400 
ON THE MEDICAL TOPOGRAPHY OF SINGAPORE. 
centre room well ventilated, but not exposed to currents or the sun’s 
radiation. This building was of brick 18 inches thick and surround¬ 
ed by a wood verandah, the roof being composed of attaps and a ceil¬ 
ing of planks. 
In estimating the correctness of the following observations the 
reader must bear in mind the conditions in which the thermometer 
was placed. It indeed most accurately shewed the temperature of 
the atmosphere of a room in that building, but not the atmos¬ 
phere of that locality ; for in the first place the building prevented 
all currents of air from affecting the temperature,which in this coun¬ 
try are the means of reducing temperature. As well might we judge 
of the temperature of a country by observations taken in a deep dell 
surrounded by hills. In the second place the brick walls while they 
absorbed tbe heat by day radiated the same by night, therefore the 
thermometer would not rise to its proper altitude nor fall sufficiently 
at night. The mean of the observations taken at this Observatory 
will be correct; but the maxima and minima will be found to differ 
from those of all other observers. To estimate correctly the tem¬ 
perature of the atmosphere a circular building should be constructed 
with a roof of attap, and a ceiling of planks having no walls, and 
from the centre the thermometer ought to be hung.* 
The limits of this paper will not permit me to give more than the 
aggregate observations, of which the following table is the mean of 
all tlie observations of each hour for every day of the month, obtain¬ 
ed by adding up the observations of each hour for every day of the 
month and dividing by the number of days. 
Table No. 2. Thermometer. 
Month. 
Month. 
. 
t 
The mean tem¬ 
perature for the 
5 years. 
January,... 
79.55 
'July, .. . 
82 24 
February,.. 
80.25 
'August. . . 
81 80 
March,. 
81.22 
! Sfinto.mhpr 
8 I 7fi 
April, .. .. 
81.47 
Oci.nhftr 
8i 21 
81® 247. 
May.. 
82.31 
Nnvcinhpr 
80 63 
June.... 
82.29 
December,. 
80/34 
H The defects ofthe observatory for thermomctrical registers are obvious, 
but perhaps not so great as our contributor considers. The atmosphere 
