416 
We have here a set of 7 observations, the moon to the left 
of the meridian about 2 hours, the sun to the right about 44 
to 5 hours. 
The moon at first low. App. alt. 5°. 43°. 
The sun at the same time higher. App. alt. 6°. 37°. 
The moon at the last had risen 1°. 40°. App. alt. 7°. 21°. 
The sun Af » fallen 3°. 45°. ‘ 3°. 20. 
The fourth observation I suspect, but this I need not 
discuss. 
Now, firstly, the mean of the first two gives the place of ob- 
servation 5". 324° to the East of its true place, that of the last 
two diminishes this error to 55°; but as, taking the mean of the 
first two, the 4th and 5th, and the 6th and 7th, the errors 
marked by them are in the proportion of 54, 34, 34, 2, nearly; 
a mean of them all cannot fairly be taken. 
Secondly: all these observations give an earlier Greenwich, 
and therefore local time, than the true time; because, allow- 
ing that the distances tabulated in the Nautical Almanac are 
correct, they make the distances between the sun and moon 
less than they ought to be, as the moon was receding from the 
sun at the time. 
Query. Can this be the error of the instrument ? 
Tt cannot be the Index Error, commonly so called, because 
T repeatedly examined this under the most favourable circum- 
stances, by taking the distance between the two opposite (hori- 
zontal) edges of the sun. 
It cannot be from simple false centering, because, at 
this point of the graduated arc, there is so small a discre- 
pancy between the readings on the two sides, and that not 
continuous. 
In order to discuss the matter properly, I will refer to some 
other examples. 
