While Nordenskiöld was at Mossel Bay he attempted a journey to the north, but was stopped by the ice at Seven Islands, and returned round North East Land. It took him five days to pass across the twenty-three miles between Phipps Island and Cape Platen over pyramids of angular ice up to thirty feet high. On the coast, which he found extending, as Leigh Smith had reported, much further to the east than was shown on the charts, he met with the inland ice ending in precipices from two thousand to three thousand feet high. Ascending this ice they had scarcely gone a quarter of a mile before one of the men disappeared at a place where the surface was level, and so instantaneously that he could not even give a cry for help. When they looked into the hole they found him hanging on to the drag-line, to which he was fastened with reindeer harness, over a deep abyss. Had his arms slipped out of the harness, a single belt, he would have been lost. Along the level surface every puff of wind drove a stream of fine snow-dust, which, from the ease with which it penetrated everywhere, was as the fine sand of the desert to the travellers in the Sahara. By means of this fine snow-dust, steadily driven forward by the wind, the upper part of the glacier—which did not consist of ice, but of hard packed blinding white snow—was glazed and polished so that it seemed to be a faultless, spotless floor of white marble, or rather a white satin carpet. Examination showed that the snow, at a depth of four to six feet, passed into ice, being changed first into a stratum of ice crystals, partly large and perfect, then to a crystalline mass of ice, and finally to hard glacier ice, in which could still be observed numerous air cavities compressed by the overlying weight; and, when, as the surface thaws, the pressure of the enclosed air exceeds that of the superincumbent weight, these cavities break up with the peculiar cracking sound heard in summer from the glacier ice that floats about in the fjords. Occasionally broad channels were crossed, of which the only way to ascertain the depth was to lower a man into them, and frequently he had to be hoisted up again without having reached the bottom; such danger areas causing so circuitous a route that much progress was impossible.