244 I. P. Kock. 
from which it appeared that the corrections of the directions never 
reached 6”. 
A considerable uncertainty, however, is implied by the circum- 
stance that the greater part of the base was only measured once. 
This appears very strongly by a comparison between the latitude of 
the most northerly station, Muschelberg, computed by means of the 
triangulation net and the latitude observed in the same place. 
The latitude of Muschelberg is determined by means of obser- 
vations of the zenith distance of the sun near its upper culmination 
at 75°11'29".3 + 0”.7 (probable error). This gives a difference of 6.9 
from the value computed through the triangulation net, given on 
p- 237. As a starting value for the computation of the latitude of 
the triangulation stations, we used for the observation in Germania 
Harbour ф = 74°32'15”.5, which is the most probable value of all 
the determinations of latitude in the Observatory. If when fixing 
the latitude of the Observatory one had only used the zenith distance 
observations of the sun performed in that place at the upper cul- 
mination, one would from the Observatory have got ф = 74°32'16”.7 
+. 0".6 (probable error). In this manner the difference between the 
computed and the observed latitude would, as far as Muschelberg is 
concerned, become 5”.7 or about 177 metres‘). 
BORGEN and CoPELAND have attempted to explain part of the 
very marked differenee between the two values of the latitude of 
Muschelberg through the deviations of the plumb. Under certain 
plausible, but of course doubtful presumptions they compute the 
deviations of the plumb, as far as the Observatory and Muschelberg 
are concerned, at 2”.45 towards north and 1”.98 towards south 
respectively, and then they write: 
“Wenn daher die so gefundenen Ablenkungen der Lothlinie 
richtig sind, so würde der Breitenunterschied um 4.43 zu gross be- 
obachtet werden sein, und die Abweichung des geodätisch und des 
astronomisch gefundenen Abstand der Parallele würde von 182.198 
auf 44.882 Meter herabgebracht werden, eine Grösse, welche man 
wohl den Fehlern der Messung zuschreiben kann.” 
The difference of 44.882 may be explained in a plausible manner 
through the uncertainty of the two determinations of latitude in the 
Observatory and at Muschelberg. In particular, the possibility is not 
excluded that the determination of latitude in the Observatory may be 
encumbered with a systematic error of a few seconds (see pp. 252—255). 
As, however, the base has only been measured once, the possibility 
of a gross error is not excluded. In order to be on the safe side, I 
1) BORGEN and COPELAND give the difference at about 182 metres. 
