Gravity in South Africa. 
91 
The observations in the tables were reduced by the 1895 caUbration. 
The final results (p. 89) are then corrected to 1910 values. 
Eemarks to Tables. 
The intervals given are for 100 coincidences in the observations at 
Johannesburg (Telephone Tower), 50 in the other cases. 
The amplitude of swing is expressed in scale divisions, each division 
being approximately 3 mm. The distance from scale to mirror and back 
to telescope (doubled distance) is measured, and allows of reducing the 
scale divisions to minutes of arc. 
When a mean-time clock or chronometer was used, its rate was 
expressed by stating the ratio of the length of its second to that of 
the sidereal clock with which it was compared. 
c = period of coincidence. 
s = ^ ^ — ^ = period of pendulum. 
When a mean-time clock or chronometer was used, the value of s 
found by it is converted to the value in terms of the sidereal clock before 
recording in the tables. 
u = correction for rate of sidereal clock. 
a = amplitude. 
t -= temperature. 
d = density of air. 
S = corrected period of pendulum. 
S is subsequently corrected for the new calibration of thermometer 63 
by the following amounts : — 
At Johannesburg (Telephone Tower) + 5 units 
Vereeniging + 17 
Cape of Good Hope + 20 
Lourengo Marques + 17 ,, 
Johannesburg Observatory +11 
Sidereal clock rates : — 
Johannesburg (Telephone Tower) ... 0-25s./day gaining 
Vereeniging 0-06s./day losing 
Cape of Good Hope 0-15s./day losing 
Johannesburg Observatory. Sidereal clock gaining 2-32s./day. 
Eiefler (mean) losing 2-32s./day, cf. to sidereal. Hence Eiefler rate 
was correct. 
