30 NATURAL CONTROL OF WHITE FLIES IN FLORIDA. 
threads interlace to form a dense papery membrane covering the 
lower surface of the leaf, and mycelial threads growing down the 
petioles and along the branch to the next leaf are often so numer- 
ous as to form a like coating on these. The authors have on many 
occasions seen watershoots 5 feet long with the undersides and 
petioles of the leaves, and the stems of the shoot, wholly coated with 
this dense mycelial growth. In one instance there were brownish 
sporelike bodies, above mentioned, scattered over the entire mycelium 
on the stem of the watershoot and along the edges and upper surface 
of the leaves. (See PL VII.) 
DEVELOPMENT. 
The development of the brown fungus on the larvae and pupse 
does not differ materially from that of the red and yellow Ascher- 
sonias already described, with the exception that after the hyphae 
have filled the insect body and have broken out around the edges, 
the stroma which then forms does not produce fruiting bodies but 
from them there grow out slender mycelial filaments which extend 
to a greater distance than those of the Aschersonias and partly take 
the place of the spores of the latter in infecting other larvae and 
pupse. As with the other fungi, insects may be killed without the 
formation of the characteristic complete stroma, or the stroma may 
be restricted in its growth to the margin of the insect. Often when 
several insects close together are infected, one large irregular stroma 
will develop over them all. 
The junior author has followed from day to day the growth of the 
mycelium of the brown fungus toward dead pupse, and the subse- 
quent development thereon of the characteristic stromas. This 
fungus is therefore definitely known to be partially saprophytic. 
This was previously suspected, since on leaves infected by it nearly 
all specimens within reach of the mycelium are overgrown and the 
usually large percentage of specimens dead from unknown causes is 
not apparent. The stroma frequently does not develop normally 
except around the margin, leaving the greater part of the body of 
the insect and the segmentation easily distinguishable. This con- 
dition is probably due in some cases to the effect of dry weather on 
the growth of the fungus, but it is considered by the authors to be 
due more often to the development of the fungus on the body of a 
dead insect. 
DISSEMINATION. 
Although Dr. Webber was unable to discover any fruiting bodies 
of the brown fungus, his observations led him to believe that the 
mycelial filaments, spreading out over the surface of the leaf from 
larvae already infected, have the power to infect other larvae and 
pupae with which they come into contact, and that it seemed probable 
