These studies constitute part of the geographical results of the Wordie Arctic Expedition, led by Mr. J. M. Wordie of St. John's College, Cambridge, to Melville Bay and the east coast of Baffin Land in the summer of 1934. It was found that some of the arctic soil forms had strong similarities to structures observed in the Pleistocene gravels of the Cambridge area, and that, treated together, these forms and structures would throw new light on the mechanics involved in their formation. The studies are divisible into three portions: frost cracking, polygonal structures, and solifluxion. I. Frost Cracking Travellers in northern and eastern Siberia report that the tundra, in places, is scarred and seamed by ramifying linear depressions flanked by low ridges, the depressions being marked out by a darker growth of mosses, generally interlacing to form roughly polygonal features many feet across. They have not, so far as is known, been excavated (Sukatchev 1911, Nikoroff 1928). Leffingwell (1919) observed frost cracks occurring in Alaska only in “muck”, though Smith (1909, p. 272), writing on the Nome deposits, noted a “series of ramifying streaks of a black peaty material cut in irregular directions across the layers of sand and gravel”, which streaks he attributed to frost cracking. Leffingwell (1919, p. 211) also noted “frequently straight lines leading across the gentle undulations of sandspits”, though he saw no polygonal forms in such places. However, his classic treatment of the theory of the formation of ground ice led him to generalize on