Survey of Northeast Greenland. 183 
Horizon II are due to a different cause. It owes its existence to the 
inversion, which must necessarily occur during the warm season, 
when the air passes closely over the relatively cold water (ice). The 
rise of temperature must necessarily be strongest immediately above 
the water, which also appears from the three values given for the 
period July 9th—August 12th: EN — 0.072 (mean altitude of the line 
of vision 52 m), k + 0.178 (mean altitude of the line of vision 15 m) 
and kl! + 0.359 (mean altitude of the line of vision 7 m). 
During the period May 31st—June 22nd the co-efficient of re- 
fraction appears very differently over land and over water (ice). The 
rather high value of the low lying lines of vision above water are 
due to similar causes as during the period July 9th to August 12th. 
As the atmospheric temperature in June is nearly three degrees lower 
than in July, though still positive, the inversion immediately above 
the ice becomes considerably smaller than in July. In the case of 
the lines of vision to the cairns IX and X one would expect a lower 
value of the co-efficient of refraction, partly because the sea wind is 
less frequent, and partly because it is only about one degree colder 
than the land wind. The month of June, however, shows such a 
low co-efficient of refraction over land that it might be reasonable 
to suppose the existence of a different cause. This is also the case. 
Fig. 14 shows that the difference of temperature between 5 m 
and 200m has a marked minimum in May. The reason of this is 
that the surface of the earth which had become considerably cooled 
in the winter, is warmed very quickly during the spring and early 
summer, on account of the radiant heat of the sun"). The heating 
of the air therefore cannot keep time with that of the earth, but lags 
behind. The relatively hot surface of the earth comes to exercise a 
heating influence on the lower strata of air. The reflexion from the 
snow, which during the month of May covers the greater part of 
the land, presumably plays no inconsiderable part in this process. 
The effect is that during the month of May, instead of a rise in 
” 
1) According to BRAND: "Die Temp. in der Ausguckstonne & c.” the actinometric 
difference (the difference between the atmospheric temperature and the black 
bulk temperature) should be greatest in June, which harmonizes with the fact 
that the sun stands highest during this month. It must however be borne in 
mind that the series of observations for the determinations of the actinometric 
difference was only made for one day in every month, and as far as possible 
on days which might be expected to be free of mist. However, the weather 
journal for 1907 shows that the month of May, out of ninety-three observations, 
only has eight with mist and nimbus, whereas June out of ninety observations 
has thirty-four with mist and nimbus. The mean value of the actinometric 
difference must therefore be supposed to be considerabiy greater in May than 
in June. 
