The existence of relatively high values of the calculated bed shear stress on the traverse A-142 to D-153, and of the greater rates of flow implied, prompts me to suggest the possible causes of this condition.

Most of the exceptionally low values of τ on the other traverses appear to be associated with the line of the traverse being appreciably different from the line of greatest slope, but even neglecting these low values the A-142 to D-153 values are relatively high. Some variation in τ on any radial traverse might be expected. Sliding on the bed is not likely near the centre of an ice cap, but it must occur towards the edge since it is the only way in which rock debris can be brought to the surface of the ice under these conditions. Thus one might anticipate some increase in τ from the centre towards the zone where shearing on the bed has involved the entrainment of rock debris, because the shearing resistance of rock debris on rock is pressure-conditioned. Close to the edge, where the flow is compressional, and overthrusting occurs over stagnant ice, the value of τ might decrease. The values of τ in traverses A–B and A–C suggest this type of variation.

The shape of the whole ice cap has one particular feature that is perhaps significant in respect of the high values of τ along traverse A-142 to D-153. The ice cap is highest along a ridge that lies roughly parallel to, but curiously closer to, its north-east side (see Fig. 2 in Part 1 of this series of papersReference Baird ¹ ), and in the south-eastern lobe at least, the bed profile does not account for this displaced ridge.

The causes may be directly climatological, but I am inclined to discount this view. No large variations in the quite small total accumulation were found, and the current marginal accretion by perennial snowdrifts mentioned by Goldthwait,Reference Goldthwait ² particularly along the north-east side, would tend to flatten the slope and decrease τ It has been shownReference Ward and Orvig ³ that much of the ablation energy is discharged to the lakes adjoining the ice cap. At present, however, only a small part of the north-east side forms cliffs fringing these lakes. This is unlikely to cause a significantly greater discharge of ice, except in the immediate vicinity of the lakes where underwater ablation might make sufficient difference. There is plenty of evidence, however, that the lakes on the north-east side have been much deeper and more extensive in recent times. A noticeable strandline, some 200 ft. (61 m.) above the present level of Generator Lake, is visible about one-third of the way up the hillside in the background of Fig. 3 in Part 11 of this series of papers.Reference Ward ⁴ Goldthwait has mentioned other strandlines to me, and a study of the R.C.A.F. aerial photographs of the larger lakes near the north end of the ice cap reveals the existence of scores of miles of high strandlines surrounding the present lake limits. In the past, almost the whole of the north-east side of the ice cap must have impounded water with a line of fringing ice cliffs. The effect of a general lowering of the water level of these formerly extensive lakes is to remove a large stabilizing force from the north-east side of the ice cap. This would cause a noticeable increase in the rate of flow of the ice towards the lakes that is likely to persist for a long time. The action is similar to the incidence of landslips following the lowering of lake levels.

In conclusion, the particular instance mentioned above calls for some consideration to be given to the numerous records of extinct glacial lakes and the substantial effects that their disappearance may have had on the motion of adjoining ice fronts.