30 FRIDTJOF NANSEN. M.-N. KI. 
in which great ice-masses melt, there is much less difference between 
the salinities in spring and autumn; and also less difference in temperature 
near the surface. As example, observations from a few depths of the 
Station 5, of May 4, 1903, mentioned above, may be taken and compared 
with those at the same place taken in August, 1903. 
Stat. 5 Stat. 16 ° 
: May 4. 1903 Aug. 9, 1903 
D rol 3 
Pee 73° 32' N. 73° 30° N. Difference 
; 33° 25' E. 33° 20' E. 
Depth 294 m. Depth 290 m. 
G C. i | € Yes ® C. Giese 
o ØST 35'01 ST 35/01 3'6 0'00 
40 207 35'05 57 135/03 3°63 0'02 
50 203 35'07 459 35'05 2'56 0'02 
Ioo 191 35'05 2102583 5:0%7 Lh — 0102 
200 WOR 2507 2°89 35°10 126 — 0'03 
250 149 3507 1:80. “35105 0'37 0'02 
280 1.44 3507 095 3503 |— 049 0°04 
This is exactly what might be expected; there is practically no 
difference in salinity. At some dephts (100 and 200 metres) it is rather in the 
opposite direction. The difference in temperature is much less near the 
surface than at the former Stations; but the heat wave has been able to go 
much deeper, owing to the uniformity of salinity which has much facilitated 
the intermixture of the upper water-strata. The difference in temperature 
is therefore the same at 40 metres as near the surface; at 50 metres it 
is 2:56" C. whilst is was only 08" C, at the former Stations. In the 
deeper strata the difference is also considerably higher. Near the bottom 
some different water has evidently come in, and made the temperature 
lower. 
Formation of the Cold, Heavy Bottom-water. 
From what is said above, it may be concluded that bottom-water 
is formed in many parts of the Barents Sea by surface cooling during 
the winter; but the temperature as well as the salinity of the bottom- 
water will greatly depend on the local conditions, and may even vary 
on the same bank, if somewhat extensive. The necessary conditions 
for the formation of very cold bottom-water are, that the waters at all 
depths are fairly stationary, and free from any rapid horizontal circulation, 
so that the vertical circulation caused by cooling at the surface during 
winter, may gradually reach the bottom, and give to all water-strata a 
nearly uniform salinity, temperature, and density. After this has been 
attained, continued cooling will reduce the temperature of the whole 
