Whitby . — Variation in Hevea brasiliensis . 319 
A number of observations were made on the problem of collecting 
seeds from selected trees. 
Owing to the circumstance that the seeds are often projected to 
considerable distances from a tree when the ripe capsules burst, the seeds 
lying on the ground under a given tree are not usually all derived from the 
tree in question. In order to secure seeds whose origin was definitely 
known, the author at first placed conical bags of wire-netting round the 
capsules on the selected trees ; but it was later found sufficient to secure 
one true sample of the seeds from a given tree, as it was observed that the 
seeds from any one tree were exactly similar in appearance. 1 The 
differences which the seeds from different trees exhibit in regard to tint, 
mottling, and size, 2 are very marked, but perhaps even more noteworthy 
than such differences is the accuracy with which the tint, the mottle- 
pattern, and the shape, down to such peculiarities as slight striations on 
one side, are repeated in all the seeds from a given tree. With a little 
experience it was found possible, once a true sample had been secured to 
act as a pattern, readily to pick out the seeds from a given tree from those 
lying on the ground in its neighbourhood. It was often possible to avoid 
the trouble of placing a number of wire-netting bags on the tree for the 
purpose of securing a true sample by keeping the tree under watch for 
a short time on a hot afternoon at the period when the capsules were 
bursting in numbers. 
Variation in Girth . 
The girth of the trees was measured at two points: 22 in. and 36 in. 
from the ground. The girths recorded in the following frequency table are 
the mean of the girths at these two points. 3 4 
Table IV. 
Girths of Population of 1,011 Seven-year -old Hevea brasiliensis. 
. ,. NO 
Girth in cm. 
1 
ON 
CO 
Frequency 2 12 29 49 70 106 114 146 127 114 91 60 49 20 8 14' 
Mean girth = 80-30 + 0-25 cm. a = 11-91+0-18. C V. = 14-85+0-22 percent. 
1 For a similar observation made in Brazil, cf. Cramer: Rubber Rectieil, Amsterdam, 1914, 12. 
2 For data on the variation in weight and size, see Sprecher : Bull. Jardin Botanique de 
Buitenzorg, No. 19, 1915, p. 112. 
8 22 in. from the ground represents the lower level and 36 in. from the ground the upper level 
of the section of the trunk which was being tapped during the year in which the observations were 
made. 2 2 in. from the ground is also the level frequently fixed for measuring the girth of trees in 
order to decide whether they are large enough for tapping ; and it may therefore be stated that the 
mean girth given in the table was on the average 5.75 per cent, less than the girth measured 
at 22 in. 
4 Mean of this group, 1 1 3 cm. 
