330 MR. H. MARSHALL WARD ON THE 
About one month was occupied in forming the new leaves, a 
period during which they cannot be looked upon as very useful 
to the tree. During the next five weeks or so the leaves may be 
considered as normally active, 7. e. they would during this period 
elaborate much material not used in constructing their own 
tissues, which would be passed down into the tree for the benefit 
of other portions. 
After this, however, the disease-spots appeared; and the in- 
creasing mycelium would consume more and more of the elaborated 
materials until the leaf became exhausted and fell. This latter 
period must be one of struggle between the leaf-cells and the 
fungus-cells, and, especially towards the end of the time, the tree 
cannot derive much benefit from the devastated leaf-tissues. As 
here shown for a pair of leaves so with a whole tree—in pro- 
portion as the amount of fungus-mycelium increases the lease of 
life enjoyed by the leaves decreases, other things being, generally 
speaking, equal. This also I have proved by actual experiment 
and observation, by marking certain young trees and watching 
the results as they became badly affected with the disease. 
Young trees were chosen because on such the development of 
the leaves takes place very nearly equally at all the branch-tips ; 
hence one can be certain of comparing structures of equal value. 
Some healthy plants of Coffea arabica were two years old in June, 
when they were planted in the open in a sheltered hollow, at 
equal distances apart ; at this time they had no disease, were not 
developing new leaves, and had lost but few. During August 
and September a typical plant was watched more particularly ; 
and the following facts are true, especially for it, generally for 
the rest, since all the plants were equal and developed very 
similarly. 
On Sept. 8 there were ten pairs of primary branches; a bud 
from the axil of each topmost leaf represented the eleventh pri- 
mary. The foliage was dark green, fine, and healthy. Neigh- 
bouring coffee (to leeward) had lately suffered from virulent leaf- 
disease, and was still badly affected; but very few spots were 
apparent here. 
The total number of leaves presented on this plant at the date 
given was 243; the number of cicatrices whence leaves had 
fallen amounted to 85. The terminal buds of the 11 branches 
represented 22 leaves in posse, and that of the main stem of course 
==2 leaves more, 24 in all. It is therefore clear that such a two- 
