A morainal sequence in south-west Yukon Territory, Canada, records at least four major, successively less extensive glaciations from ice fields in the St. Elias Mountains south of the glaciated area.

The Nisling Moraine flanks the Klondike Plateau in a belt t km. wide to an altitude of 1,040 m., 12 km. north-east of Snag. The northernmost lobe of this moraine terminates at the junction of the Donjek and White Rivers, 120 km, from the nearest source of ice, Klutlan Glacier. 11 km. north-east of Snag, the prominent front of the Donjek Moraine lies 180 m. below the front of the Nisling Moraine. The northernmost lobe of the Donjek Moraine terminates 106 km. north of Klutlan Glacier and occupies the lower courses of canyons cut into the Nisling Moraine. The front of the Snag Moraine crosses the White River valley 210 m. below the front of the Donjek Moraine and 96 km. north of Klutlan Glacier. The Tchawsahmon Moraine, 38 km. north-west of Klutlan Glacier. consists of a series of concentric ridges, the oldest of which impounded Tchawsahmon Lake.

Provisional correlations suggest that the Nisling Moraine is pre-Illinoian; the Donjek, Illinoian; the Snag, pre-classical Wisconsin; and the Tchawsahmon, classical Wisconsin.

Meighen Island lies in the centre of the north coast of the Queen Elizabeth Islands and fronts on the Arctic Ocean. An ice cap of about 76 km.2 covers about one-tenth of the island. Its greatest thickness of 150 m. occurs under the summit, near the south end, which was 268 m. above sea-level in 1960. The northern half of the ice cap is less than 30 m. thick; and the total volume is of the order of 2,000 × 106 m.3.

Precipitation is low in the northern Queen Elizabeth Islands, and Meighen Island lies in an area where summer temperatures are lowest. In the winters of 1959–60, 1960–61 and 1961–62, the snow accumulation was 12.6, 18.2 and 14.1 cm. of water equivalent. Some snowfall remained on the higher part of the ice cap in the cold summer of 1961; but the ice cap diminished in volume in each year; by 36 × 106, 72 × 106, 22 × 106 and 91 × 106 m.3 in the 1959, 1960, 1961 and 1962 ablation seasons.

If the conditions of these four seasons were maintained the ice cap would disappear in about 100 yr. However, a radio-carbon dating of a saxifrage exposed by the retreat of the ice from a small nunatak near the northern edge gave a date of less than 100 yr., and it appears that the existence of the ice cap might be sensitively related to recent climatic change.

Careful surveys were made in 1959, 1960 and 1961 in an attempt to detect movement in the ice cap. Unequivocal evidence is not available from these surveys; but the stake network has been maintained and another survey has recently been completed.

The history of glaciological studies in the Svartisen area is outlined, and observations made by earlier workers are summarized. The present position of the ice limit in the area is indicated and is compared with those of the past Evidence of extensive, and almost continuous, retreat of the glaciers is presented, and emphasis is placed on the rapidity of the changes since the 1930’s. In the present century, the formation of marginal lakes and the detachment of glacier tongues have characterized the decay of many of the Svartisen outlets.

The upper 180 ft. (55 m.) of Lake Vanda in Wright Valley, south Victoria Land, is essentially potable, whereas that part below 200 ft. (61 m.) is more than three times as saline as sea-water. The salinity below 200 ft. (61 in.) resulted from the evaporation and freezing, mainly during interglacial (Loop-Trilogy) time, of a larger, less saline body of water.

An alluvial fan in Wright Valley has been dated as interglacial on the basis of (1) ice-marginal channels formed in Loop time that cut across it, and (2) the fjord-like longitudinal cross-section of the valley formed by the ice of the oldest glaciation or glaciations.

Fossils found in till and glacio-fluvial deposits in the McMurdo Sound region, south Victoria Land, date from both early and late Pleistocene interglacial time.

An ocean-bottom core sample obtained in the Ross Sea contained interglacial material. The presence in the sample of glacial marine sediments deposited in interglacial time suggests that the Antarctic Ice Sheet maintained itself throughout the Pleistocene.

Sub-surface efflorescences consisting of a layer of pure salts as much as 3 in, (7.6 cm.) thick are found 2−5 in. (5−13 cm.) below the surface in Loop deposits in Wright Valley. The absence of similar thick occurrences in Trilogy deposits indicates that the efflorescences are, in part at least, interglacial.

Interglacial cinder cones and lava flows are found in the McMurdo Sound area, and some of the widespread scoria present in moraines, glacio-fluvial deposits and beaches dates from interglacial time.

