SIXTY YEARS ago, P. F. Kendall published a thorough interpretation of 'glacial retreat' in the Cleveland Hills,1 and the ideas contained in his two papers on the subject met with little opposition at that time.2 The anomalous valleys, described as overflow channels, were soon recognized in other areas and used as indicators of the former existence of proglacial lakes. In the succeeding thirty years, many studies in particular areas of this country3 used Kendall's methods of interpretation to elucidate the mode of recession of the ice margin during deglaciation. In the absence of any more recent research on the North York Moors, the sequence proposed by Kendall is still quoted in textbooks of geomorphology4, although some other areas which were originally considered in the light of Kendall's interpretation have been reexamined.5 The distribution of proglacial lakes in north-east Yorkshire as proposed by Kendall6 is illustrated in Figure I. Later workers suggested slight amendments to this interpretation. F. Elgee7 offered some new evidence concerning the glaciation and subsequent history of north Cleveland, but adhered closely to the ideas proposed by Kendall. J. T. Sewell8 proposed a minor elaboration for part of Eskdale, while work on the drifts of north Cleveland9 also followed the established interpretation. In I956, a paper on north-west Cleveland10 marked the beginning of renewed interest in the North York Moors and in the problems which they present. In this account, R. H. Best retained the conventional interpretation of overflow channels and the maximum limits of the last ice sheet as proposed by Kendall. However, the recognition of a more complete sequence of overflow channels led Best to invoke two glaciations to account for the channels present in north-west Cleveland. More recently, a brief account has been produced of the Stonegate valley11 and the landforms described, taken to indicate the former presence of stagnant ice. Finally, in a description of eastern Eskdale,12 the present author suggested that the ice was finally stagnant during deglaciation. This interpretation differs from that proposed in I902 in that Kendall envisaged deglaciation accompanied by the recession of an active ice margin towards the north-east. In eastern Eskdale as a whole, it is now thought that two distinct stages of meltwater drainage were followed by stagnant ice conditions. At the head of the Stonegate valley (Fig. 2), three minor stages of meltwater drainage can be recognized, but these do not necessarily correspond directly with the stages recognized elsewhere in eastern Eskdale. Reconstruction of the gradients of former ice sheets,13 using the gradients of the marginal drainage channels, suggests that the second major stage at Goathland corresponds to the three minor stages identified at the head of Stonegate. Although this interpretation differs from that proposed by Kendall for eastern Eskdale, it does not necessarily affect the interpretation of a proglacial lake in the remainder of Eskdale, at least during the early stages of deglaciation. During the later stages, when the ice in eastern Eskdale was fragmented,14 a proglacial lake could not have been