ON CORAL REEFS AS A CAUSE OF FEVER. 444 
monsoon when the wind blows from the sea over the 
islets surrounded by coral exposed at low water. 
But whatever be the causes, I would reiterate my recom¬ 
mendation that all dwelling houses be built on the N. E. 
side of the Island, and that the intervening jungle betwixt it 
and the Harbour be left in its primitive state, taking care 
however to drain the swamps. The immunity enjoyed by 
the N. E. side of the island does not extend to the houses 
built on the eminence. 
I find that it is impossible to finish this most important 
subject, without inflicting on the reader part four , which will 
treat upon the feve ish localities in Sumatra, Nicobar, Mal¬ 
dives &c. &c, and the objections that may be advanced 
against this theory. 
but the secret is that he was always there when the wind blew from the island 
and never during 'he S, E monsoon for any lime, which was the reason of 
the good health of his crew, as the same precautions are now u-ed when they 
are feverish. 
