294 1. P. Косн. 
mination of position from Sælsøen through Аппекзоеп from station 
to station, by means of latitude and azimuth only. The construction 
on the map of the position of the station at the edge of the inland 
ice gives in this manner the longitude 21°42'. This agreement between 
the two observations is of course greater than might a priori be 
expected. 
For the journey across the inland ice to Ymers Nunatak it was 
impossible to use the azimuth for the determinations of position, 
because the only accurately determined points on the coastland, 
which could be identified, were situated in the northwesterly part 
of Anneksgeu, and the direction at these points only deviated very 
little from the east. The azimuth at those could, therefore, only be 
used as a control of the latitude, which was less valuable. (On the 
other hand the levelling at these points was of considerable value 
in determining the altitude of the station). The determination of 
longitude at the five stations on the inland ice was, therefore, to be 
based on the watch. A valuable control of the astronomical ob- 
servations on the inland ice I got by pointing from every station 
certain points in Dronning Louises Land, in particular the point in 
the most northerly part of the country, designated by the altitude 
1710, a very sharp and clearly defined point on a dark rocky slope. 
By marking off the stations on the map by means of latitude and 
longitude and then by drawing the sights at point 1710, it turned 
out that the five sights came very near to intersecting at one point. 
An immaterial shifting of a few of the stations was the result of this 
control. 
It might perhaps be thought that also through the levellings at 
a sharply defined sighting} point, such as point 1710, it would be pos- 
sible to achieve a control. In the following, when I come to deal 
with the levellings, I will show that the control attained in this way 
it a very unreliable one. ; 
The Geographical Sketching Survey. 
In the foregoing I have already mentioned the astronomical obser- 
vations constituting the framework of the survey upon which the 
sketching survey depends. The principles which hold good for astro- 
nomical observations under difficult conditions — simple methods 
and a control of the survey — one also ought to aim at in the 
course of the sketching surveys. Especially on far-going sledge trips 
to regions to which one is not going to return, it is likewise of great 
importance on the journey itself to draw a rough map, based upon 
a rough computation of the observations. Only in this manner can 
