Thus the climate of southern Iceland is insular and not excessive. We have a notorious instance of the same disposition in England. With us Devonia represents the south-western coast of Iceland, and justifies Carrington’s high praise:

“Thou hast a cloud
For ever in thy sky; a breeze, a shower
For ever on thy meads. Yet where shall man,
Pursuing spring around the globe, refresh
His eye with scenes more beauteous than adorn
Thy fields of matchless verdure?”

The northern climate of Iceland, distant only 3° or 180 direct geographical miles, is distinctly continental; the difference ranging between 14° and 17° (F.). This is easily accounted for by the Arctic current, by the proximity of Polar ice, and by the prevalence of northern and north-western winds, which, in south Iceland as in Palestine, drive away rain. Whatever discrepancy of opinion there may be concerning the Gulf Stream, there can be none about the cold drift which, between Greenland and Iceland, measures some fifty miles in breadth, and many hundred feet in depth. Hence the north-western digitations are more subject to floes and bergs than the Breiði Fjörð, which again is oftener invested than the Faxa Fjörð, the latter being rarely beset more than once during the century. According to Uno Von Troil, the sea-ice, now so rare, came regularly in January with the north-eastern gales, and was never far from the north-east coast. At present the season is about April and even later.

In the north, according to Metcalfe (p. 152), the winter is much keener, and the summer is proportionally milder than in the south; some observers deny the truth of the latter part of the proposition, and make the hot months average about the same figure. The snow often begins with October and lasts till mid-May when the temperature stands at a mean of 35° (F.). For Akureyri Baring-Gould (quoting the Almanak um Ár 1863), gives the year as 32° (F., freezing point = Eyjafjörð), the winter as 20°·7, and the summer 45°·5. He therefore determines that, while the mean of Reykjavik is very nearly that of Moscow, Akureyri almost corresponds with Julianshaab in Greenland.

At Stykkishólm on the mid-west coast (N. lat. 65° 4´ 44´´, and W. long. (G.) 22° 43´ 17´´), observations have been taken by Hr A. O. Thorlacius for nearly thirty years. The gross results are given in the following table, taken from the Journal of the Scottish Meteorological Society, iii. 148-304:

Mean Temperature of the Months at Stykkishólm,
during the Years 1845-71.
Jan. Feb. Mar. April May. June July. Aug. Sept. Oct. Nov. Dec. Aver.
28·1 26·9 27·8 33·1 39·8 45·6 49·1 48·2 44·0 37·7 33·1 30·4 37°·0
Highest mean, 38·0 34·7 40·1 41·9 43·8 50·5 53·1 51·8 48·7 43·9 38·4 37·4 39°·8
Lowest mean, 17·2 13·3 12·4 19·8 31·4 41·5 44·2 43·0 37·2 32·5 26·4 24·0 29°·7

Mr A. Buchan, the learned Secretary of the Scottish Meteorological Society, has printed in the same Journal (1873, pp. 304-307), the following highly interesting notice on the climate of Iceland, and especially of Stykkishólm, which appear to have great differences of temperature in the same months of different years.[91]

“The mean annual temperature of the twenty-six years (1845-71) is 37°·0. The highest annual mean of any of the years was 39°·8 in 1847, and the lowest 29°·7, giving thus the enormous difference of 10°·1. This very low annual mean of 29°·7 occurred in 1866 under very exceptional circumstances, which were detailed by Mr Thorlacius in a letter 15th October 1866. Spitzbergen ice surrounded Iceland on the north and north-east coast from January to the close of August in a greater or less degree, and did not wholly disappear till about the middle of September. Its effect on the temperature of the summer was therefore perceptible. What enormous masses of ice filled up the ocean north of Iceland may be conceived from the fact that, in clear weather, its gleaming appearance could be observed from Stykkishólm twenty geographical miles, not only during the day but also at night. The depression of temperature which followed was very great, amounting on the mean of the year to 7°·3; of the nine months from January to September to 8°·1, and of February and March to 14°·5. Leaving, then, this exceptional year out of account, the next lowest annual mean was 33°·6 during 1859. Hence the coldest year fell short of the mean annual temperature to the extent of 3°·4, and the warmest year exceeded it by 2°·8.

“With 1859 began a marked diminution of temperature. For the previous thirteen years the annual mean was on each, except 1848 and 1855, above the average—the mean of these thirteen years being 38°·2, or 1°·2 above the average. For the next thirteen years the mean was only 35°·8. Thus the first half of the period was 2°·4 warmer than the last half.

“As regards the annual mean of temperature, the lowest (26°·9) occurs in February, and the highest (49°·0) in July—the difference between the coldest and the warmest months being thus 22°·1. The three coldest months are January, February, March, the mean temperature of which is 27°·6, that of December being 2°·8 higher. In the northern part of the British Isles, and at the western station of the Atlantic, these are also the three coldest months, but the difference between their mean temperature and that of December is comparatively small, whereas in the south-east and interior of Great Britain, December, January, and February are the three coldest months.