ORIGIN AND PROGRESS OF GEODETIC SURVEY IN SOUTH AFRICA, 239 
For Cape Colony and Natal. 
Abstract of Triangular Errors. 
Limit of Errors of Triangles 
4 No |— Probable 
ia ‘ 5 £ " Tak Ueaitt wool Hiednate aeae " E f 
% | Portion of Chains | of | | rror 0 
2 as | 00) 05/10) 15 | 20) 25 80 | 85 | 4:0 | a Single 
(o) to | to | to | to | to | to | to | to | to | Angle 
05 | 10/15 | 20) 25) 30 35 | 4°0 | 45 | 
Troughton § Simms 18-inch Theodolite. 
M.| Zwaartkop, New- 1) é ‘ | Beedle a 
castle eae ee lies 4 | 3 1|/— .—/ 40-492 
M.| Zwaartkop, Gri- |} : ills Melba rept ed [ek ; 
qualand ; Wie q 3 3 | = — |'— | +0°394 
aloe aie | 30 12/4/7 | BA aut tenileae ft sey. |S gene 
L. | Griqualand Tri- | | ., : “ : ; : 
angulation if pal e | hi eee | Supeans lw Eee Th ero23 
| 100 41] 13 | 23) 11] 5] 6) 2/—|—| x0493 
Repsold 10-inch Theodolite. 
M.| King William’s |, 
Town to Port | Fas seats pee te 
Elizabeth i SHAE yuiteM rabet — | — | #0362 
M. | Port Elizabeth ' 
base extension. Ij TB ry NE Pe eae Ee eae ee re ole: 
M. | Port Elizabeth to | 
Caledon | \ 22)' 8) Gi) 5) kel — | — P12) — | — | 40-497 
M. | Port Elizabeth to | 1 
M.| Kimberley base | 
extension . =| } 25}10); 8} 6} 1); —|]—]}]—j|— |] —] 40-324 
M.} Hanover to Cal- | 
vinta =. ~—. {$2311 ] 8| 3) 1] —~|—|—|—| —] zoos 
p. | Tie Chain N. of |5 
Mossel Bay I 6 4) 2/—}]—)—]—]— |] — | — | 40-224 
134) 62 | 42 | 22 7}—}— 1/— | — “£0326 
» | Sir T. Maclear'’s oer 
5 A* employed in 
(2) 
& the. geodetic Bey OMS | ca Te OL fe i | |) 0622 
& circuit 
M.=Colonel Morris, R.E. L. = Lieutenant Laffan, R.E. P.=Mr, R, Pillans. 
1 This triangle has one line which crosses an arm of the sea, and its error is 
obviously abnormal. If this triangle is rejected, the probabie error of an observed 
angle in the are Port Hlizabeth-Caledon becomes +0/"329, and the general 
probable error of an angle observed with the Repsold theodolite becomes + 0'':306. 
accidental errors of measurement are certainly considerably less than the 
uncertainty of the determination of the lengths of the bars—which uncer- 
tainty is of the order of 1 : 1,000,000. 
With regard to the accuracy of triangulation, the probable error of 
