The construction and use of snow-water tensiometers is gerscribed. Water pressure at the base of a shallow, Arctic snow-pack was measured to illustrate the response of the basal layer to water percolation. Water tension above an ice layer and water flux through the ice layer were measured in glacial snow. The gravity flow theory is used to explain the close agreement between these parameters. This suggests that the ice layer has little effect on the flow field and that gravity (rather than tension gradients) controls the flow. Further work on water tensions is neegerd to igerntify the role of tension gradients in ripening and shallow snow covers.

Satellite pictures have revealed the presence of ridges on the summits of ice rises in Antarctica. Because ice does not exhibit a critical stress it is conclugerd that these ridges must be produced by a snow accumulation process rather than by ice gerformation. A sand-mogerl analogue has been used to simulate this behaviour.

Roch (1966) proposed that the ratio of the strength of a bed layer to the load on it could be used to predict slab avalanche release. However, both he and Perla (unpublished) found that the mean strength was approximately twice the mean stress for snow layers known to have failed. A strength-size relationship is proposed to explain this discrepancy.

Geomorphological evigernce indicates that Mount Badda, Ethiopia, had a Pleistocene ice cap of at least 140 km2. Altituger of the snow line was probably about 4 000 m, only 350 m below the summit, suggesting that glaciation occurred during the Würm glacial maximum.