ON THE DISPERSAL OF PLANTS. 283 
that they construct their nests. In that island these birds 
have often been observed to be greatly incommoded by the 
number of the seed-vessels that have been entangled by the 
broken stalks in their plumage ; and I have been informed by 
the residents that sometimes the bird has been killed by this 
cause. Although in the main distributed by sea birds, the 
seed-vessels of Pemphis acidula are also probably transported 
in the crevices of floating pumice and drift-wood. As already 
remarked, the young plant has been seen growing out of a 
piece of pumice stranded near one of the trees, 
Other well-known littoral trees in these islands, trees that 
are also to be found, like nearly all the larger plants, on the 
coral islands of the Pacific, have, without doubt, reached this 
isolated group through the agency of ocean currents. Such 
are Ochrosia parviflora and Guettarda speciosa, the fruits of 
which are often to be observed amongst the vegetable drift 
stranded on the beaches. The fruits of Guettarda speciosa 
float when newly picked. After they have lost their outer 
green covering they float more buvyantly. Those that I 
experimented on in this condition continued to float after 
remaining 50 days in sea-water. Two fruits were subse- 
quently sown out by Dr. Treub, and one seed germinated. 
The wide-ranged Hibiscus tiliaceus and Thespesia populnea, 
that originally spread their branches over the sheltered waters 
of the lagoon of Keeling Atoll, where they may be still ob- 
served, have evidently reached these islands through the 
Same agency of the ocean currents. After the capsules of 
Thespesia populnea had been about a week in sea-water, they 
began to get rotten and to break up, so that the seeds 
escaped and floated buoyantly. It happened in one instance, 
however, that the fruit sank before the seeds were liberated. 
Out of several seeds that I placed in sea-water, a few sank in 
about a month, but the majority floated during 40 days 
without any apparent injury, and would have doubtless 
floated for a still longer period. After the experiment, seven 
or eight of the seeds were sown out at Buitenzorg, and of 
these only one germinated, a proportion, however, of from 12 
to 14 per cent., which is quite sufficient to establish the fact 
that the seeds of this tree can germinate after a transportal 
over a wide tract of ocean. Unlike those of Thespesia 
populnea, the capsules of Hibiscus tiliaceus dehisce on the tree, 
when the seeds drop, sooner or later, to the ground. Its 
small, hard, reniform seeds are more likely to escape notice 
than the much larger seeds of Thespesia populnea; but they 
will float on salt water for a long time. Thcse that I exper- 
VOL. XXIV. x 
