422 
SURVEY OF THE INTERTROPICAL 
mo. difficult, and even dangerous, for the stones were 
Oct. 5. so loose and decomposed, that no solid footing 
could be found. The top of the rock is covered 
with a thick brush of acacia leucophtea, (of La- 
crosse Island,) many trees of which were obliged 
to be cut down or cleared away before the various 
objects could be seen from the theodolite. Mr. 
Cunningham collected here specimens of eighteen 
different sorts of plants. 
Bat Island is a mass of sandstone superincum- 
bent upon a quartzose basis, and intersected by 
nearly vertical veins of white quartz, the sur- 
face of which was in a crystallized state. The 
floor of the cavern was covered with heaps of 
water-worn fragments of quartzose rock, con- 
taining copper pyrites, in some of which the 
cavities were covered by a deposit of greenish 
calcedony. The sides of the cavern had a sta- 
lagmitical appearance, but the recess was so 
dark, that we could not ascertain either its forma- 
tion or extent ; it did not, however, appear to be 
more than twelve or fourteen yards deep. On 
first entering it we were nearly overpowered by 
a strong sulphureous smell, which was soon ac- 
counted for by the flight of an incredible number 
of small bats, which were roosting in the bottom 
of the cave, and had been disturbed by our ap- 
proach. We attempted to grope our way to the 
