8 BULLETIN 1380, U. S. DEPARTMENT OF AGRICULTURE 
ing bodies either formed a continuous resupinate portion within the 
cavity up to an elevation where it became necessary to become re- 
flexed for the proper development and dispersal of spores or re- 
mained entirely resupinate and became pileate only on the sides 
of the surface roots some distance from the base of the tree (PL IV). 
On fallen trunks elevated above the ground the fungus was usually 
entirely resupinate on the under side. On large trunks, such as 
Ceiba, this resupinate condition may be very extensive and develop 
over an area of 10 or 20 feet. On trunks lying directly on the 
ground the fungus for obvious reasons always produced the typical 
brackets. Occasionally the fungus would be found apparently 
growing from the ground, but it was always found attached by 
means of a false stem to a dead root or a buried piece of wood. In 
such cases the substratum was partially enveloped by white mycelial 
sheets or strands which extended away from the wood into the 
forest duff. Between the bark and the wood in such cases the my- 
celial layers were conspicuously developed. Beneath the bark the 
mycelial development took the form of a fine thin sheet of a silky 
sheen, without the development of cords (PL III, A). On the 
surface of the roots the sheets are interlaced with smooth white 
strands or cords forming a network or radiating at the margin of 
the sheets in fanlike extensions. 
These cords are at first white and soft, but with age they become 
tough, yellowish, and discolored. This mycelial development is 
characteristic and may serve to detect the presence of the fungus. As 
is the case in some of the root fungi of the Temperate Zone there 
is every evidence that the fungus may travel by means of these 
mycelial threads from a seat of infection to uninfected roots of 
neighboring trees. This has been demonstrated in the East. For 
this reason the fungus is difficult to eradicate from plantation sites 
unless all jungle stumps are removed. 
The depth to which the roots of trees are infected by this fungus 
depends upon the character of the soil and the kind of root system. 
In the hard fine-silted baked soil of the river flats the roots of a 
Cecropia stump were sound at a depth of 10 inches. The decay 
on the laterals (which extend into the earth at an angle of about 
40°) and the taproots apparently ended abruptly at that depth. On 
high ground where the soil was porous and better drained the roots 
of the same species which had been exposed by a bank slide a few 
days before had been penetrated to a depth of 40 to 50 inches. The 
roots showed the characteristic white mycelium, with sporophore 
production at the root collar. 
There are two main t} 7 pes of root systems of tropical trees, those 
with a well-developed taproot having laterals penetrating the earth 
at a sharp angle and those without a taproot but with laterals sprawl- 
ing over or near the surface of the ground. The lateral roots in the 
last-named group are frequently conspicuously buttressed, a neces- 
sary provision in the absence of a taproot to give strength to the 
root system. The rubber tree belongs to the former group and may 
be infected and in a serious stage of decay without any evidence of 
the presence of the causal organism until after the tree is dead or 
overthrown by the wind. 
On the roots of the second group the fungus has an opportunity 
to extend its activities over a wide area, in which it comes in contact 
