294 
Thirdly. It cannot attack healthy uninjured buds, but is in fact 
a wound parasite only. It can attack young imperfectly developed 
leaves. 
Fourthly. In such attacks as on a leaf or wounded bud the plant 
defends itself from further injury by throwing off the infected dead 
portion, whether infected leaves or internode. This indeed is one of 
the great defences of the Para rubber tree, its deciduousness, and it 
is from this power of shedding its leaves and their actual short life, 
that the leaf-fungi Pestalozzia etc., do it so little harm. Before 
the fungus has been able to spread over the leaves of the tree so as 
to cause a serious injury to the foliage, the tree sheds its leaves. 
Many . of the spores must be lost on the ground, and the infection of 
adjoining leaves is naturally slower. Compare with this the infection 
of long lived nondeciduous leaves, such as Coffee with Hemileia or 
Cloves with Cephaleuros. Each leaf is infected at several different 
spots and the spread of the fungus, or in the second case the Alga, 
is much larger and more continuous. The life of the Hevea fungus 
is practically shortened with the short life of the leaf, and often it has 
apparently hardly time to produce spores before the leaf falls. 
In Borneo a lot of sixty trees together were attacked and killed 
besides isolated trees here and there in the plantations. It does not 
seem likely that all of these sixty were wounded, unless by topping, 
so that it is possible that it may be able to attack at a weak point, 
such as the young leaves as seen in Experiment E. 
Much probably depends on the power of the plant to cut off the 
infected dead portion as in the case of the leaf stalk in F. and in the 
cases where after destruction of the bud the next internodes died and 
were isolated by a growth below the dead portion. This would not 
be possible in a thick stem of a two year old tree, as the shoot would 
be too thick. 
It must be noticed that the length of the life cycle in the case of 
the petiole in F. was very short, especially when compared with that 
of the trunk of the tree sent from Borneo. The latter had been sent 
from a considerable distance to Singapore and remained for a week 
in the office before it produced spores. The upper part of this tree 
must have been diseased for a long time before it was sent as the 
cambium was destroyed for nearly its whole length. Possibly the 
development of the fruit requires the complete death of the portion 
attacked, or its almost complete deprivation of water, which would 
naturally take longer in the case of a thick stem than in a slender 
leaf stalk, which indeed had become detached before the spores were 
matured, or it might depend on the humidity of the air, for as has 
been mentioned, the fall of a heavy shower of rain after a Jong dry 
spell on the night of May 16 was followed immediately by develop- 
ment of spores. The rain ceased about 7 a.m. and the stems were 
seen to be covered with masses of spores about 8 a.m. There were 
none on the previous evening. In the case of the young leaf and 
