126 CAPTAIN ELLIOT ON THE LUNAR ATMOSPHERIC TIDE AT SINGAPORE. 
height of the barometer at each hour was deducted from the height given by obser- 
vation at the corresponding hours during the month ; if the height of the barometer, 
at any hour of observation, was less than the mean height at that hour, the difference 
was put down with the sign—; if greater, with the sign +. By this process the 
diurnal variation was eliminated. The residual quantities tvere then arranged in 
Tables, and the observation corresponding the nearest in time to the moon’s culmina- 
tion being marked for each day, the whole were again re-arranged in lunar tables as 
follows : — 
The moon’s superior culmination was assumed as 0 hour of lunar time, and the 
differences corresponding to that hour placed in the first column ; the remaining 
columns were similarly formed. 
The means of the sums of these differences are exhibited in Table I., which consists 
of two parts, the first part containing the barometric differences at the lunar hours 
from the superior to the inferior passage, and the second part, from the inferior to the 
superior passage ; the means for each period of six months are shown in the final 
column of the second part of the Table. 
Table II. shows the differences between the mean results and the several numbers 
in Table I. It must be borne in mind, that had the sets of weekly observations been 
complete, the sum of the minus differences would have equalled the plus ones : this, 
however, is not the case owing to the omissions above noted, and an inequality is 
thereby produced which occasioned the formation of Table I., in which the range of 
the mean values is shown, the lowest number being assumed as zero. 
The means of the complete years 1841, 1844 and 1845, are shown in Table III.; 
and as two-hourly observations were taken during the first six months of 1842 and 
of 1843*, and hourly observations during the latter portions of those years, the re- 
sults of the first six months of 1842 have been combined with the first six months of 
1843, and the hourly observations of the latter halves of these years have been com- 
bined for the same reason. 
In Table IV. the whole of the two-hourly observations, for a period of twenty- 
four months, have been added together for a general mean, and similarly the whole 
of the hourly observations for a period of thirty-six months. The results are exhibited 
in Plate X., which is drawn to '001 of barometric pressure to 0*74 of an inch of 
scale. 
Finally, Table V. exhibits the observations of three years combined, so as to show 
the effect upon the barometer of the moon when similarly situated both in its superior 
and inferior passage ; and in a column in juxtaposition is placed a similar table 
derived from two years’ observation at St. Flelena, extracted from Colonel Sabine’s 
valuable paper on the subject; from which it will be observed that the effect produced 
by the moon upon the barometer at Singapore, in the vicinity of the Equator, is 
slightly greater than at St. Helena, more distant by 14^ degrees of latitude, 
* In 1843, in consequence of a deficiency in the number of assistants. 
