8 



BIRKELAND. THE NORWEGIAN AURORA POLARIS EXPEDITION, 19021903. 



The wind would sometimes come like a rushing river at the one station, while it was fairly calm 

 at the other. On the igth January, for instance, on Sukkertop, the velocity of a wind from the SSW 

 was found to be 36 metres, while on Talviktop at the same time there was no wind, this being ascer- 

 tained by telephone. The wind was heard on Talviktop, however, as a tremendous rushing from the 

 south; and an hour and a half later the wind blew with tremendous force over both mountains. 



In extreme cold and a high wind, it was uncomfortable on Talviktop. Water once froze there a 

 couple of yards from a glowing stove; and the lamp was blown out on the table in the middle of the 

 room, although in a general sense the house was well enough built. 



The worst trouble was the repeated breaking of our telephone-wires, occasioned by the snow- 

 storms. At first the telephone wires between the two summits were hung upon poles in the usual 

 manner; but this proved to be useless. Either the wires themselves were blown to pieces, or the insulators 



,\orm. pressure atO t 678 & 

 Correspond to 760 nt sea-level 



Fig. 9. 



torn down, and the line in either case destroyed. On the other hand, the wires, when laid upon the 

 ground, keep fairly well, except on hills, where great snow-drifts are heaped up upon them. In such places 

 they often came to grief; and our first work after a fall of snow and storms used to be to get them 

 repaired. 



In the same way we at first had a double line from Sukkertop down to Kaafjord; but here too 

 the wires were often broken, and we had great difficulty in repairing them. 



A couple of hours before violent winds came over Haldde, great changes were generally observed 

 in the barometer, which sometimes went up and down at intervals of a few seconds; and when this 

 occurred, we knew that it would not be safe to start from one observatory to go to the other. 



During the storms this vibration of the barometer, owing to dynamical causes, was very considerable, 

 as will be seen from the barograms, and could serve as a relative gauge for the violence of the storm. 



Figure 9 shows a couple of correlated barograph and thermograph curves drawn on Sukkertop. 

 They show the conditions during these very January storms mentioned, which moreover were the cause 

 of many casualties on the coast of Norway that year. 