Density-depth curves constructed from seismic travel-time data observed at 25 locations, show a large variation over an essentially isothermal area (T = −22±2.6° C). This is correlated with the wide range of mean annual accumulation values. (A = 20 to 50g./cm.2yr). Quantitative estimates of mean annual accumulation can be obtained from seismic refraction data in this area by the expression A = 23.5+0.049(3500−V200) where V200 is the compressional wave velocity (m./sec.) observed at 200 m. from the shot point. The density-depth curves can be used to give estimates of accumulation by the expression:

where ρ40 is the density (g./cm.3) at 40 m. depth. The internal consistency of these methods is indicated by the standard deviation of ±2.9./cm.2 yr. for comparisons of accumulation estimates. The theoretical basis for these results is discussed.

Art examination of bullet crystals in precipitation at the South Pole indicates that combinations of bullets originate as primary growth structures and that individual bullets are formed as a result of the disintegration of these primary growth forms rather than by independent crystallization of pyramidally terminated columns.

The seasonal distribution of snow at the South Pole and its relationship to stratigraphy was investigated to pits dug beside a number of four-year-old accumulation stakes. Results show that conventional stratigraphic methods yield thoroughly reliable values of accumulation rates. Even hiatuses in accumulation can be identified from the intensity of sublimation of layers of depth hoar in the stratigraphic section. Such hiatuses are due almost invariably to the prolonged absence of accumulation rather than to widespread scouring of pre-existing layers of snow. The bulk of the year’s accumulation is deposited as dunes during winter. The majority of dunes are subsequently transformed into linear sastrugi by wind with the result that the amplitude of surface relief observed at the end of winter frequently exceeds the average thickness of snow accumulated annually. Such gross relief does not persist to the end of summer, however. Instead the dunes and sastrugi arc gradually worn down by a process of sublimation-deflation. It is this leveling of the surface relief in summer and the resultant redistribution of snow more uniformly over the surface that are believed to be the significant factors in the formation of the systematic stratigraphy observed in pits at the South Pole.

Early photographs of a cirque glacier complex in the Arrigetch Peaks of the south-central Brooks Range date from 1911, and are the second oldest known set of glacier photographs from northern Alaska. Matching photographs and supplementary observations made in 1962 indicate that changes over the past 51 yr. in this area include general glacial recession and thinning, the emergence of pronounced trimlines and partial melting of ice cores beneath recent moraines.

Renewed post-hypsithermal glacial activity in the Arrigetch Peaks area attained two maxima in the recent past. The younger moraines lie close to glacier positions shown in the 1911 photographs, contain ice cores which extend practically to their surfaces and were probably formed in the latter part of the nineteenth century. More weathered deposits, underlain by ice cores at relatively greater depths, were formed at some unknown earlier date. The moraines are correlated with the two-substage recent Fan Mountain advance of northern Alaska and may correspond to dated mid-eighteenth century and middle to late nineteenth century advances of alpine glaciers in the north Pacific coastal mountains of North America.

Gravity measurements have been used to determine ice thicknesses across the western part of the Devon Island ice cap in the Canadian Arctic. A detailed profile of the ice-cap edge and a profile across an adjoining glacier are also given. The ice cap has been found to have a largely rock core with ice thicknesses generally less than 500 m. A deep valley has been found in the bedrock beneath the ice cap some 15 km. from the start of a draining glacier. The measured depths on the ice cap should be within 15 per cent and those on the glacier within 20 per cent of the true values.

During the summer of 1963. a drift-veneered mass of ice suddenly emerged at the surface of “Miller Lake”, art ice-walled lake at the terminus of the Martin River Glacier, Alaska. Subsequent fathometer traverses, together with the fact that the ice was not at the pressure melting point, revealed that it had been derived from a hole approximately 40 m. below the surface of the lake. There are numerous large debris-covered icebergs in the lake and presumably at least some are formed in this manner. Others are derived through differential ablation of projections of the glacier into the lake. Only the small icebergs are formed through calving. The interpretation of the mechanism of release of ice from the bottom of the lake, and the observation and interpretation of a rising ice-cored island in the same lake suggest that the ice on the bottom is deforming plastically and that it is undergoing compressive flow. Much of the movement may be along shear planes actually found in icebergs derived from the bottom of the lake.

If a bore hole of circular cross-section is made in an ice sheet or glacier, will it remain circular, or will it perhaps by some process of instability tend to depart from the circular shape? An argument is given to show that the circular shape is stable.