S U M A T R A. 
ward— the laud wind has a correfpondence with the next day's fea wind— 
and thus the divination is accounted for* 
The effed: of the fea wind is not perceived to the diftance of more 
than three or four leagues from the Ihore in common, and for the moH 
part it is fainter in proportion to the diftance* When it firft fees in, ic 
docs not commence at the remoter extremity of its limits, but very 
near the fliore, and gradually extends itielf farther to fea, as the day 
advances; probably taking the longer or Ihorter courfe as the day i& 
more or leis hot. I have frequently obferved the fails of ihips^ at the 
dillance of four^ fix or eight miles, quite becalmed, whiiil a freih fea 
breeze was at the time biowmg upon the Ihore* In an hour afterwards 
they have fek itseffed^ 
Palling along the beach about fix o'clock in the evening, when the 
fea breeze is making its final efforts, I have perceived it blow with a 
coniiderable degree of warmth ; owing to the heat the fea had by that 
time acquired, which would foon begin to divert the current of air to- 
wards it, when it had firft overcome the vis inerita^ that preferves mo- 
tion' in a body after the impelling power has cealed to operate. I have 
likewife been fenfible of a degree of warmth on pafling, within two 
Kours after fun fet, to leeward of a lake of freih water ; which proves 
the aficrtion of water imbibing a more permanent heat than earth : — in 
the day-time the breeze ^vould be rendered cool in crolfing the fame 
lake. 
Approaching an ifland fituated at a diftance from any other land, I 
was ftruck with the appearance the clouds, about nine in the morning,, 
which then formed a perfed: circle round it; the middle being a clear 
azure; and refembled what the painters call a Glory* This I account 
for from the reflefted rays of the fun rarefying the atmofphere Immedi- 
ateiy over the ifland, and equally in all parts, which caufed a conflux 
of the neighbouring air, and with it the circumjacent clouds. Thefe laft, 
tending uniformly to the centerj comprefTed each other at a certain dif-* 
tance 
