This investigation began as a study of fossil patterned ground in East Anglia, but was broadened in scope when it became obvious that existing published reports of comparable active patterned ground only allowed very limited interpretation of the patterns of East Anglia. The final investigation included field observations on active patterns in Arctic North America and Europe, especially the Seward Peninsula, Alaska and Finnmark, Norway. Much of the evidence was visual and not easily expressed in a quantitative form. A large number of diagrams and plates are presented as an essential supplement to the inevitably subjective description. The review of literature on patterned ground was largely based upon the work of Washburn(1956), but information relevant to the mechanism of origin was drawn from a wide variety or sources. The descriptive classification suggested by Washburn was reluctantly abandoned in the light of the field evidence, and a new synthesis of descriptive terms is proposed. When possible existing terms were used, taking care to avoid distorting existing definitions. The new classification is based upon surface form, grouping and marking. Proposed terms for form are 'equiform'(replacing the ambiguous terms 'circle', 'polygon' and 'net'), 'stripe', 'elongate'(intermediate between stripes and equiforms) and 'step'. The grouping of patterns is divided into 'isolate', 'contiguous' and 'grouped'(the latter for doubtful intermediate cases). The terms 'relief', 'stone' and I 'vegetation', or combinations, are proposed for the description of pattern marking. The terms 'sorted' and 'non sorted' are rejected because of the contradictions arising from the dual descriptive and genetic use of these terms, An 'other variations' category is added to both form and marking descriptive categories to allow for occasional occurrences not covered by the proposed terms. Permafrost, the heat flux in ground subjected to freezing temperatures, possible ways in which pressure in the ground is produced by freezing, frost susceptible materials, and movements in patterned ground are all discussed as major factors affecting patterns. The field evidence does not differ markedly from the type of evidence produced by previous workers, though perhaps the size and number of excavations is notable, Air photographs were used extensive1y. Important observations included gradations between almost all forms, groupings and markings of patterns; clear evidence of circulatory mover.tent in pattern sections; and e1ongated patterns apparently parallel to lines of drainage rather than parallel to the maximum slope Discussion of the field data includes the evaluation of evidence on the development and perpetuation of patterned ground once the pattern is established; direct and indirect evidence of mechanisms of movement; evidence of actual movements; the initiation of the pattern of patterned ground; the utility of patterned ground as a palaeoclimatic indicator; and the rate of pattern development. An important conclusion is that the pattern forming processes may all be incorporated into two models for pattern development - a radial movement model and a circulatory movement model, the former perhaps being a special modification of the latter. Additionally each of the three main field areas is treated independently for presentation of evidence and deductions. The fossil patterns of East Anglia are thought to indicate former continuous permafrost conditions. A major part of the concluding section is entitled "Towards a more rational understanding of patterned ground", which summarises many of the writers ideas.