92 
MINUTES OF PROCEEDINGS OF 
REMARKS 
ON THE 
EMPLOYMENT OE THE SEXTANT, FOR OBSERVATIONS REQUIRING 
GREAT PRECISION. 
By Captain E. W. HAIG, E.A. 
So many books on navigation and practical astronomy liave been written 
containing descriptions of the sextant, and explanations of the method of 
using it, that it may be thought presumptuous to attempt to add anything to 
what has been already said. 
The principle which I am about to allude to must have forced itself upon 
the attention of many observers, but I have never seen it dwelt upon or 
impressed upon the beginner as of any importance, in the books that have come 
under my notice. It is to make observations in such a manner that the instru¬ 
mental errors of the sextant must necessarily be nearly eliminated in the final 
results. Thus, in determining latitude on land by circummeridian observa¬ 
tions, it is necessary, if accuracy is required, to observe objects both north 
and south of the zenith, and the same principle applies in the case of single 
altitudes for time: in this case stars both east and west of the meridian should 
be observed. In determining the longitude by lunar distances, the distances 
of objects on both sides of the moon should be measured. In all these cases 
as many observations should be made on one side as on the other, or if this 
is impracticable, as it sometimes may be, then having obtained a result from 
all observations upon one side and another from all upon the other side, the 
arithmetic mean between these two results should be adopted, without regard 
to the greater number of observations that may have been taken to produce 
one result as compared with those for the other. 
Another point to be attended to as much as possible is to choose stars 
equally distant from the zenith or moon as the case may be. 
The ordinary reconnaissance observations of daily recurrence are those for 
time and latitude, and the usual method adopted by travellers is to observe 
the sun only. Now unless they start late in the day they must halt at least 
twice on their journey to make the requisite observations, and then the 
observations for time and latitude being made at different places have to be 
reduced to the same place by dead reckoning or judgment. This also is the 
method used at sea w r here, as the ship's course and rate of speed are constantly 
recorded it is not liable to the errors occasioned by the variable circumstances 
of travelling on land. 
The method recommended is to dispense with observations of the sun 
altogether, unless for occasional convenience ( e.g . in cloudy weather where 
any opportunity is welcome), and rely upon stars only. The observations 
made should be ten altitudes of a star about the time of its culmination 
south of the zenith, and ten similar observations of a star north of the 
