Ml 
while the former averaged 2'2 cm. and some of them were only 19 cm. 
The volumes of two lots were found by placing them in graduated 
pvli rulers filling up to a known volume with sand, and then deduc g 
the volume of the sand. 100 fresh seeds from untapped toe. w*- 
itia 411'5 grams had a volume of 594 cubic centimetres while 100 seeds 
from tapped trees weighing 353 grams had a volume of o^y 411 ^ 
Thus the volume of the latter is only about two-thirds that of the 
former But the decrease in volume is greater than the decrease m 
weight,' and from the figures last given the density of the 
seeds is 69, while the density of tapped seed is 86. * . 4 
fche seeds of tapped trees lose more weight on drying, this increased 
density is probably due to an increased content of water. The metho 
of determining the density is only a rough one, no account being taken 
of the ^ ™ perime^t!' therefore, shows that the seeds from tapped trees 
are small^eigh less per 1,000 seeds are actually denser, but lose 
more weight in drying, than those from untapped trees. 
The Germination Tests. 
Turning now to the actual germination tests, the second part of 
each table gives the percentage germination, and the time of germma- 
e. 7 u J fil infQ Tn lots 1 and 2 of both series several seeds 
were°dull brown, without the usual mottled outer layer. This layers 
formed last of all in the fruit and its absence may evidently be taken 
as a sign that the seed has fallen before maturity, since none of these 
hrown seeds germinated. Unfortunately an unusually large proportion 
of °b ro wn ^ed^ccurred in the, first hundred, and this spoils he per- 
centage germination of the sample, though it does not affect the com 
1,ar T AsS kept for four weeks did not germinate, and those 
kept for three weeks showed only 3 percent germmation beginning 
after twenty-four days (they were kept for six weeks longer.) Piacti- 
cally the seeds from untapped trees were worthless if kept longei 
than two weeks. In series B seeds kept for five weeks did not 
germinate, but those kept for four weeks showed 28 per cent germina- 
tion beginning after seventeen days. Both in percentage, gela- 
tion and time of germination the seeds tw^msufts dea 
throughout. But it must be pointed out that these results dea 
with germination only, and they give no ini formation ^atevm- as to 
the quality of the trees which would result from the two set i of seeds- 
Indeed, the earlier germination of senes B, after the fa rst week, mig t 
indicate a precocity which is undesirable An instance of J this pi ‘eco- 
eity is recorded in the Agricultural Bulletin of the Straits &e , Yok 
VI , p. 176, where Mr. Bidley states that he was shown a little lot of 
trees of 31 years old, which had already commenced fruiting, and also 
seedlings coming away nicely from the trees. The detai s given * 
show that one or two factors have been overlooked m the P ubb ^ 
estimates of the return to be obtained by extracting oil from H* vea 
seed. Wright states (2nd edition, p. lc-5) . 
Henaratgoda and Peradeniya produce annually about 200 000 seeds 
equal approximately to one ton by weight. This assumes that 1,000 
U. 
