observations oE HeaVENlV bodies With sextant. 137 
be taken with the transit theodolite* The instrument should be caretally 
levelled, care should be taken to remove the effects of parallax ( see p. 26), 
and all observations must be taken in pairs with the face of the vertical 
circle to the left and right. The correction for level error {see p. 201) 
should be applied. In nearly all theodolites, observations taken with 
the face of the vertical circle to the left are altitudes, those taken with 
the face of the vertical circle to the right are zenith distances, and must 
therefore be subtracted from 90° to convert them into altitudes. The 
only difference in computing the results from theodolite observations and 
sextant observations is that in theodolite observations, taken face right 
and face left, there is no index error, and as the altitudes are measured 
direct they are not divided by 2 as in the case of the sextant when an 
artificial horizon is used. In all other respects the computations are 
exactly the same as those given in the examples. 
Observations of Heavenly Bodies with the Sextant. 
Before any good results can be expected from sextant observations, 
the observer must be able to read the angles quickly and accurately; 
the only way to become proficient in doing this, is by practising with the 
instrument, especially at night, when the angles have to be read by the 
light of a lantern. 
Methods of obtaining accurate results .-^From the presence of the different 
sources of instrumental error mentioned on pp. 17 to 20, it is necessary* 
in order to ensure accurate results, that observations should be taken so 
as to eliminate them. 
The precise methods will be described under the head of each observa¬ 
tion, but the general principle is, that any altitudes for any purpose 
should be balanced by others taken in the opposite direction, either by 
waiting until the heavenly body has travelled to the opposite side of the 
meridian or by observing another on the opposite side taken immediately 
after, as in observations for time, or, in case of latitude, by observing 
another body on the opposite side of the zenith, as in meridian observa¬ 
tions of a star for latitude. 
Owing to the instrumental errors acting in different directions on the 
results in each case, the mean of those results will be the true time, or 
latitude, as the case may be. 
For ordinary purposes of rough mapping, these niceties are not neces- 
