Stresses occur in the uppermost 10 m of a glacier as a result of temperature fluctuations at the surface. A model is set up of a typical year's surface temperature variation, and the progress of temperature waves through the glacier is calculated using Fourier theory of heat conduction. Short-period fluctuations are rapidly attenuated, and at 10 m depth the annual cycle is reduced to 5% of its surface amplitude. As the temperature of the ice varies it undergoes small volume changes; stresses are calculated on the assumption that any tendency of the ice to expand or contract laterally results in the creation of just enough stress to cause the ice to remain unstrained. It is found that in the top 2 or 3 m stresses of thermal origin are generally in excess of those due to gross deformation or overburden pressure. For the case of high-density ice Glen's flow law is used, and conditions are found to be favourable for the formation of surface rumples of wavelength about 10 m. For the case of firm or snow a Newtonian flow law is assumed, and it is found that under cold conditions fracture under tension can occur. Cracks of thermal origin may be responsible for the initial formation of crevasses, and they also provide an explanation for background noise encountered when seismic shooting at low temperatures. Calculations are made of the strain-rate field surrounding a crack and it is found that thermal effects can lead to appreciable Strain-rate anomalies for strain-rate measurements near cracks. The magnitude of the effect is easily sufficient to account for anomalous fluctuating strain-rates found by workers using wire strainmeters on the Barnes Ice Cap.

Laboratory studies suggest that neither bubbles nor dirt particles migrate rapidly enough in glacier ice to be responsible for the alternating layers of bubbly and clear ice or dirty and clean ice which constitute foliation. We therefore suggest that these variations in bubble or dirt content are inherited from primary inhomogeneities such as may occur in sedimentary stratification in the accumulation region, in crevasse fillings, or during debris entrainment at the base of the glacier: the appearance of these inhomo-geneities is later modified by strain during flow to produce foliation. We consider six types of inhomogeneity, or components of foliation, and show that, at the very large total strains expected in glaciers, all are eventually flattened, stretched out, and rotated to form a layered structure roughly perpendicular to the direction of maximum total shortening. Most characteristics of observed foliation can be explained by this hypothesis. For example, in the marginal zones of polar ice sheets the rapid decrease in dip of foliation with depth and with distance up-glacier from the margin can he explained by a model in which the foliation is assumed to be nearly parallel to the base of the glacier some distance from the margin, and is deformed passively with the ice thereafter. However, some observations of cross-cutting foliations may require localized inhomogeneous shear parallel to the "new" foliation.

The local variability in surface mass balance (net snow accumulation) up-glacier from "Byrd" station, Antarctica, is due to the combined effects of year-to-year "climate" variations and of the surface microrelief clue to snow drifts and sastrugi. These variabilities are consistent with the variability in surface mass balance obtained from core stratigraphy (Gow, 1968), and are used in a discussion of the difficulties encountered with the deep "Byrd" station core in detecting annual layering by the stable oxygen-isotope ratio and the microparticle concentration techniques. The recognition of annual layers by these techniques requires that the snows of certain seasons be present in the measured section, but near "Byrd" station the microrclief is such that summer snow layers are not horizontally continuous and may be absent from a given section. At other sites on ice sheets, where the microrelief is less (less wind activity) or where the surface mass balance is larger, or both, less difficulty is anticipated in using the stable oxygen-isotope ratio and micro-particle-concentration techniques to identify annual layers.

Temperatures from 10 m bore holes have been analysed to determine the spatial variation of mean annual surface temperature over the Antarctic Peninsula. In general there is a decrease in temperature of 0.84 deg per degree latitude southwards combined with an altitude lapse rate of 0.68 deg per 100 m. There is a sharp divide between the continental type climate of the east coast and a maritime climate of the central and western regions. Temperatures are approximately 7 deg lower along the east coast compared with those at sites of a similar altitude and latitude on the west.

