Though they were mated, laying had scarcely commenced, as we found only two eggs. They made small grottoes in the snow-drifts, and many pairs were seen billing and cooing in such shelters.

The small Wilson petrels were found living in communities under slabs of rock, and Hoadley one afternoon thought he heard some young birds crying.

Skua gulls were present in considerable force, notably near the penguin rookeries. They were breeding at the time, laying their eggs on the soil near the summit of the island. The neighbourhood of a nest was always betrayed by the behaviour of these birds who, when we intruded on them, came swooping down as if to attack us.

Although many snow petrels were seen flying about, we found only one with an egg. The nests were located in independent rocky niches but never in rookeries.

Vegetable life existed in the form of algae, in the pools, lichens on oversell rocks and mosses which grew luxuriantly, chiefly in the Adelie penguin rookeries.

Weddell seals were plentiful about the island near the tide-cracks; two of them with calves.

Though the continuous bad weather made photography impossible, Hoadley was able to make a thorough geological examination of the locality. On December 2 the clouds cleared sufficiently for photography, and after securing some snapshots we prepared to move on the next day. Dovers built a small cairn on the summit of the island and took angles to the outlying rocks.

On the 3rd we packed our specimens and left for the mainland at 9.30 A.M., arriving at the land ice-cliffs at 2 P.M. The snow surface was soft, even slushy in places, and the heat amongst the bergs along the coast of the mainland was very oppressive. After we had dug out the second sledge and re-arranged the loads, the hour was too late for sledging, so Dovers took another observation in order to obtain the rate of the half-chronometer watch. While on the island, we had examined the coast to the west with gla.s.ses and concluded that the only way to get westward was to ascend to a considerable alt.i.tude on the ice-cap, which, as far as the eye could reach, descended to the sea-level in long cascades and falls. We had expected to place a depot somewhere near Haswell Island, but such procedure was now deemed inadvisable in view of its distance from what would probably be our direct return route.

A start was made next day against an opposing wind, the sledges being relayed up a steep hillside. Later on, however, a turn was made more to the west, and it was then possible to haul both sledges at the same time. The surface was soft, so that after every halt the runners had to be cleared. The distance for the day was five and a half miles, and the night's camp was at an alt.i.tude of about one thousand five hundred feet, located just above the broken coastal ice.

During December 5 and 6 a snowstorm raged and confined us to our tent.

The high temperature caused the falling snow to melt as it touched the tent, and, when the temperature fell, the cloth became thickly coated with ice.

On the 7th the march was resumed, by skirting a small valley at an approximate alt.i.tude of two thousand feet. The ice-cap ahead descended in abrupt falls to the floe. Having a fair wind and a smooth surface, we made good headway. In the afternoon we ran into a plexus of creva.s.ses, and the surface was traversed by high ridges. The s...o...b..idges in many cases were weak and several gave way while the sledge was crossing them.

A chasm about fifty feet deep and one hundred feet long was pa.s.sed, evidently portion of a creva.s.se, one side of which had been raised.

Later in the afternoon the surface became impa.s.sable and a detour to the south was rendered necessary. This difficulty arose near the head of the valley, in which situation the ice-cap fell in a series of precipitous terraces for about one thousand feet.

At midday on the 8th we were compelled to continue the detour over a badly creva.s.sed surface, ascending most of the time. On that night, camp was pitched again amongst creva.s.ses. The sledge-meter showed only two miles one thousand one hundred yards for the afternoon, relaying having been necessary.

The sledges slipped along in the morning with a fresh breeze in their favour. The sky was covered with rapidly scudding, cirro-c.u.mulus clouds which, by midday, quite obscured the sun, making surrounding objects and even the snow at our feet indistinguishable. After continuing for four and a half miles, we were forced to camp. In the afternoon a heavy snowstorm commenced and persisted throughout the following day.

Though snow was still falling on the morning of the 11th, camp was broken at 10 A.M., and we moved off rapidly with a strong wind. During the morning the surface was gently undulating, but it mounted in a gradual ascent until nightfall. In the latter part of the afternoon the sun was clouded over, and steering had to be done by the aid of the wind. To the north we had a fine view of Drygalski's "High Land" (Drygalski Island), perceiving a distinct seaward ice-cliff of considerable height.

As there were no prominences on the ice-cap that could be used for surveying marks, Dovers had considerable difficulty in keeping a reckoning of our course. The trouble was overcome by building snow-mounds and taking back-angles to them with the prismatic compa.s.s.

At this juncture we were about ten miles from the sh.o.r.e and could see open water some thirty miles to the north. Frozen fast within the floe were great numbers of bergs.

