ORIGIN AND PROGRESS OF GEODETIC SURVEY IN SOUTH AFRICA. 237 
If we adopt instead of the result of column @ (table A) that derived in 
column c, which may possibly, for reasons explained, be the more correct, 
we should obtain a shorter length by about 1 : 200,000; on the whole IT 
am disposed to think that the probable error of the base is not much 
greater than that. 
For measurement of the base lines in the Transvaal and Orange River 
Colony much better appliances were available. The steel-bar measuring 
apparatus was transported in all cases to the site of the base, and a 480- 
foot ground standard was measured before and after the measurement of 
each section of the base, both with the steel bars and the wires used in 
measurement of the section. 
A large amount of time was devoted at Belfast to the training of the 
base-measuring staff, determining temperature coefficients of wires, and 
making experiments and tests of various kinds, so that a most effective 
working staff was created. 
I feel certain it would be true economy in all future large geodetic 
operations to have a base-measuring staff first well trained, and to measure 
the bases so soon as the reconnaissance has been completed, and before the 
angles of the chain are definitively measured. 
There have been five base lines measured in the Transvaal and Orange 
River Colony, of which the results are as follows :— 
No. of Total Length. Probable Error, 
Es Sections Matres 
m, 
* Belfast Base 9 18,994:027 +0009 =1 : 2,423,000 
Ottoshoop Base 8 17,438:178 ‘005 - 1: 3,487,000 
Wepener Base 6 21,655°280 009 +1: 2,406,000 
Kroonstad Base 4 19,787°341 ‘004 1: 4,946,000 
*Houts River Base 8 33,961°340 ‘005 1: 6,792,000 
Each section of the base was measured three times. 
Those marked * occur on the great arc of meridian. 
_ These probable.errors of the total length of the base are derived from 
the differences of the three independent measures of each section, and 
include, therefore, the effect of the accidental errors of the measurement 
with the Jaderin wires, the errors of their comparison with the ground 
standard base of 480 feet, and the measurement of the ground base with 
the steel bars. but they do not include systematic errors due to imperfect 
knowledge of the absolute length of the steel bars, which may be 
1 : 1,000,000. 
The three independent measures of the Belfast base give— 
18,994-027 métres. 
Hi fess 
“066+. 
The Houts River base is on the are of meridian north of Belfast, 
about 100 miles south of the Limpopo. The following particulars about 
its measurement may be taken to represent the accuracy that can be 
attained by the Jaderin method when properly employed. 
