THE TROPICAL 
AGRICULTURIST. [December i, 1881. 
of dry, Lot days caused all quick growth to cease, but the 
opening of the bud commenced at once after the showers 
which set in towards the end of March, and continued for 
the most part throughout April. 
On April the 2nd, in fact, after a week of showery grow- 
ing weather, the swollen bud had burst, and presented two 
leaves each § inch long, and by April 25th these were fully 
formed, handsome, bright green structures, about 4£ inches 
long by 2J inches broad in the middle : — on or about this 
date the leaves ceased to enlarge, and may be considered 
adult, and in full working order. 
During the last week in April and the first week in May, 
much rain fell in heavy showers ; but from May 8 to May 
25 a hot period intervened, the mornings being frequently 
close and steamy, however. At this time a few spots of 
" leaf disease" were observable here and there on sur- 
rounding coffee trees. 
Now, from April 25th, the period at which we may con- 
sider the leaves adult, to June 1st, no trace of Hemileia 
was discovered on either of the leaves ; but on June 1st a 
distinct, chough small " pin-spot " was seen on one of the 
leaves, which I shall call the left-hand leaf, and on Jnne 
3rd a few spores were seen proceeding from this. On the 
latter date, also, I found a minute yellow "pin-spot" on 
the right hand leaf of the pair. 
The question is, what connection had the weather, the 
presence of spores, and other circumstances with this 
definite appearance of two " leaf disease spots," on leaves 
which had been to all appearance perfectly healthy from 
April 25th to June 1st — i.e., during some five weeks ? It 
may first be stated that the weather from May 25th com- 
menced to indicate the wet usually associated with the 
incoming of the south-west monsoon, and wind and rain 
prevailed more or less up to the end of June, June 20th to 
26th being, however, fine. 
The disease was first noticed on June 1st: if the "pin- 
spot " arose from the germination of a spore as described 
above (§ 2), this spore probably commenced its action 
within three weeks or so previously. On comparing the 
notes made about the requisite period, two points are clear; 
1st, there were spores being shaken and blown about at 
that time ; 2nd, it rained heavily up to May 8th, and a 
series of hot, close, steamy mornings occurrred thereabouts, 
and it also rained during the week preceding the discovery 
of the spot. It is evident, in fact, that an odd spore of 
Hemileia germinated (probably in the dew) on each leaf on 
or about the 16th — 18th of May, and sent its tube into the 
leaf to form the mycelium of which the presence was dis- 
covered on June 1st — 3rd. 
On June 15th, the spot on the left hand leaf was large, 
and producing abundance of orange spores which were 
being widely distributed by the high winds (from June 3rd 
to 15th), as well on surrounding trees and leaves as on 
other portions of the same leaf surface. 
On June 29th, many more disease spots were apparent 
for the first time : these rapidly came to produce spores, 
and on July 1st numbered 35 new patches, each pouring 
forth hundreds of spores to be distributed as usual. The 
spores which produced these probably germinated about 
June 15th in the showers so prevalent during the month. 
By this time, also, the disease was bad all over the tree. I 
think it highly probable that the 35 new spots arose from 
spores detached from the one spot of June 3rd. 
By July 6th, the leaf was badly diseased : the one older 
spot (of July 3rd) beginning to turn brown in centre, but 
still active in spores ; the others which might be termed the 
second generation, shining through above with an orange 
hue. 
On July 15th, there appeared about 12 new spots, evid- 
ently from spores which germinated during the last week 
preceding July 1st. Each series of spots on the left hand 
leaf could now be distinguished as follows : — One large and 
old spot, with a black patch in the centre, and few spores — 
that of June 3rd. A number (35) of very active spots 
which are just commencing to become brow in the 
centre— those of June 29th. A smaller number (12) of 
new active spots and only just tinged yellow above — those 
of July loth. 
On July 23, the leaf was evidently becoming destroyed by 
the numerous (48) virulent spots draining it, and curious 
green rings around the spots of June 29th alone represented 
the normal colour of the leaf ; by July 2Gth, these had faded, 
and the leaf was quite yellow and exhausted, and it fell 
during the night of that date. 
As to the right hand leaf, its hietory is very s imil ar, On 
June 15th there were 2 Hemileia spots on it, a new one 
having appeared in addition to that of June 3rd ; on the 
29th June appeared 57 new ones, which spread rapidly, and 
covered the greater part of the leaf by the 6th July ; on 
July 15th were 13 still newer spots : the three generations 
were quite evident on July 27th, when the leaf was yellow. 
Before the last day of July this leaf also had fallen. 
It is clear from the foregoing that what the planters term 
an " attack " of leaf disease, i.e., a sudden outburst of the 
" rust," results from the coming to maturity or at about 
the same time of a series of mycelia which have been 
formed from the successful sowing of a certain number of 
spores, and since all were exposed to similar conditions we 
must look for the origin of the rust to the conditions 
previously present. It is clear, however, that we cannot 
say exactly when a given disease spot commenced to form ; 
we can only argue from the known data. I do not think 
that any mycelium takes less than one week, or more than 
three weeks to form, as a rule, however, and hence the 
above argument may be widely applied. 
But the serious matter to face is the evident cumul- 
ative power of the fungus ; in the examples cited, the 
few isolated spots produced (no doubt by the germina- 
tion of spores from fallen leaves, &c), during May, in 
their turn sowed spores which came to a head in June : 
these became sown and were successful before July, 
and so on. The effect on the tree cannot be wondered 
at, where the above recorded history of a pair of leaves 
may be approximately put as follows: — 
One month to form the pair (March 25th — April 25th) ; 
one and a quarter month in the enjoyment of normal 
functions, &c. (April 25th — June 3rd) ; three-quarters of 
a month with one or two disease spots (June 3rd — 29th) ; 
one month of continual struggle with increasing spots 
(June 29th — July 27th) ; finally resulting in the destruc- 
tion of the leaves. 
On the connection between Climate and " Leaf 
Disease," &c. 
§ 7. An early obseiwation in the present investigation 
was that the more carefully coffee plants were shelt- 
ered, the less " leaf-disease " did they appear to incur. As 
I progressed, the following truths became apparent also. 
In the S. W. monsoon, coffee plants placed in a veran- 
dah in the teeth of the wind suffer more from the 
disease than similar plants in the N. E. verandah, and 
therefore sheltered, while the reverse holds good, gener- 
ally speaking, for the alternate monsoon. In the gard- 
ens at Peradeniya, moreover, the disease hardly affects 
plants which are sheltered, at a time when the S. W. 
monsoon is blowing through exposed trees with hardly 
a leaf left on them. One patch of trees is very in- 
structive in this respect. There is one tree at the end 
through which the S. W. wind blows before reaching 
the others ; and it is dark green and bushy, long after 
the other trees are badly diseased and nearly leafless. 
The explanation seems simple : all the spores formed on 
this tree are blown away to the other trees — any spores 
to reach this tree must first travel about a quarter of 
a mile, a public road, bamboos, and the river inter- _ 
veiling. When the wind changes, however, this tree jj 
becomes very bad, being in its turn the possible re- 
cipient of every spore formed on the trees now to wind- I 
ward. With this change of wind, also, the trees at the I 
other end of the patch recover more rapidly than those I 
more to leeward. 
An extension of these observations to estates and § 
larger masses of coftee, demonstrates the fact that the I 
