CLINTON: NORTH AMERICAN USTILAGINEAE. 
415 
nitz herbarium at the Philadelphia academy of natural sciences and 
from Schweinitz’s statement of the host as Digitaria, there is no 
doubt that he described what is now known as Ustilago Rabenhors- 
tiana and not Sorosporium Syntherismae as has been supposed. 
Fortunately a change in the current usage of names is made unnec¬ 
essary by Peck’s use of the name Ustilago Syntherismae for the 
form on Cenchrus three years before this name was first applied by 
Cooke to the smut on Panicum (Digitaria). There are some varia¬ 
tions on the different hosts, especially on Cenchrus, where the sori 
are often limited to the individual spikelets and the spore balls are 
of a more permanent character. The forms on this host approach 
Sorosporium Eoerhartii. The form on Panicum has been reported 
erroneously a number of times as Ustilago Panici-miliacei , a species 
that so far has never really been found in this country. Norton has 
described the germination of Sorosporium Syntherismae. Litera¬ 
ture: 129. 
Sorosporium Ellisii Wint. 
Sorosporium Ellisii Wint., Hedw., 1 22 : 2. 1883. 
Sorosporium Syntherismae Amer. Auct. p. p. 
Exsiccati : Sorosporium Ellisii Wint., on Andropogon scoparius , Seym. 
& Earle, Econ. Fungi, Clinton Ust. Supp., C 38, on Andropogon Vir- 
ginicus, Ell. & Ev., N. A. Fungi, 1099, on Aristida dichotoma , Ell. & Ev., 
N. A. Fungi, 1494. 
Sori elongated, including the entire inflorescence or more rarely 
confined to the individual spikelets, chiefly 1-5 cm. in length, often 
hidden by enveloping leaf sheaths, provided with evident false mem¬ 
brane within which is the black brown dusty spore mass; spore balls 
dark reddish brown, subopaque, rather temporary, oblong to sub- 
spherical, chiefly 40-100 p in length; spores somewhat irregular, 
oblong to chiefly subspherical or polyhedral, thick walled (wall often 
irregularly thickened and lighter colored where spores have been 
in contact) verruculose, chiefly 12-19 p in length. 
Hosts: Andropogon scoparius , 2 Conn., Ill., Kans.; A. Virgini¬ 
ans, 2 N. J. (type); Aristida dichotoma, Ohio, Penn. (type). 
‘Also in Bull. Torr. Bot. Club, 10: 7. 1883. 
' 2 Reported on these hosts also from Ala. and Miss., but the specimens exam¬ 
ined by the writer have all been S. Everhartii. 
