—and the standard thermometer, a very carefully manufactured thermometer, with all its errors determined and tabulated. The other thermometers were all checked from this one. On top of the screen a Robinson's anemometer was screwed. This consisted of an upright rod, to the top of which were pivoted four arms free to revolve in a plane at right angles to it. At the end of these arms hemispherical cups were screwed. These were caught by the wind and the arms revolved at a speed varying with the force of the wind. The speed of the wind could be read off on a dial below the arms.
In addition there was an instrument called a Dines anemometer which supplied interesting tracings of the force, duration, and direction of the wind. There was an added advantage in the fact that the drum on which these results were recorded was comfortably housed down below, so that one could sit in a comparatively warm room and follow all the varying phases of the blizzard which was raging without. The barometer used was of the Kew Standard pattern. When the ship was crushed, all the monthly records were saved, but the detailed tracings, which had been packed up in the hold, were lost. Though interesting they were not really essential. Continuous observations were made during the long drift on the floe and while on Elephant Island the temperature was taken at midday each day as long as the thermometers lasted. The mortality amongst these instruments, especially those which were tied to string and swung round, was very high.
A few extracts from the observations taken during 1915—the series for that year being practically complete—may be of interest. January was dull and overcast, only 7 per cent. of the observations recording a clear blue sky, 71 per cent. being completely overcast.
The percentage of clear sky increased steadily up till June and July, these months showing respectively 42 per cent. and 45.7 per cent. In August 40 per cent. of the observations were clear sky, while September showed a sudden drop to 27 per cent. October weather was much the same, and November was practically overcast the whole time, clear sky showing at only 8 per cent. of the observations. In December the sky was completely overcast for nearly 90 per cent. of the time.
Temperatures on the whole were fairly high, though a sudden unexpected drop in February, after a series of heavy north-easterly gales, caused the ship to be frozen in, and effectually put an end to any hopes of landing that year. The lowest temperature experienced was in July, when —35° Fahr., i.e. 67° below freezing, was reached. Fortunately, as the sea was one mass of consolidated pack, the air was dry, and many days of fine bright sunshine occurred. Later on,