We started off early on December 12 with the aid of a fair breeze over a good surface, so that both sledges were easily hauled along together.

The course was almost due west, parallel to the coast. Open water came within a few miles of the ice-cliffs, and, farther north, a heavy belt of pack was observed. When the sun sank lower, the bergs on the northern horizon were refracted up to such a degree that they appeared to be hanging from the sky.

The aid rendered by the sail under the influence of a fair breeze was well shown on the following day. In four hours, on a good surface, both sledges were transported seven miles. When we moved off, the wind was blowing at ten to fifteen miles an hour. By 10 A.M. the sky became overcast and the wind freshened. Camp was pitched for lunch at 11 A.M., as we hoped that the weather would clear again later, but the wind increased and snow began to fall heavily in the afternoon, so we did not stir. The storm continued throughout the following day and it was impossible to march until the 15th.

Continuing the ascent on the 16th out of a valley we had crossed on the previous day, we halted on the top of a ridge within view of German "territory"--a small, dark object bearing due west, evidently bare rock and presumably Gaussberg. The course was altered accordingly towards this object and everything went smoothly for ten miles. Then followed an area where the ice fell steeply in waves to the sea, crossed by creva.s.ses which averaged fifty feet in width. The snow-bridges were deeply concave, and the lower side of each chasm was raised into a ridge five to ten feet high. Making fast the alpine rope on to the sledges, one of us went ahead to test the bridge, and then the sledges, one at a time, were rushed down into the trough and up on the other side. After crossing ten or more creva.s.ses in this fashion, we were forced to camp by the approach of a rapidly moving fog driven before a strong westerly wind. While camp was being prepared, it was discovered that a tin of kerosene on the front sledge had been punctured causing the loss of a gallon of fuel. Fortunately, we were well within our allowance, so the accident was not serious. Soon after tea our attention was drawn to a pattering on the tent like rain, caused by a fall of sago snow.

In the morning the weather was clearer, and we saw that it was impossible to reach Gaussberg by a direct route. The ice ahead was cleft and split in all directions, and, in places, vertical faces stood up to a height of one hundred feet. The floe was littered with hundreds of bergs, and in several localities there were black spots which resembled small rocks, but it was impossible to approach close enough to be certain. Retracing the way out of the broken ice, we steered in a south-westerly direction, just above the line of serac and creva.s.sed ice. The coast here trended to the south-west, forming the eastern side of Drygalski's Posadowsky Bay. The going was heavy, the surface being covered by a layer of frost-crystals deposited during the night. A fog came up again early in the afternoon and had quite surrounded us at camping time. During the day there were fine clouds of ice-crystals in the air, and at 8 P.M. a fog-bow was seen in the east.

Turning out in the morning we saw Gaussberg peeping over a ridge to the west, but were still prevented from steering directly towards it by the broken surface. When we had advanced ten miles, a heavy fog brought us to a halt at 5 P.M.

On Friday the 20th, in spite of a sticky surface, thirteen miles was covered on a west-south-west course. The ice-cap continued to be undulating but free of creva.s.ses. The alt.i.tude was between two thousand five hundred and three thousand feet.

In the morning, after travelling two miles, we came in sight of Gaussberg again and steered directly towards it. The surface was good with a downward grade. At five and a quarter miles a depot was made of the small sledge and most of the food, in expectation of a clear run to the mountain. Not far ahead, however, were two broken-backed ridges intersecting the course, and a detour had to be made to the south to cross them higher up.

Midsummer's day, December 22, was spent in the tent, a move being impossible on account of the high wind. In the afternoon we walked ahead a short distance and reconnoitred six or seven crumpled ridges. Though the barometer had been falling ominously for twenty-four hours, the bad weather did not continue.

Gaussberg was reached in the afternoon, after our track had pa.s.sed through seventeen miles of dangerous country. For the first few miles the surface consisted of a series of steep, buckled ice-ridges; later, it was snow-covered, but at times literally cut into a network of creva.s.ses.

The only approach to Gaussberg from the plateau is from the south. To the east and west there are magnificent ice-falls, the debris from which litters the floe for miles around.

December 24 and Christmas Day were devoted to examining the mountain.

Dovers made a long series of observations for longitude, lat.i.tude and magnetic variation, while Hoadley examined the rocks and took photographs.

On the southern side, the ice-cap abuts against this extinct volcano at an elevation of about four hundred feet above sea-level; the summit of the mountain rises another eight hundred feet. On the north, the rock descends to the floe. Gaussberg is pyramidal in shape, falling steeply, from a ridge at the summit. The sides are covered with a loose rubble of volcanic fragments, square yards of which commence to slide at the slightest disturbance. This renders climbing difficult and accounts for the large numbers of isolated blocks fringing the base.

