NOTES ON DIPTEROCARPS. 283. 
9()1° 
ee placenta attached about the middle of the larger part.. 
eu do 
IDR} att 
IKSIOhs 
180° placenta attached at 45° from the centre of one half. 
When the fruit wall is split into three parts after the manner 
of diagram 3, ascertained measurements were 
120° 136° 124° 1112° 125° X 110° 
84 194° 125° 
138° 184° 162 193° TE) 0) 
#138" 105° 109° 
It is easy to understand what happens in these seeds from 
these few measurements. Take the diagrams, which have purposely 
been oriented for the sake of this explanation; the pressure of the 
embryo is greater transversely than in any other direction and 
results in the giving way of the fruit-wall at either side: if the 
giving way occurs at points more or less diametrically opposite, 
the pressure needs no further cracking: but if the first cracks 
appear rather to the lower side of the diagrams, then the embryo 
continuing to. grow produces a new crack more or less mid way 
upon the larger part, that is generally in the neighbourhood of the: 
attachment of the placenta. If as in diagram 6 the crack is too 
much to the upper side then a third crack must appear upon the 
lower or larger part to allow of the germination proceeding. I 
measured only two fruits of this type. and I found them :— 
106 97 
127 Y 127 146 Y 117 
This method of fruit-sphtting, easy to demonstrate in Jsotoma 
borneensis, is characteristic of the greater part of the order. It 
is not dehiscence, for the product of fertilisation in the order never 
ceases growing from the moment of fertilisation to the time when 
the produced plant dies: the embryo grows into the seed and 
devours its albumen, which done it is normally cast from the parent 
tree, not quiescent as so many seeds are, but still growing, and in 
the course of its growth it ruptures the fruit-wall as described. 
Under abnormal conditions it may not be cast from the parent 
tree, and then germinates suspended (van Romburgh). 
If only we knew the workings of the process by which the 
tree cuts off nutriment from its offspring, we should know the 
strength of the barrier preventing vivipary from being anything 
but a - phenomenon exhibited by few and pecuhar Phanerog: ams. 
R. A. Soc., No; 86 1922. 
