298 
related to, but probably quite distinguishable from, A. hulbifer and A. 
tuhercuUger, the two species hitherto known which exhibit this character. 
The east side of this island has outside the Pandanus fence, which is 
there about three times as broad and thick as on the west, a belt of 
Thespesia populnea and Ouettarda speciosa, with patches of Pemphis 
acidula and Clerodendron inerme, and some trees of Cordia suhcordata 
and Champereia Griffithiana as well as a few thickets of Vitex Negundo 
and Desmodium umbellatum. 
The sandy isolated spit on the reef between Great Coco and Jerry 
Island is not covered even by spring-tides—it is about 70 feet long from 
north to south by some 30 feet across, and at the time of the writer’s 
visit there could be counted on it (mostly near the east side, and towards 
the south end) about a dozen germinating coco-nuts; three seedling 
Hibiscus tiliaceus, a seedling Thespesia, some seedlings of Gyrocarpus, four 
seedling Mimina, two seedling Erythrina, six seedling Garapa moluccensis, 
one seedling PavTingtonia speciosa, one seedling Entada scandens, some 
young Ipomcea biloba, and one young Gynometra, with two or three other 
species not recognised. 
In general features Little Coco so greatly resembles the other 
islands that it is unnecessary to deal with it in detail. The chief 
feature is perhaps the great abundance of Gorypha elata and Siphonodon 
celastrineus; still both species were met with, though sparingly, on the 
Great Coco. 
Before concluding, however, this general account of the vegetation of 
the islands the two fresh water accumulations deserve to be more parti¬ 
cularly noted. That on the Great Coco consists of a small lake in the 
narrow neck of land that joins the outlying north-eastern peninsula to 
the rest of the island. This lakelet is about 300 yards long and hardly 
100 yards wide, with its longer diameter across the isthmus. Its depth is 
a little over 3 feet; it is uniformly deep from side to side and from end 
to end, with a hard, even bottom. At either end it is only separated from 
the sea by some 80 to 100 yards of shingle bank, and it seems difficult 
to understand why the water it contains does not ooze out, and how it 
is that it is unaffected by the adjacent salt water, since the bottom of 
the lake is lower than the point reached by the waves that beat up on 
the single beach, if not actually lower than the level of the highest tides. 
The bottom seems to be no more than the floor of what has formerly 
been a shallow bay on the fringing-reef, and the shingle banks which 
separate it at either end from the sea seem to be nothing more than the 
ultimate embankments that would result when the causeways connecting 
outlying islets with the main island are so enlarged by accretion as to 
cease to be covered by the tides. This postulates that the present out- 
108 