The basal ice layer of Alpine glaciers, the thickness of which can reach more than one metre, generally appears as a sequence composed of layers of bubble-poor ice 0.5-3 cm thick and dirt layers containing dispersed rock fragments. Locally, debris-free bubbly ice is also present in the sequence. This basal layer exists on the lee side of rock protuberances or along the marginal part of the glacier. Three geochemical characteristics shed some light on its origin:

• (a) The high calcium content of basal ice sliding over limestones from the north side of Glacier do Tsanfleuron can be successfully predicted from considerations of the phase relations in regelating subglacial solutions in approximate equilibrium with subglacial calcite deposits.

• (b) The (Na + K)/(Ca+Mg) ratio of bubble-poor basal ice from Glacier de Tsijiore Nouve which flows over gneissic rocks is significantly higher than this ratio in other samples taken in the same environment (glacier ice, melt water). This may be explained by refreezing of squeezed water, the ionic composition of which is governed by selective diffusion of ions during squeezing and/or by the water being forced through mud layers as demonstrated by us at the base of Glacier d'Argentière (Souchez and Lorrain. 1975).

• (c) The Lead-210 activity per kg of basal ice from the south side of Glacier de Tsanfleuron is greater than that of the glacier ice immediately above, which indicates that it is younger despite percolation effects. This can be explained if recent ice accretion has occurred at the base of relatively old glacier ice. Accretion of blowing snow has been found to occur under the southern margin of the glacier.

A theoretical density distribution of the Dumber of bonds per grain as a function of the mean number of bonds per grain is derived from the assumption of randomness and isotropy of grain and grain-bond location and orientation. The knowledge of the theoretical density distribution allows the determination of the effective distribution parameters from section planes or thin sections. The concept of the fundamental unit to describe the strength of snow is introduced. Structural parameters developed on the basis of the fundamental units show improved correlations with the tensile strength.

The basic features of the statistical models to describe brittle and ductile strength of snow are outlined and discussed. The concept of the "fundamental unit" is introduced. The models are applied to estimate the development of slab stability during heavy snowfalls and to simulate measurements of the tensile strength of snow using the centrifugal method as a function of the load rate. The models developed in this paper are compared with Sommerfeld’s applications of statistical methods to estimate the strength of snow in the starting zone of a slab avalanche.

Dielectric properties of ice microcrystals have been studied in the frequency range 3 x 105 to 10-2 Hz. Two separated dispersions are revealed : the Debye dispersion of ice and another dispersion at very low frequencies. Both dispersions present an ageing effect, with decreasing activation energy for the Debye dispersion and with increasing activation energy for the second one. It is shown in this paper that the sum of these activation energies retains a constant value of (1.05±0.02) eV.

The aim of the paper is to analyse landscapes of glacial erosion associated with the Laurentide ice sheet at its maximum and to relate them lo the three main variables affecting glacial erosion, namely former basal thermal regime of the ice sheet, the topography of the bed, and the geology of the bed. The key to the analysis is the comparison of the distribution of landscape types with the simulated pattern of the basal thermal regime of the former ice sheet.

Landscapes of areal scouring are found to be associated with zones of basal melting and occur beneath much of the former ice-sheet centre and in those places where the topography favoured converging ice flow. The landscape type may also have formed beneath cold-based ice when it was carrying debris inherited from an up-stream zone of regelation. Areas with little or no sign of glacial erosion occur primarily in the north in the Queen Elizabeth Islands but they also occur on uplands associated with diverging ice flow; they coincide with areas calculated to have been covered by cold-based ice devoid of debris. Landscapes of selective linear erosion are common on uplands near the eastern periphery of the ice sheet. In these situations, pre-existing valleys channelled ice flow and created a situation where there was warm-based ice over the valleys and cold-based protective ice over the intervening plateaux. Variations in the permeability of the bedrock base have modified the landscape pattern, mainly in those areas where there was a change from one basal thermal regime to another. In general, permeable rocks tend to have experienced less erosion than impermeable rocks.

Using lake-basin density as an indication of the intensity of glacial erosion, a zone of maximum erosion is identified and this forms a ring between the centre of the former ice sheet and its periphery. This ring coincides with a zone where melt water from the ice-sheet centre is calculated to have frozen on to the bottom of the ice sheet. This regelation incorporated basal debris into the ice, forming a basal layer 20-50 m thick and afforded an efficient means of debris evacuation.

