OF THE SEEDS OF DAVIDIA INVOLUCRATA. 557 
sometimes only one. By the falling away of the dorsal valves 
the solitary, pendulous seeds become partially visible, but they 
are not released. On the contrary, they are held fast until after 
germination and considerable development of the young plant 
has taken place. 
The seeds are pendulous from above the middle of the inner 
angle of the cell, with the micropyle near the top on the outside. 
The embryo is straight, nearly as long as the seed, with foliaceous, 
flat cotyledons, twice as long as the straight radicle, and it is 
embedded in endosperm somewhat thicker than itself. 
After the dehiscence of the dorsal valves of the carpels, which 
apparently only takes place when the conditions are favourable 
to germination, tbe radicles of as many seeds as are present in 
the fruit emerge simultaneously. There is also rapid elongation 
of the cotyledons, by which the axis of the plantlet is carried 
outside of the testa. Then, according to the position of the 
fruit, the radicle turns towards the soil, which it enters, when 
the developing cotyledons absorb the reserve-materials of the 
endosperm, soon free themselves and assume a horizontal position, 
become green and persist a long time. 
This stage is soon followed by the full development of the two 
first foliage-leaves, which are opposite and at right angles to 
the cotyledons, whilst all succeeding leaves are alternate. This 
primary pair of leaves reaches a conspicuous stage in the dormant 
embryo. 
The advantages, if any, of the simultaneous germination of a 
number of imprisoned seeds in such close proximity are not 
obvious, at least to me. Unfortunately I was unable to 
follow the subsequent behaviour of seedlings produced under 
such conditions, or I might have been able to offer some ex- 
planation of the phenomenon. Of course it is no unusual thing 
in nature for a whole podful of seeds to germinate in a cluster, 
and for one, or few, to smother the rest in the struggle for 
existence. But in Davidia the proximity is unavoidable.* 
* I may mention incidentally that seedlings do sometimes, in a sense, devour 
each other. Mr. Hackett, the foreman of the Tropical Department at Kew, 
gave me the Brazil-nut as an instance. In a fruit full of nuts (seeds) he had 
under observation, the seeds germinated in the shell and decayed one after 
another until only one was left to grow up. He could not assert that this 
behaved as a saprophyte, though appearances favoured the view that it did to 
Some extent. 
