Survey of Northeast Greenland. 213 
computation of this kind would be loss of time, and besides the 
basis of a scientific treatment can hardly be said to exist. The 
forced work during the survey did not permit us to pick and choose 
our weather; we had to survey under all sorts of weather conditions, 
even in storm and at a low 
temperature, and consequent- 
ly there was no a priori reason HARE= FI. 
to suppose that the survey | % 
work done at the various 
stations possessed the same 
degree of accuracy. Add to 
this that there was a con- 
siderable difference between | 
the certainty with which the 
cairns could be pointed, so 
that we could not even 
reasonably take it for granted 
that the bearings from the ` 
= 
THERM - FJ. 
XL 
— 
Scale: 
single stations were deter- om "à 7 $ À kim. 
mined with nearly the same Fig. 28. The base net. 
sharpness. 
When the bearings had been corrected for eccentricity, one might 
have rested content with this and thus have left out the adjustment 
altogether, without the accuracy of the results suffering thereby to 
any considerable extent; but the fact that observer and computer 
were one and the same person, that the observation journal contained 
full particulars of the circumstances under which the survey had 
taken place on the stations, and that the computation followed so 
quickly upon the survey that we still had a lively recollection of 
the details of the latter, made it rather probable that by a purely 
practical estimate one would be able to increase the accuracy by 
adding such corrections to the angles that the sums of the angles 
became 180° in the triangles, through which the computations were 
carried. À direct profit from this would be — besides the some- 
what problematical increase of the accuracy — that it would be 
possible to control the computations of the single triangles. This 
latter circumstance is of no small importance from a practical point 
of view, in particular when one has to make one’s computation 
without an assistant. 
A treatment of the observation material as here proposed has 
no bearing on the systematic adjustment; it will always be a some- 
what arbitrary correction on the angles, partly determined by the 
knowledge possessed by the computer of the circumstances under 
