90) 
of the same species (e. g., H. Nos. 5 and 6), and between H. No. 20 and No. 
13, presumably of the same species, and therefore they are not of reliable 
taxonomic value. Even at the end of three months there were no sclerotia, 
pycnidia, or perithecia in any of the tubes. Ravn (91) has stressed the 
importance of sclerotia (which this medium yielded in his studies) as a 
means of distinguishing certain races or species; but no such value attaches 
to it for the races that I had under observation. 
Fresh wheat-leaves—Green wheat-leaves were prepared in the same 
manner as the wheat straw. On these leaves growth of H. No. 1 was 
much as on the wheat straw except that many more conidia were pro- 
duced both on the lower portions of the leaf and on the water-surface. 
On the upper part of the leaf, about 6 cm. above the level of the water, 
where the leaf was too dry for conidia-forniation, a rather extensive, white, 
floccose, loose aerial mycelium developed. 
SUMMARY CONCERNING THE FOREGOING VEGETABLE MEDIA 
But little of differential value resulted. The characters of aerial 
mycelium, sclerotia, clumping of mycelium, and pellicle-formation were 
but slightly marked, all being highly dependent on conditions of environ- 
ment. 
CEREAL SHOOTS GROWN FROM ASEPTIC SEEDS AS MEDIA 
Cereal seeds were disinfected and sprouted in sterile moist cham- 
bers. When the shoots were 2-3 cm. long they were cut off, placed in 
water in a test-tube, and autoclaved. Next, washed agar was placed in 
Petri dishes and inoculated with H. No.1. About four days later, when 
the colonies were several centimeters in diameter, various cereal shoots, 
prepared as above, were laid on the washed-agar plates, the basal end of 
the shoot in each case touching the edge of the colony. In this way re- 
peated tests were made with wheat, oats, corn, rye, and barley, with the 
invariable result that growth was most luxuriant and dense on the corn, 
which soon became completely black. No constant difference was evident 
between the other cereals, except that rye seemed a medium slightly less 
favorable than the others. 
It seemed probable that the more luxuriant growth and conidia-devel- 
opment on the corn shoots was due to quantity rather than to quality of 
nutrients. To test this hypothesis an entire corn shoot, a longitudinal 
half of a shoot, a longitudinal quarter of a shoot, and a mere longitudinal 
filament were laid on a colony of H. No. 1 growing on washed agar. On 
the smallest fragment, growth and conidia-production were about as on 
