( 456 ) 
adjustment was effected by equally distributing the error of closure of 
each triangle over the 3 angles. The computation was made in a 
plane by transference by means of a Mercator’s projection according 
to the method of ScHors. 
The adjustments of the triangulation of the West Coast of Sumatra 
were made in portions; only for the base-net and for the Northern 
part the method of least squares was applied; in most cases an 
approximation method was used. The computations were made on 
the ellipsoid. 
The following remarks may be useful for a judgment of the 
accuracy which may be expected in the different connections. 
(1). The distance between the base of Simplak, on which the 
triangulation of South-Sumatra rests, and that of Padang is about 
850 kilometers; the least number of triangles, necessary for the 
transference of the length of the side Poetri-Dago of the Simplak base- 
net to the first side G3 Gadoet P1 — Poelau Satoe P2 of the 
triangulation of the West Coast of Sumatra, is 49. 
(2). The distance between the stations Gs Dempoe and the West 
extremity of the base at Padang, used for the orientation of the net, 
is about 700 kilometers; the least number of triangles, by means of 
which the azimuth of the line GS Dempoe — G+ Tenggamoes can 
be transferred to the first side of the triangulation of theWest Coast, is 40. 
(3). The distance between Telok Betong and Padang, where the 
tidal observations were made, is over 700 kilometers; the least number 
of steps, necessary for the transference of the altitude of the pillar 
at Telok Betong to the zero mark of Padang, is 24. 
The difference between the two values of the logarithm of the 
length of the connecting side is expressed in units of the 7 
decimal and corresponds to about a of the length of the side 
or to 43.2 mms. per kilometer. 
The differences found are comparatively so small that their origin 
may be easily explained by the accumulation of errors of observation 
and by the irregularities of refraction. The difference between the 
values found for the latitudes does not indicate a local deviation 
at GS Dempoe with respect to Padang. 
For the length of the connecting side a better result might have 
been expected, if for the base measurement at Padang a more 
suitable apparatus had been available. 
For the rest, the differences are such that for the purpose of the 
triangulation, namely, to afford a basis for the topographical work, 
they do not come into consideration. 
