As the absence of the sun lengthened, so the cold increased. Arctic Expeditions have almost invariably registered their lowest temperatures in February and March, the months in which the earth is coldest even in England. The darkness and the low temperature of winter do not occur together: the cold, indeed, belongs rather to spring than to winter. In our case, it was not till after darkness had left us and dawn was well advanced that the state of our thermometer became a subject of general interest. We did not expect an unusually cold winter. Maps marked the “pole of cold” far south of our position, and it seemed likely that the great polar sea, though much the reverse of open, would make our winter warm. The thermometer stands were conspicuous objects as we came out from the ship to the floes. The first was supported on a barrel and snow pedestal only seventeen feet from the ship, so as to be convenient for hourly or half-hourly registration. Then came the self-registering thermometer, elevated on a tripod about thirty yards from the ship. Others were placed on the floe near shore, and on a hillock close to the beach. It may be said to be always freezing in the far north. Even in a warm summer day, when the air is perhaps 40° Fahrenheit, flakes of ice rise up from the cold sides of the floebergs, and in the shade float in a thin pellicle on the water in the ice-cracks. Meat exposed to the air keeps all the year round, and for many months our rigging was decorated with sides of musk ox and carcases of mutton. In connection with the keeping of meat, it is worth while to mention that a piece of musk ox meat, exposed for six months in the rigging, and sealed up in the cold air, remained, very unexpectedly, unchanged when the temperature rose, and was exhibited perfectly fresh three months after the Expedition returned to England. The temperature of the air sank permanently below freezing in the middle of August before we had reached winter quarters, and continued below for nine months. Fifty-four degrees of frost were registered during the October sledging. In November, mercury froze and the spirit thermometers fell to forty-five below zero (_i.e._, 77° of frost). The lowest in December was one degree colder. Then hopes of a warm winter were given up, and we watched the spirit shrink degree after degree past the coldest recorded by our predecessors. January’s lowest was 58°.7; February brought 66°.3 below zero; but on the third of March, three days after sunrise, the unparalleled temperature of 73.7 degrees below zero was indicated by our Kew-corrected thermometers, and for many hours the temperature remained more than one hundred degrees below freezing. [Illustration: EXAMINING THERMOMETER: -73.4°.]