At the summit two cairns were found, the bamboo poles which had previously marked them having blown over. Further examination revealed many other bamboos which had been used as marks, but no other record of the visit of the German expedition, ten years before, was met. Bird life was not plentiful, being limited to a few skuas, Wilson petrels and snow petrels; the latter nesting under slabs of rock. There were large quant.i.ties of moss where thaw-water had been running.

The ice and snow near the mountain showed evidences of marked thawing, and we had difficulty in finding a favourable spot for our camp.

Christmas Day was gloriously fine, with just sufficient wind to counteract the heat of the sun. At midday the Christmas "hamper" was opened, and it was not long before the only sign of the plum-pudding was the tin. In the afternoon we ascended the mountain and left a record in a cairn at the top. By the route followed, Gaussberg was two hundred and fifteen miles from "The Grottoes" but relay work had made the actual distance covered three hundred miles.

We had been away from home seven weeks, and, though there was sufficient food for an outward journey of another week, there was no indication that the country would change. Further, from the summit of Gaussberg one could see almost as far as could be marched in a week. Accordingly it was decided to commence our return on the 26th, making a course almost due east, thus cutting out numerous detours which had to be taken on the outward journey.

We left the mountain on December 26, pursuing a course to the south of our outward track so as to avoid some creva.s.sed ridges. Ascending steadily against a continuous headwind, we picked up the second sledge at midday on the 28th.

Next day all the gear was transferred to one sledge and a course made direct to the Helen Glacier; the other sledge being abandoned.

On December 31, after a day's blizzard, the surface was found to be covered with sastrugi of soft snow eighteen inches to two feet in depth.

In crossing a wide creva.s.se, the sledge became bogged in the soft snow of a drift which had a deceptive appearance of solidity. It took us ten minutes to extricate ourselves, and, after this, creva.s.ses were negotiated at a run.

A violent blizzard raged during the following day--the first of the New Year 1913. This proved to be a blessing, for it made the surface more crisp and firm. In the morning the sun was obscured and nothing was visible but the snow at our feet, so that steering was very difficult.

In the afternoon the sun broke through, a strong westerly wind sprang up and we moved along at a good pace, covering more than thirteen miles before camping.

On January 3 the track bordered on the edge of the plateau, the surface being almost level, rising gently towards the south.

After a violent blizzard of three days' duration, which confined us in the tent, we continued on the same course for four days, averaging about eleven miles each day. The surface was good, but a strong south-easter blew practically all the time and reduced our speed considerably.

At 10 A.M. on January 9, a fog-bank was observed in the east. This rapidly approached, and in fifteen minutes was quite close. There was now a splendid display of rings and arcs, caused apparently by minute ice-crystals which filled the air without obscuring the sun or sky.

First an arc of prismatic colours appeared in the east, and in a few seconds the sky seemed literally to be covered with other arcs. At first they seemed to be scattered indiscriminately, but after a short time several arcs joined and we could discern a symmetrical arrangement. The sun was surrounded by a ring, the lower portion of which was broken by an inverted arc; two other arcs were visible on either side. A large ring appeared encircling the zenith, intersecting the first and pa.s.sing through the sun. Two pairs of arcs were also seen, one pair in each ring. Excepting the arcs and ring about the zenith, which was grayish-white against the blue sky, the arcs showed prismatic colouring.

The display lasted ten minutes and ended with the disappearance of the ice-crystals.

[ILl.u.s.tRATION IN TEXT]

The diagram shows the arrangement of the arcs:

S = Sun. Z = Zenith.

At A, B, C, mock suns could be seen.

From our camp on the night of January 10, broken country could be seen ahead. To the north, open water was visible, and to the north-east the Shackleton Shelf, so that we were nearing home at last. Here, a heavy snowstorm delayed us for two and a half days, and it was not till the afternoon of January 13 that we were able to move ahead.

The next day was dull, the sun being quite obscured; and the only check upon the steering was the south-easterly wind. At midday the thermometer registered 35 degrees F. in the shade, and the surface became quite sticky. After tea we walked ahead for a couple of hundred yards to the summit of a ridge where the full extent of the Helen Glacier was laid before us. It was evident that our position was some miles north of the true course, but, considering the absence of steering marks and the constant overcast weather, we considered ourselves lucky in being so close to it.

The bad weather continued and snow fell during the following day. On the 16th the light was better, and we pushed into a strong wind which freshened to the force of a moderate gale before we had travelled two miles. Approaching a steep ascent we were compelled to camp. The morning brought an improvement, and the crossing of the Helen Glacier was commenced a mile or two above the outward course.