412 ARTHUR AND MAINS: GRASS RUSTS 
striata. Then the telia were found in P. Chaseana on a host 
belonging to a family preceding the Paniceae. The final triumph 
came in the discovery of telia with the much-studied form, Uredo 
pallida (Puccinia pallescens) on Tripsacum, a grass belonging to 
the tribe Maydeae. These telia would doubtless not have been 
discovered through the usual method of examination by scraping 
the spores from the leaf with a scalpel. Sections show the sori 
to be small, indehiscent, and filled with pale, compacted spores 
(Fic. 2A) that are thin-walled and practically without pedicels. 
Many of the teliospores are three-celled, especially toward the 
center of the sorus. The whole appearance of the telia and telio- 
spores is quite unlike that of the ordinary grass rusts, and reminds 
one of those in the group of melampsoraceous rusts represented 
by Phakopsora. 
Even with this experience the telia of still another species of 
rust on Olyra latifolia were passed over fora time. The uredinia 
had been referred to Puccinia deformata Berk. & Curt., on the 
same host, a species with prominent telia, known for over fifty 
years, but with no described uredinia.* Thin-walled, nearly 
colorless urediniospores, surrounded by incurved paraphyses, 
were found, and on the same leaves were discovered telia by the 
sectioning method. So unusual was the appearance of these 
telia, however, that they were tentatively referred to the genus 
Phaksopsora, and might have remained there longer had the grasses 
been known to harbor species of this genus. It was then thought 
there might have been an error in determination of the host, and 
that it was a bamboo, instead of the Olyra. After much study the 
following description of a new species of rust was evolved, and 
incidentally the true urediniospores of P. deformata were found. 
Puccinia phakopsoroides sp. nov. 
II. Uredina amphigenous, small, round, cinnamon-brown, 
early naked, ruptured epidermis inconspicuous; paraphyses 
peripheral, incurved, clavate, 10-12 by 35-50 u, the wall cinna- 
mon-brown, I yu thick, usually thickened up to 2-3 u on the convex 
side; urediniospores ellipsoid or broadly obovoid, 19-26 by 27-35 #3 
wall slightly brownish or nearly colorless, 1-1.5 « thick, closely 


*See Arthur & Johnston, Uredinales of Cuba. Mem. Torrey Club 17: 136. 
1918. 
