358 
AMERICAN JOURNAL OF BOTANY 
[Vol. 9, 
a millimeter, diagram B shows that each mycelial strand is being broken 
up or split as a new set of leaf bases {L') appear. Only a tenth of a milli- 
meter farther (C) we find that each strand has been completely split 
and that there are now six distinct regions in the cortex penetrated by 
the fungus, a pair of mycelial strands for each leaf base. At D hyphae 
have cut in beneath one leaf vein, and the union of the members of the 
other two pairs of hyphal strands is completed thirty-three sections above 
Fig. 2. Diagrams A to F made from cross sections of a branch infected with Gym- 
nosporangium clavipes, Aug., 1915, cut March, 1917. Shaded portions show location of 
mycehum and sori. A longitudinal diagram would have shown these mycelial strands 
interlacing, forming a closed network as they weave around the veins leading into the 
leaves. See text for explanation and discussion. 
at E. Note that there is some mycelium invading the epidermis on the 
proaxial side of the leaves near the lines of attachment. At the point F, a 
few sections above, two leaf blades are detached, and subcuticular and inter- 
cellular hyphae have penetrated still farther into the leaves. The last of 
these hyphae is seen in the twentieth section (not shown) above F, proving 
that the mycelium penetrated only 210 /jl into the leaf above the line of 
its attachment to the stem. These gaps between strands of mycelium 
continue up and down stems even in those showing three rings of wood, 
and one wonders if the fungus would ever occupy the entire cortex, espe- 
cially as he finds old trunk infections confined to only a small part of the 
circumference. But in an infection such as the one diagramed in figure 2 
the strands actually interlace, weaving in and out around the veins of the 
leaves and forming a closed network system in the cortical region of the 
