Survey of Northeast Greenland. 365 
the conditions were unfavourable (temp. + 29°) В. XXIV and В. XXVIII 
everywhere seem to give too low values; on the other hand B. XXVI 
seems to give particularly reliable altitudes. The result of the research 
is that the altitude probably lies between 2100 and 2200 metres. 
Established value 2100. 
Point 670. Both distances are here so small that the un- 
‘certainty in the refraction becomes of no importance. The working 
map shows that the two bearings intersect one another under a 
rather acute angle; the sketch shows that the sighting point has not 
been very clearly defined. The uncertainty of the position of the 
point in relation to the stations may be estimated at 1 to 2 kilo- 
metres in the direction of the sight. For station W. & Г. УП cotz 
— 0.0178; for W. & L. VII cotz — 0.0650. An error of one kilo- 
metre in the distances would thus, in the altitudes computed, give 
an error of about 18 and 65 metres respectively. Result: the value 
determined from Station W. & L. VII is the more reliable. The result 
of the investigation becomes: The tens, it is true, are uncertain, but 
ought to be given in the map; established value 670. 
Ås far as this point is concerned the extreme consequence of 
the argument would suggest its removal nearly one kilometre in 
the direction of the stations. In this manner a practically complete 
agreement between the altitudes computed would be obtained, and 
it would be possible to fix the final value at 650 metres. Here where 
beyond the levellings there are no other circumstances to suggest 
that the position of the point is wrong, I have not considered my- 
self justified in moving it. 
Point 1710. The sighting point extremely well-defined. The 
working map shows that the point is determined with great certainty. 
The leveliings from K. XII, B. XXIV and B. XXVIII which give too 
low values, ought here to be condemned. Of the five remaining 
levellings the four give values between 1692 and 1705 metres; the 
fifth levelling, taken at a distance of 65 kilometres, gives a value of 
1745 metres. The two most reliable levellings at a distance of 23.2 
and 30.0 kilometres give 1698 and 1705 metres. The most natural 
proceeding would be to put the altitude at 1700 metres; but according 
to the system followed, this would suggest that the hundreds are not 
quite reliable, which would be absurd. Established value 17101). 
1) In itself it is not a pleasant thought to put the figure of altitude 1710 on the 
map, while considering 1700 a more probable value. The drawback is, how- 
ever, not essential, compared with the evident advantages of the system. 
One might avoid the difflculty by expressing the altitudes in hectometres 
and adding a number of decimals, corresponding to the certainty with which 
the altitude was determined. In the case in point we might then have put the 
