Survey of Northeast Greenland. 291 
If we compare the clock rate before the beginning of the journey 
with the mean clock rate for its whole duration, it appears that the 
influence of travelling conditions on the rates of the two watches, 
Ju, is the same, in that 4?u for both of them is negative. In the 
case of 57203 the change in the rate is — 35.5, as regards watch K 
— 65.4. It must, therefore, be supposed that the rate, particularly 
during the early part of the trip, has varied somewhat. After 12'IV 
one must take it for granted that the watches have adjusted themselves 
to the particular travelling conditions, and have become relatively 
balanced. Therefore, I have not hesitated exclusively to use the value 
Ju = + 375454 at the computations of the longitudes. The value 
of Ju which is obtained by means of the two determinations of 
time at Peary’s Cairn, deviates, it is true, considerably from this one; 
but I have not dared to ascribe any decisive importance to the value 
determined at Peary’s Cairn, because the observation on 21/V was 
conducted under most unfavourable conditions. Not only was the sun 
rather far removed from the prime vertical — the hour angle was 
about + 48550 — but the observation was performed in fog and 
without using coloured glass. In the observation journal the following 
entry is made: “During the first two pairs of determinations limb 
of the sun dim”. As regards both of the determinations of time at 
Peary’s Cairn it holds good that they were undertaken without the 
use of a tripod. Finally it is to be borne in mind that nearly all 
the series of observations performed in Peary Land after 7/У bear 
testimony to the fact that I was so worn out and ill that I was not 
able to strain my attention to the same degree as I was wont to do 
(see the section: Extract of Kocx’s Diary). Under conditions of this 
kind, an error of 10$ to 208 in the determination of time may easily 
occur. As the interval between the two determinations of time is 
only seven days and a half, it appears that the uncertainty of the 
observation on 21/V is in itself sufficient to explain away the deviation 
from the value + 3™545.4, found through the computation. The eircum- 
stance that watch K for the same period shows quite a similar deviation 
with the same sign points further in the direction that the cause 
must be looked for in an unfavourable observation of the sun and 
not in the watch. 
From what has been indicated above it appears that the longi- 
tudes would necessarily become more uncertain than was desirable, 
if they were to be determined solely by means of the chronometer 
57203. This was, however, not the case. 
The fact that the sledge journey in 1907 in the main took place 
in north-south we used to its full extent to determine the positions of 
the stations by latitude, in connection with a favourable azimuth in 
