Lutman—Life History and Cytology of the Smuts. 1203 
These may produce finally the tree-like groups such as Brefeld 
(4 & 6) has figured. The conidia (Fig. 1) during all of this 
budding are usually uninucleated. This is the case in liquid 
media that are still rich in nutrient substances. 
If, however, the conidia are taken from the dense masses on 
the plate cultures they show quite different nuclear conditions. 
Sometimes there appears only a single nucleus but there are 
more likely to be two or more in each cell (Fig. 2). The num¬ 
ber does not seem to be definitely fixed though it does not seem 
to exceed four or five. It seems to be a general rule that the 
conidia when formed in the crowded masses in the plate cultures 
may become multinucleated but when growing free in liquid 
cultures full of nutrient substances they remain uninucleated. 
In addition to becoming multinucleated these conidia lose the 
oval shape typical of the oat smut and become swollen and ir¬ 
regular (Fig. 3). A germ tube may be pushed out (Fig. 2) 
and the material in the conidium travel out into it. As progress 
is made forward by this tube, successive walls are put in behind 
the advancing cytoplasm. The forward end loses the form of a 
conidium and also becomes multinucleated (Fig. 3). In very 
old cultures these germ tubes may be seen radiating out in all 
directions from the edges of the colonies on the agar, each of 
them with a bit of nucleated protoplasm in its tip and a long 
empty tube behind it. It appears that as the conidia become 
too crowded and are enfeebled they cease budding and spend 
their last efforts in putting out germ-tubes; if these do not find 
a host plant, of course, they die. 
Sections made from parts of the leaf sheathe immediately un¬ 
der the smears will show the fungus making its way through the 
outer epidermal walls. The conidium apparently fastens itself 
to the wall and proceeds to penetrate it (Fig. 4-5) while the 
host in response to this stimulus builds the wall thicker. The 
fungus pierces further and further into the cell until finally the 
wall is passed and the conidial tube makes its way into the 
protoplast through a small opening. This continued dissolving 
and thickening of the walls produces the funnel-like openings of 
the conidial germ-tubes observed by Brefeld (5) in surface 
