220 
SURVEY OF THE INTERTROPICAL 
18 ) 9 - into the mangrove bushes which concealed them 
July 4. from our view. This manoeuvre was evidently 
intended to decoy us into their power, and served 
to increase our caution. 
Soon afterwards their fires were seen about a 
mile behind the mangroves, and in the evening 
the canoe was observed to pass up the river with 
s. the same two natives in it. On the 5th, we 
landed at the long north sandy point, and mea- 
sured a base line of 231 chains from the point to 
the end of the beach, where it is terminated by 
a rocky head that forms the base of a steep 
hill ; this we climbed, and, from its summit, ob- 
tained a very extensive view of the reefs near the 
coast ; but as the weather was too hazy to allow 
of our making any observation upon distant ob- 
jects, very few of the reefs in the offing were 
distinctly seen. 
On the beach we passed the wreck of a canoe, 
large enough to carry seven or eight persons ; it 
measured nineteen feet in length, and twenty-two 
inches in the bilge, and appeared, like that of 
Blomfield’s Rivulet, to be made of the trunk of the 
erythrina indica, hollowed out either by fire or by 
some blunt tool. A piece of teak-wood, one side 
of which bore the marks of green paint, was 
found washed up on the beach ; it had probably 
dropped or been thrown overboard from some 
