774 
Journal of Agricultural Research voi. xxvn, No. i© 
(PI. i, C-F). The bulbous outgrowths next elongate severally to 
club-shaped stalks which project perpendicularly from the leaf surface 
and grow rapidly larger (PI. i, G, H). At its swollen apex, each.of these 
produces successively the two to four stout primary branches (PI. i, I-K); 
while from these latter similarly arise the secondary branches (PL i, L-O). 
In like manner, successive series of branches develop until eventually the 
more or less extensive, usually dichotomous, branch system is complete, 
and the ultimate tips terminate in tapering sterigmata (PL 2, A). The 
very apex of each sterigma, beginning as a small globular swelling 
(Pl. 2, B-E), enlarges gradually, until finally it attains the shape and size 
of the mature conidium (PL 2, F), which then is separated from the neck 
of the sterigma by a wall. 
Obviously, the development of the conidiophores and conidia in 
Sclerospora graminicola as just outlined agrees very closely in its several 
stages with that already established for philippinensis and S. spon¬ 
tanea. Moreover, S. graminicola apparently agrees with these Philippine 
species in the way in which its conidia are shed. While studying the 
Philippine species, the writer became convinced that the conidia do not fall 
passively from the sterigmata as has been assumed, but rather are forcibly 
snapped off when the outbulging of the opposed walls of the basal apiculus 
of the conidium and of the sterigma tip suddenly overcomes the adhesion of 
their contiguous surfaces. This point has not been settled for S. gramini¬ 
cola; but such indications as the outbulging of the formerly flatly apposed 
walls of the sterigma tip and conidium base when released, lead to the 
conclusion that the species resembles those of the Philippines in this 
respect also. 
dependence on nocturnal moisture 
The process of conidiophore development in Sclerospora graminicola 
adheres to a regular schedule. When the infected leaves were wet with 
dew, at about 8 p. m., the outgrowths had protruded from the stomatal 
slit at about 11 p. m., and the first conidiophores and conidia were mature 
at approximately 2 a. m., while others which had begun to develop mean¬ 
while, matured successively, so that production continued, reaching its 
greatest abundance at perhaps 3 a. m., and only ceasing when the dew 
dried from the leaves in the morning sun. This nightly schedule agrees 
very closely with that found in 5 . philippinensis and S', spontanea (. 32 ) 
under conditions of dew deposition normally obtaining in the 
Philippines. 
Conidiophore production in S. graminicola , as in the Philippine species, 
is vitally dependent on the presence and persistence of dew or other 
moisture on the leaves. Infected Setaria plants that were kept dry dur¬ 
ing the night never showed conidiophore formation even though their 
leaves were obviously thoroughly invaded by vigorous mycelium, while 
similar plants exposed to dew supported abundant conidiophore pro¬ 
duction. After the leaf surface of infected plants had been wet with 
dew for about five hours, production began; if the moisture dried off 
prematurely, production coincidentally ceased; if moisture persisted un¬ 
duly into the morning, production was thus much prolonged. As in the 
Philippine Sclerosporas also, conidium production in S. graminicola is 
very sensitive in its response to moisture changes, so much so that,when 
studied at the University of Minnesota, the Setaria plants in such differ¬ 
ent localities as in a glade among trees, at the edge of a wood on a small 
hill, and in an open lot, showed slight variations in their schedule of 
