Survey of Northeast Greenland. 205 
horizontal axis, excentricity of the horizontal circle or parallax of 
the verniers. 
The measurement of angles was performed as is shown by the 
following table of observations. The horizontal angles were measured 
by the “method of directions” in three sets. In the case of points 
which were not marked off as triangulation points, only two sets 
were measured. The cairn at the actual station, where the measuring 
took place, entered into the series of the points observed, in respect 
of the determination of the elements of centration; and in order to 
eliminate all doubt on later occasions as regards the position of the 
instrument, a sketch was likewise made of the position of instrument 
and station cairn and with an indication of the direction at one of 
the other cairns. Not until the three horizontal sets had been fully 
measured, were the vertical angles observed. By means of this 
arrangement one achieves a more accurate measuring of the horizontal 
angles, among other things because the attention has not at the same 
time to be strained by the measuring of the vertical angles. For the 
vertical angles only two levellings were used. 
All that in the table is put with italics is recorded in the field; 
the rest is entered in the observation journal at home, before and 
after the actual surveying operations. 
Ås the dimensions of the cairns were entered in a pocket book 
during the reconnoitring trips, they are not given in the observation 
journal. They ought, however, to have been mentioned here, as we 
should then have had all the particulars necessary for the computation 
collected in one place (The height of the cairn plays a part, in case 
it is the top that has been pointed. As a general rule the foot of 
the cairn was pointed). 
Attention is drawn to the fact that the four upper readings of 
the level correspond to the first levelling, the four lower ones to the 
second levelling. As the two observations of the level are computed 
together (example: in the case of Cairn VIII 4.4 + 12.7 + 5.0 + 14.2 
— 36.3 and 11.2 + 2.3 + 10.6 + 1.5 — 25.5) a direct comparison 
between the two values of the vertical angle does not give a complete 
picture of the accuracy arrived at. Nor is this necessary. The object 
of the two levellings is first and foremost to secure oneself against 
grave mistakes in the reading, and this security is obtained. 
The centration corrections noted in the case of the means appear 
from a special computation, further particulars of which will be 
given below. 
It has already been mentioned that the triangulation net does 
not contain a base net in the ordinary acceptation of the term, and 
that in the small triangles I contented myself with providing very 
