72 H. G. Fourcade.—On the [April 26,- 
between the several axes could have been more accurately determined 
and the pivot errors investigated. 
As an example of the precision attainable with ordinary instru-- 
ments of small size, by following the methods of elimination of error 
which have been pointed out, the data of a simple quadrilateral, the 
last in the small chain of triangles already referred to, are appended : 
Observed Angles: Nov. 1890. 
+ Ne. of 
w repetitions.- 
At Spijoenkop 
/ 
Portland Height 0° 0: 0-0 
Rondevallei ... 9 21Giof 7 1 -E(1) 9°43 10 
Cornelis Kop .. 2/8°29° -O0°7 + (2) 9°43 10: 
Roode Pill 9-21-7 453) 9°43 10 
At Roode 
Spijoenkop .. 07,0; (0;0 
Rondevallei ... 44° 7°16°3+ (4) 2°40 20 
Cornelis Kop ~ 93° 9:12°1+4(5) 2°40 20 
At Cornelis Kop 
Spijoenkop ... 0; 0: 0:0 
Rondevallei ... 01°29°28°8 + (6) 9°43 10: 
Roode . 805°49°32°0 + (7) 2°40 20: 
At Rondevalle 
Touw’s Berg scat M00" 00 
Cornelis Kop 58°24" 18+(8) 9°48 10 
Roode he 83°42°12°44+(9) 25°15 10 (a) 
Spijoenkop . 125°22°39°9+(10) 9°48 10 
The length of the side Cornelis—Spijoenkop is 4066°540 Cage Roods. 
(a) The observations at’ Rondevallei between Touw’s Berg and Roode were 
made in a high wind, interrupted by rain, and the end reading taken in the’ 
dark after probable slight shifting of the vernier plate. The station could not 
conveniently be revisited, but a smaller weight was assigned to the angle in 
the calculations. 
