528 
THE TROPICAL AGRICULTURIST. [December i, 1881 
germination, as well as fresh and recently -blown 
specimens : — 
a 
a 
of the 
ase. 
ice of 
ip fro 
» 
H. V. 
tOUlnl 
istrict 
onth. 
IS kite 
Dise 
Wind. 
ass si 
coff 
pj 
ethod 
sli 
a 't 
3 
No. of 
pores 
P 
hrs. 
Kakwana ... 
April 
8 
1 feet 
3 feet 
14 
1 
Dikoya 
May 13 
§ 
C „ 
_ II 
b 
8 
1 
Pundaluoya... 
Slight 
12 „ 
12 
Peradeniya... 
June 3 
High 
12 „ 
J f " 
flat 
8 
7 
Do. 
July 8 
Very high 
do. 
18 „ 
5 feet 
d 
12 
117 
Do. 
do. 
25 „ 
I „ 
d 
12 
21 
Do. 
J-uly 9 
do. 
i 
d 
8 
Do. 
do. 
t 
do. 
12 " 
5 „ 
d 
13 
* Very little "rust" visible. 
t A few spots of "rust" here and there. 
X "Bust" patches becoming very bad. 
§ Scarcely prcceptible. 
II 2 feet below coffee. 
f On the grass. 
a Plat on a rock. 
b Flat on rock in middle of a>tream. 
c Plat on a charred log. 
d Vertical : the smeared side towards the coffee. 
F. 
The following may serve to convey some idea of 
the enormous masses of fallen leaves which are often 
left to blow about estates : — 
(1) On a patch of 100 coffee trees, in the neighbour- 
ing of the Ramboda district, swept clean on January 
1st, 1881, the quantity of leaves which fell from the 
trees during January and February was as follows : — 
They weighed 18 lb. (very dry). Since 2,092 went 
to the pound, the experimenter estimated nearly 
38,000 leaves in all. On the same patch, from March 
1st to April 31st, the weight was 19 lb., 1,290 leavas 
to the pound. The March and April showers cause 
flushes of new leaves which are larger and heavier ; 
at the same time fewer leaves fall in those months. 
(2) In Madulsima, on a swept patch of 100 trees as 
above, there fell from 20th December, 18S0, to 20th 
January, 18S1, 22 bushels of air-dried leaves fairly 
heaped. 
(3) In Pundaluoya the following observations were 
carefully carried out. All the leaves and primings, 
which had fallen during the past ten or twelve 
mouths, on an oblong field of 6 acres, were collected 
and heaped, and the cubic contents measured. It 
was estimated that more than 1,200 cubic feet of 
leaves and primings had been collected. 
On a swept, square area, measuring about one- 
tenth of an acre (160 trees) and enclosed with 
cadjans, the leaves were allowed to fall for two 
months (February and March), and 20 bushels of 
leaves were collected ; during April and May there 
fell 11 bushels on the same area. 
G. 
I have obtained, among others, the following estim- 
ates as the cost of sweeping up fallen leaves, &c, 
in answer to inquiries, and as the result of experi- 
ments actually tried. Bach of the following para- 
graphs is quoted from the letter of an independent 
observer : — 
1, KotmaU. — "I have done 80 acres, which gives 
me a fair idea of what the cost should be, as in 
some cases the branches, &c, bad to be carried as 
much as 200 yards. Thoroughly sweeping the 
ground and carrying branches and leaves to ravines 
or other suitable places for burning has, I find, cost 
me 10 coolies per acre, which would be equal to 
R3'50 per acre." 
2. Uda Pmeelldwa. — "The cost [of sweeping up 
leaves and burying them] is rather heavier than I 
anticipated, as each man on an average did but 100 
to 200 trees a day, tho field being widely planted ; 
this would take 7 men to the acre = R2"50 per 
acre. But I only employed lads to do it, rated at 
25 or 29 cents each. They certainly can do the 
sweeping more nimbly than men, but ior the holing, 
lining and filling up, bigger men would have done 
quicker work ; and in this case I am certain 112 
per acre would be ample. Then, again, if this field 
had not been in contract, I would have made the 
sweeping work act for a weeding, and so saved 
III -50 per acre for this month." 
3. Maduhimu. — "The cost of removing the leaves 
primings, &c, was a3 nearly as I can make out, 
116 per acre. You will have noticed that they 
were removed only to a short distance— on to the 
grass fields adjacent." 
4. Mdtali. — "My official estimate for weeding and 
sweeping combined for 12 months — about ftl'25 
per acre. I have done it 12 times a year for 8 
>ears It cost me more at first by from 50 to 75 
cents an acre by contract. I do it now by estate 
labour People who do not manure so much 
as we do might weed anil sweep more cheaply." 
H 
The following illustrations give some idea of the 
proportion between the numbers of successful and 
unsuccessful flowers in the clusters up to May this 
year.* They were obtained by counting the flowers 
and young berrieB in the various clusters, and are 
all taken from "spikes" which appeared capable of 
producing much more than they actually bore : I 
estimate about 200 flowers to each " spike" during 
the season. 
The tables are not intended to prove by how much 
a tree is benefited by manure, but to illustrate (1) 
that a very large percentage of flowers become fert- 
ilized but do not ripen as crop, and (2) that more 
fertilized flowers on the whole arrive at the stage of 
green crop when the tree has been manured. It 
must be remembered further that the 5th column 
does not refer to ripe crop: the trees had to pass 
through a second attack of leaf-disease ere that 
stage, and it need only be said that a much larger 
proportion ripens on manured than on neglected 
trees, other things being equal. This proportion I 
have computed. 
In reading the results, the example No. 4, Table 
II., may be instanced : — A flowering shoot on a fine 
tree possessed nine double clusters of flowers ; of the- 
200 or so of flowers which are typically produced, 
we find that 131 were apparently fertilized, and 
commenced to swell as berries. Of these, however, 
39 never came to anything beyond the size of a large 
pinhead; the remaining 92, however, had swollen 
and (in April— May) become green crop about the 
size of a large pea. Tbe question, how many of these 
92 will ripen up as mature crop, depends on the 
condition of the tree as regards leaves, and therefore, 
on the degree of damage done to it by the forth- 
coming attacks of Hemileia. From comp-irirons with 
other trees of the same age and condition, I have 
estimated that 60 of the cherries may ripen. This 
is certainly not below the number. 
Both the tables were compiled in April — May, and 
represent what occurred largely at that time in 1881 :t 
leaf-disease appeared in June, and became bad in 
July and August, denuding most shoots of the 
majority of their leaves. This causes a further re- 
duction by the dropping of the half ripe crop, or by 
its turning yellow and shrivelling up to "light" 
brown and black masses. As far as I can ascertain, 
* The crop begins to ripen about August ; picking 
concludes about December. 
+ The seventh and last columns were compiled in 
September, after examining many trees. 
