462 I. P. Косн. 
field of vision is, at the background, bounded by mountains, the 
reflected image of the celestial vault gives the impression of being 
open, crystal water, in which the icebergs, holms and mountains 
are reflected. This “open water” has a doubly deceptive effect, be- 
cause in the autumn one must be prepared to find the current ripples 
open; as a rule one only realises that one has been the victim of an 
optical illusion, when approaching the imaginary opening, cautiously 
testing the strength of the ice. . If one’s attention is particularly 
directed towards the phenomena, one will however generally be able 
to recognize the mirage by a trifling movement in the reflected 
images or in the contours of the openings. 
The upward mirage is conditioned by the denser medium — 
the cold stratum of air — being the nethermost one. The pheno- 
menon is of rather a common occurrence in all seasons, but it 
occurs particularly often in spring on the sea ice, where it is, so to 
speak, to be seen any sunny day. Just as the temperature of the 
ice in the autumn cannot follow the rapid cooling off of the air, 
but lags behind, in the same manner it cannot keep up with the 
heating in the spring. The stratum of air nearest above the ice is 
consequently cooled off, when coming into contact with the colder 
ice, and in calm weather a stratum of cold air may form which is 
sufficiently sharply defined to bring about a total reflection. 
The common objects of the reflection under these conditions 
are the coast and the pack ice. The coast gives the impression of 
being finished off with a terrace falling off vertically towards the sea, 
which seems to be built of. vertical columns. The details. of this 
column structure are not easily got hold of, because they change so 
quickly, and because the mirage is often enveloped in a glimmering 
haze, which circumstances are connected with an extremely rapid 
change of temperature in the boundary between the strata of air. 
As an example see the thermogram for March 9—16 1908, Medd. om 
Grenl. XLII, Pl. XVIII, p. 318. In reality the elementary fluctuations 
take place much more rapidly ihan the thermogram shows; the 
instrument is not sufficiently sensitive to be able to follow the rapid 
variations of the temperature. This boundary, in other words, 
consists of an unduiating plane, which must give a changing, blurred 
and streaky effect to the reflected images (compare the moon bridge, 
the reflection. of street lamps in water and similar things). The 
pack ice when reflected appears as a rather sharply defined ice wall 
with a column structure. It is rather a common occurrence to see 
two ice walls, the one above the other, showing a reflection against 
two different bounding surfaces. (As the Danmark-Ekspedition had 
no photos of mirages of pack ice, I have in this place reproduced 
