136 
Journal of Agricultural Research 
Vol. XXIII, No. 3 
wholly collapsed. Practically no substomatal vesicle was formed in 
this case, and the infecting hypha passed off directly at one end of the 
stoma. In the upper part of this drawing is shown the tip of an infect¬ 
ing hypha from another fungus which had entered an adjoining stoma. 
Plate 1, G, also from Mindum, shows the fungus a day older. It is a 
longitudinal section cut obliquely through the stoma. The substomatal 
vesicle here is well developed and has given off two hyphae at each end, 
into which the protoplasm has moved, leaving the substomatal vesicle 
nearly empty. 
Almost without exception the primary hypha or hyphae grow along 
the end of the guard cells and either run closely applied to the inner 
surface of the epidermis for a short distance in a direction parallel to 
the length of the leaf or, more rarely, slant down directly into the meso- 
phyll. The nuclei and cytoplasm follow the growing tip or tips of the 
fungus, leaving the vesicle and the older portions of the hyphae near the 
stoma practically empty. This conserves the limited resources of the 
fungus until connection with the food supplies of the host is established 
by means of haustoria. 
By the fourth day after inoculation the mycelium in Early Baart has 
attained a considerable size. The older hyphae near the stoma are 
empty or nearly so and the first two or three mesophyll cells attacked 
by the fungus are overgrown with hyphae and contain full-grown haus¬ 
toria. From this center the mycelium spreads downward obliquely to 
the vascular bundle. As the hyphae work their way through the inter¬ 
cellular spaces they usually remain in contact with the walls of the 
mesophyll cells, following the irregular curved surfaces, but sometimes a 
hypha will cut across an open space. 
Certain of the hyphae when unimpeded form long slender “runners.” 
The long straight surfaces of the cells covering the vascular bundle form 
a particularly favorable place for this, and hyphae will run lengthwise 
of the leaf along these cells for considerable distances. Plate 2, A, at a 
and its continuation a' shows a relatively short runner. The tip of 
such a rapidly growing runner is slender and tapering, the cytoplasm 
fairly dense, and the two nuclei of the terminal cell are generally left 
behind and are to be found at some distance from the tip. These nuclei 
often are elongated, and the denser-staining portion inside the nucleus 
may also be considerably drawn out as if temporarily misshapen by the 
flowing onward of protoplasm to the narrow tip of the growing cell. 
Rust nuclei elsewhere are nearly isodiametric, being usually oval. Pole- 
Evans (13) saw these elongated nuclei of the runners and interpreted 
them as cases of direct nuclear division, but I have seen no evidence of 
this. 
When the tip of a growing hypha strikes a host cell wall end-on or 
becomes wedged into an angle between two or more cells, or grows into 
a “blind alley” in the intercellular spaces of the leaf, it may make a 
haustorium. In fact, that is commonly the way in which the changes 
preparatory to the formation of a haustorium are initiated. It is not 
always essential that the growth in length of a hypha be forcibly checked, 
for a small minority of cases have been seen in which a haustorium 
was formed by a hypha whose tip appeared to be free to continue growth. 
Yet even in some of these cases (PI. 2, D, and C, b) a careful comparison 
of the sections before and after the one in question may show that the 
obstructions were present in the plane above or below. 
