Nov. 1905 ] The Genera Bala?isia and Dothichloe , htc. 261 
This species is very common on grasses in the Southern 
States, especially on Andropogon in Alabama and probably in 
nearby States (Andropogon virginicus, Eragrostis tenuis and 
Eragrostis campestris; Auburn, Alabama). On Panicum agros- 
toides (No. 2373 E. & E. N. A. F.) Jackson, Miss., on grass, 
Irby, Ga., Tracy; Cocoanut Grove, Florida, Thaxter. The speci¬ 
mens which I have seen of “D. atramentaria” Rav. Fung. Am. 
No. 100, Ellis & Everh. N. A. F. No. 683 have a very thin stroma. 
They either represent depauperate forms, or a different species, 
probably the former. Cultures and studies of development are 
needed to settle this point. In some cases I have found the stroma 
of this species, Dothichloe atramentosa, so situated on the under¬ 
side of the leaves of a single plant as to indicate that in the young 
stage both of the host and parasite the young stroma entirely sur¬ 
rounded the cluster of leaves before they had elongated. As the 
leaves elongate some elongate more than others, tear apart this 
common young stroma and we then find it situated on the under 
side of all the leaves of that cluster, but separated as shown in 
Plate 88, Fig. 25. It will be remembered that one of the im¬ 
portant characters of the genus Hypocrella as given by Saccardo 
(1. c.) was the unilateral position of the stroma in contradistinc¬ 
tion to the enveloping character of the stroma of Epichloe. Where 
the stroma is effuse as it is in Dothichloe atramentosa and related 
species the host has more influence in determining its partial or 
complete envelopment. In this species the fungus may begin its 
development with the position on the host ascribed to one genus, 
but is finally cast in the character of another genus by the growth 
and unequal elongation of members of the host. This together 
with the fact that in Dothichloe aristidae, a closely related species, 
the stroma normally surrounds the stem of the host, lessens the 
value of the unilateral position of the effuse stroma as a generic 
character. On the other hand the peculiar deep and restricted 
character of the stromatic bodies of Hypocrella phyllogena 
(Mont.) Speg. and Hypocrea discoidea- (B. & Br.) Sacc. of itself 
limits the extent of the stroma without regard to the host. This 
together with the difference in color, texture, etc., well separates 
these species genericaly from a plant of the type of Hypocrea 
atramentosa. Hypocrella Sacc. might well be retained for H. 
phyllogena, H. discoidea and other congeneric species. 
Dothichloe aristidae Atkinson 33. PI. 88, Figs. 28, 29. 
Stroma dimorphic, sterile portion confluent, forming a thin black 
layer, in the specimens seen entirely surrounding the culm. Fertile 
portion much thicker, confluent or interrupted, forming small 
perpendicular elevations on the sterile portion, but not forming 
distinct stromata as in Balansia since it is always thin beneath the 
bases of the perithecia as in D. atramentosa. Projecting apices 
33 Bull. Torr. Bot. Club, 21 , 224, 1894. 
