberia, Japan, Indiana. 
PHRAGMIDIUM Link. 
Autecious. 
Spermogones subcuticular, conical or flattened, without 
ostiolar filaments. Czomata indefinite, usually encircled by 
incurved paraphyses; spores in chains, each with numerous 
germ-pores. Uredo-sori definite, usually encircled by para- 
physes; uredospores borne singly on pedicels, often with 
paraphyses intermixed, pores numerous, scattered, indistinct. 
Teleutospores two- to several-celled by transverse septa; wall 
thick, laminate, usually coarsely verrucose, the middle layer dark 
and rigid; pores two or more in each cell, placed laterally ; 
pedicels often swollen below; basidiospores globose. 
This genus is confined entirely to the family Rosacee. 
There are many species in North America, but with the 
exception of P. Potentiliae and those introduced on cultivated 
roses they are all different from those of Europe. The warts 
often found on the outer surface of the teleutospores are due to 
the contraction of the external gelatinous layer, which can be 
caused to swell up enormously by boiling in lactic acid. This 
GU. 19 
