206 PROCEEDINGS OF SECTION A. 
trustworthy, errors of several feet having been discovered in 
places. It is earnestly hoped that it will not be long before 
spirit-levelling of precision is undertaken in this colony. 
The approximate adjustment of errors of close in the altitudes 
by a system of means, which on taking up the work it was found 
had been followed, will be shortly replaced by reduction on the 
principle of making the sums of the squares of the co-efficients of 
refraction to be a minimum, recognising the refraction as the 
main disturbing factor. 
CO-EFFICIENT OF REFRACTION. 
As a preliminary to the determination of the heights of stations 
along the south coast in the counties of Camden, St. Vincent, and 
Dampier, the co-efficient of refraction has been calculated from 
the reciprocal observations on nearly a hundred lines. The mean 
of all the values obtained is 
Ke 12 162e- 0012, 
In discussing these determinations the conclusion has been 
arrived at that the generally accepted formula from which the 
co-efficient is computed, 
R Sin 1” 
|e ara 
Ss 
(€ + & — 180°) 
is inadequate, and may without great labour be much improved. 
In analysing the south coast results it was observed that the 
lower values were somewhat remarkably associated with the lines 
of greatest slope and an endeavour was fruitlessly made to connect 
the two conditions. It however suggested itself that as the 
steepest lines were usually the shortest, a connection might be 
found between the change of co-efficient and change of length in 
the line. The co-efticients were accordingly grouped and gave the 
following results :— 
Under 10 miles. 10 to 15 miles. 15 to 20 miles. Over 20 miles. 
I 
K 096 ‘113 126 135 
