LUNAR OBSERVATIONS. 
XXXlll 
These are the rates which were in strictness deducible from the observa- 
tions ; but it is necessary to explain that when the rates were actually 
assigned, which was as early in August as the observations could be collected 
from the respective observers, the mean result both of the June and of the 
August lunars, differed a few seconds from the more correct results obtained 
in a revision in the following winter, inasmuch as in their computation the 
mean refraction of the tables had not been in all instances corrected for 
variations in temperature and atmospherical pressure ; the rates then deduced 
from them were not therefore exactly those above stated ; in consequence, 
however, of the sun being on different sides of the moon, in the observations 
of June and August, and the number of distances observed on either side 
nearly equal, the errors of calculation in August counterbalanced in great 
measure those of June, the mean result being but slightly influenced, and 
affecting the deduction of rates so inconsiderably as not to require subsequent 
alteration 
The rates actually assigned were as follows ; 
228 254 259 369 404 25 
1 0.567 1^246 1^89 L 1.474 G i.986 G 9.2 
commencing at midnight of the 6th of May, and being in addition to the 
“ original differences” of that date. 
These rates were subsequently confirmed by an opportunity of comparing 
the longitude, deduced by the chronometers on the 6th of September, with 
the result of 6862 lunar distances. 
The spot, of which the longitudes were thus compared, was a low point of 
land on the coast of Melville Island to the eastward of Winter Harbour, on 
which observations were made on the 6th of September, detailed under that 
date in Appendix, No. III. ; giving its chronometrical longitude 110° 33' 59". 
This point was afterwards included in a survey of the harbour and adjacent 
* It may not be amiss to state in proof, that the ship's longitude, at noon on the 22d of July, 
using the rates actually assigned, was 60° 14' 20".7 ; whereas by using those deducible from the 
correct result of the lunars, it would have been 60° 14' 59".6. Thus it will be seen, that the 
accumulated difference of the rates for 76J days, a period much exceeding half the whole season 
of navigation of 1819, amounts to less than 40" of longitude. 
