Under-currents of the sea take them not unfrequently in directions contrary to the drift of the field-ice, which depends only on upper-currents; and abnormal winds may sometimes carry them out to seas where they have been seldom or never seen.[4] This appears to be the case even with those met with on the north-west coast of Novaya Zemlya. On the other hand, they have never been seen on the coasts of Siberia, which have no glaciers. 27. The constant displacement of the centre of gravity of an iceberg, resulting from the unsymmetrical decrease of its form, causes its periodical oversetting; and the different temperature of the internal and external ice is the principal cause of its rending asunder with a noise like thunder; a process which occurs generally in the height of summer. CHAPTER II. NAVIGATION IN THE FROZEN OCEAN. 1. Although it be impossible to give any one, who has not with his own eyes seen the Arctic Sea, a perfectly clear conception of its character, the phenomena described in the preceding chapter are sufficient to indicate the difficulties and dangers to which its navigation is necessarily exposed. And to these difficulties and dangers, formidable enough in themselves, are often added the evil influences of preconceived theories and exaggerated expectations, usually followed by bitter disillusions. The calm judgment, which, to all the bold plans of navigation within the Polar basin, opposes distrust in their feasibility, while it points to the hundred expeditions which have at last returned home after penetrating but a little way into the frozen sea, is an attainment of slow growth. Years, too, must be devoted to the theoretical study of the Polar question, to the examination of all that predecessors have experienced and recorded. But this study is very important to Polar navigators; for the discoveries which they too readily regard as exclusively their own prove sometimes to have been made centuries before them. 2. A most essential element of success is the choice of a favourable ice year; and the commander of an expedition must possess sufficient self-control to return, as soon as he becomes convinced of the existence of conditions unfavourable for navigation. It is better to repeat the same attempt on a second or even a third summer, than with conscious impotence to fight against the supremacy of the ice.