3 1 8 Whitby . — Variation in Hevea brasiliensis. 
It may be remarked, for example, that 9 *6 per cent, of the trees (trees 
giving twice the mean quantity or more) was on the average yielding 3-6 
times as much rubber per tree as the remainder of the trees. On the one 
hand, 9-6 per cent, of the trees on the area was contributing 28 per cent, of 
the total yield. On the other hand, 137 per cent, of the trees (Groups 
0-2 gms.) was contributing only 2*9 per cent, of the crop, and certainly did 
not repay the cost of tapping. The highest yielders in the population 
were four trees giving the following yields per diem: 41*45, 41*56,. 4172, 
42-77 1 grm. 
The great possibilities of seed selection in improving rubber yields are 
indicated by the above figures. 
The data obtained have a certain bearing on the conduct of tapping 
experiments. They show that the extent of variation on what is pre- 
sumably a normal area may be such that it is quite impermissible to 
assume, as is often done in tapping experiments, that small groups of trees 
(say 50 or 100 trees) chosen at random from an area of uniform age, 
situation, and appearance, will have the same yielding capacity, and hence 
that differences in yield which the groups may display when different 
tapping systems are applied to them are due to differences in the tapping 
systems, and afford a true comparison of the yielding capabilities of the 
systems in question. Thus, taking a point at random on the area with 
which the present investigation deals, counting off from it 36 rows, and 
arranging these rows in their order of succession, into groups of 3, 6, 
and 12 rows, representing respectively groups of 45, 90, and 100 trees 
we find the yields per group, under the same tapping system, to be as 
follows : 2 
Table III. 
Yields from Adjoining Groups of frees, under the same Tapping System t 
on an apparently Uniform Area . 
Grm. rubber per diem. 
Group of 45 trees. 284, 285, 204, 259, 260, 334, 392, 439, 328, 390, 325, 276. 
„ „ 90 „ 569, 463, 594, 831, 718, 601. 
„ „ 180 „ 1032, 1425, 1319. 
1 Equivalent to a yield (calculated without making allowance for the reduction which 
£ wintering &c. involves) of 33 lb. for a tapping year of 350 days. 
It may be remarked that at an earlier point in its history one of these four trees had, according 
to the estate records, been treated against an attack by white ants. It would seem certain, however, 
that the high yield which it displayed during the period of the present observations was not due to 
white ants, because ( a ) careful examination failed to reveal the presence of white ants and the tree 
was certainly alive two years after the observations had been concluded, ( b ) the greatly increased 
flow of latex which white ants are recognized as inducing is transient, whereas the present tree was 
found to give large yields over a period of two years during which it was kept under observation. 
2 The most frequent number of trees in a row was fifteen. In cases where the number was other 
than this, the yield for the row has, for the sake of simplicity, been adjusted proportionately. 
