j8o 
Journal of Agricultural Research voi.xxvn, No. io 
cylindric. Fully mature spores are perhaps most commonly broadly 
ellipsoid or broadly cylindric. A qualitative idea of the usual variations 
in form may be gained from the representative conidia grouped in Plate 
2, M. A quantitative idea of the relative predominance of the short, 
broad shape is given by the ratios of length to diameter which are 
grouped in Table 1 (p. 779) and in the diagram (fig. 1). 
Obviously, both in their small size and rotund shape the conidia, as 
revealed in this nocturnally collected material, mark Sclerospora gram- 
inicola as distinct from the other conidial Sclerosporas—namely, the 
typically ellipsoid, large-spored species of the Orient. Sclerospora 
javanica of Java, however, if we may judge from the 20 measurements 
given by Palm (77), has rotund conidia ranging from 22 to 26/* in length 
by 16 to 2 ojjl in diameter, but most commonly 24 by 18/x, a size so closely 
approaching that of the conidia of 5 . graminicola reported from Japan by 
Shirai (24) and from India by Butler (2) and Kulkami ( 12 ) that distinc¬ 
tion based on conidia alone might be difficult were it not for the con¬ 
siderably different structure and germination of these spores. 
The wall of the conidium is thin (0.5-1 jj) and of cellulose save at the 
apex where, as the conidium matures, the single so-called papilla of 
dehiscence develops (PI. 2, G, M, O, T). This is a specialized tip, 
approximately plano-convex in shape, although of variable thickness and 
area, and of modified cellulose (pectin or hemicellulose) composition. 
At the opposite (basal) end of the conidium there persists in some cases 
an apiculus of attachment (PI. 2, G), which is merely the point at 
which the conidium was affixed to the sterigma and is not at all compar¬ 
able to the apical papilla. 
At germination the apical papilla of dehiscence softens and gelatinizes, 
leaving a terminal pore (PI. 2, H) through which escape the several 
zoospores into which the granular conidium content has by then become 
differentiated. The conidium is thus in effect a zoosporangium, but 
like many other Phycomycete zoosporangia it has the potentialities of 
a conidium and may, under circumstances unfavorable to zoospore 
emergence, germinate by sending out hyphae. Under ordinary circum¬ 
stances, however, hyphal germination is rare, the writer having seen only 
a few instances among many observations. Also, even when germinat¬ 
ing by hyphae, the conidia are distinguished by the presence of the apical 
papilla of dehiscence from conidia of other species of Sclerospora in 
which hyphal germination is the rule. 
The process of germination, involving as it does the maturing of the 
zoosporangium, the development of the zoospores, the formation of the 
apical papilla of dehiscence, the deliquescence of this structure, and the 
escape and subsequent behavior of the zoospores, shows many points of 
interest which the writer hopes to work out in detail and present at a 
later date. 
It is the conidia, then, as they appear in favorable night-collected 
material which particularly distinguish the conidial stage of Sclerospora 
graminicola. Their short, thick form and small size, although relative 
characters, are of some diagnostic value in distinguishing the species. 
The possession of an apical papilla of dehiscence, however, is the one 
absolute distinction. By this character, and secondarily by the zoosporic 
germination which usually follows, 5 . graminicola , as yet, must stand as 
a unique representative of the genus. In this connection, Kulkarni’s (12) 
study on this species in India is of especial interest. This investigator 
found that, although the characteristic oogonial stage of S . graminicola 
