230 



THE VOYAGE OF H.M.S. CHALLENGER. 



coral island, the pigeons alight in their passages upon these trees and drop the germs of more inland 

 trees. I saw the pigeons thus resting on one of the two or three trees as yet growing on Obser- 

 vatory Island, a very small islet in Nares Bay. At Banda formerly the growth of the nutmegs was 

 confined by the Dutch Government to one island of the group, Great Banda, and the trees on the 

 other islands were destroyed. It was found necessary, however, to send a Commission every year to 

 uproot the young nutmeg-trees sown on the other islands, especially Gunong Api, by the fruit- 

 pigeons. Some of the wild nutmegs in the stomachs of the birds from Pigeon Island were soft 

 and partially digested, and unfit for germination. 



" The main island immediately opposite Pigeon Island consists of a low swampy fiat of coral 

 sandstone covered with a dense growth of high trees. Immediately at the water's edge, along the 

 sandy beach, are the usual littoral trees with banks of seaweeds thrown up at their roots, whilst a 

 few yards inland a different set of trees, with tall straight trunks, grow, the trees being so closely 

 set that it is very sensibly dark beneath them. Amongst these trees is one with a vermilion-red 

 fruit, 1 which fruit was also found at Aru, and, lying thickly scattered on the mud beneath the tree, 

 is a familiar object at both places, and which was further found on the sea-surface off the north coast 

 of New Guinea amongst the driftwood from the Ambernoh river. "Whilst the ground beneath is 

 bare and muddy, and beset with the bare roots of the trees, the trunks of the trees and fallen logs 

 in these dark swampy woods are covered with the most luxuriant growth of feathery mosses and 

 Jungermannice. On one of these tree-trunks I found also a very curious fern, Triclwmancs peltatum. 

 The fronds of the fern are orbicular in form, and adhere in rows (as connected by the slender 

 rhizome) to the bark. They are pressed absolutely flat against the bark, so as to look like an adher- 

 ing crust, and have all the appearance of a Riccia, or some such form, for which, indeed, I took them 

 when I gathered the specimens by cutting off flakes of the bark. At a few hundred yards inland 

 are tracts covered with young sago palms, with several species of Zingiberaceae and large swamp- 

 ferns growing beneath them, and a Sphagnum in small quantities. On a collecting expedition to 

 this part of the island I crossed the swamp, here about half a mile in width, and came to a steep 

 rise in the land of about thirty feet or so. Here the rock appeared to be volcanic, and the soil, 

 draining itself into the swamp below, was firm and comparatively dry. The vegetation here changed 

 its aspect considerably ; and a tree fern, about six feet in height, occurred at the verge of the rise, 

 and also a Melastoma. The rising ground itself was covered with a dense growth of trees, with but 

 little underwood. Beneath these trees grew in abundance isolated tufts of Triclwmancs javanicum, 

 the erect fern-like Selaginella incequalifolia (so abundant in the Fijis, the Aru Islands, and the 

 Moluccas), and a small zingiberaceous herb. I found many trees here which I had not met with in 

 the swampy ground. They were covered with climbing Aroids, of only one species of which was I 

 able to obtain fertile specimens. 



" Asplenium nidus, and several epiphytic ferns of somewhat similar habit, were abundant; but I 

 missed the large Platycerium, so abundant in the Aru Islands. The Trichomanes javanicum and all 

 the low vegetation here was bound together by a horsehair-like Rhizomorpha, which was in sucli 

 abundance as to be a hindrance in the securing of good specimens of the plants. 



" Of palms I saw, on the whole, in the Admiralty Island, five species — the cocoa-nut, sago, and 

 areca palms, a Garyota, and a small fan palm. I procured specimens of leaves only of the two latter. 

 The fan palms appeared identical with one procured in the Aru Islands. I saw no rattans : but 



Tabernccmontana sp. 