A conceptual model is developed and hangs round the following postulates:

• (1) Landscapes of glacial erosion are related primarily to the basal thermal regime of the ice sheet.

• (2) Landscapes of glacial erosion are equilibrium forms related to maximum glacial conditions. This implies that at some stage in the Pleistocene the Laurentide ice sheet was in a stable maximum condition for a long period of time.

• (3) Mechanisms allowing evacuation of debris rather than those of abrasion or fracture may be the most important in influencing the amount of erosion achieved by an ice sheet.

• (4) Cold-based ice may accomplish erosion if it contains debris.

Moraines deposited by the outermost ice advance across Judge Daly Promontory, northeastern Ellesmere Island, reflect thin, topographically controlled ice lobes extending to sea-level. The termini of two ice lobes were investigated and both produced ice shelves where they flowed into isostatically depressed embayments along western Kennedy Channel. Morphological evidence for these ice shelves occurs at the entrance to these valleys where steeply descending lateral moraines become abruptly horizontal for 2 km. In addition, both the horizontal moraines and associated pro-glacial terraces are fossiliferous down-valley from the apparent grounding line. Based on the differences in elevation between the horizontal moraines and the valley bottoms, the two ice shelves had estimated thicknesses of c. 110 and 150 m. A proglacial outwash terrace at 175 m a.s.l. is considered to represent the approximate relative sea-level during the formation and break-up of the ice shelves. This relative sea-level is consistent with the water depths required to float the calculated ice thicknesses in both valleys. Associated with these ice margins are finite 14C dates of 28 000-30 000 B.P. and amino-acid age estimates of >35 000 B.P. The importance and likelihood of additional past ice shelves in the Canadian High Arctic is discussed.

In 1899 P. A. Tutkovskiy published a theory of loess formation which depended on the presence of large continental glaciers. Unfortunately there was no glacial requirement in the theories of Berg and Richthofen and these have survived better than that of Tutkovskiy with the result that the close relationship between glacial action and loess formation is sometimes overlooked.

The report describes an experiment to evaluate the use of geophysical (Cambridge-type) wire strainmeters for the rapid acquisition of strain-rate data and to compare strains on a large ice mass over gauge distances of 5 m, 50 m, and 1 km.

Three continuously recording wire strainmeters were installed at the centre of two separate arrays of strain poles 10.6 km and 19.5 km from the ice divide on the Barnes Ice Cap. Data was collected between 24 April and 15 May 1976. The 1 km strain arrays had previously been measured in 1974 and 1975. The results show good agreement between the strainmeter data and the larger strain arrays at the 10.6 km site but differ at the 19.5 km site. When the daily means are calculated for the strainmeters at the 19.5 km site, the strain-rates show a possible periodic variation with an apparent period ofabout 11 d. Since there appears to be no direct correlation between the strainmeter signal and either temperature or pressure, the result is assumed to represent real varying strain within the ice.

An investigation into the feasibility of using a strain rosette to measure the principal strains and to locate the principal axes for a propagating strain wave is carried out for the particular application of a flexural-gravity wave in sea ice. It is found that the separation of the instruments in the rosette is extremely critical and, for a physically realizable rosette, errors are unavoidable if strain is to be monitored continuously. An alternative approach is proposed employing frequency-domain analysis and in particular a running-power spectral density. The method is demonstrated with data obtained from strainmeters on fast ice in Notre Dame Bay, Newfoundland. Two distinct wave components are found to be present of period 6 s and 13s. and it is shown with 99.9% confidence that they are propagating in different directions.

The design, construction, and testing of a portable constant strain-rate testing machine for determining the mechanical behavior of avalanche snow is described. The machine is intended for use in determining the stress strain time behavior of low-density natural snows in the field. A technique for making direct measurements of strain in the snow sample is described and stress-strain curves are presented for strain-rates ranging from 0.5 to 5.0 x 10-5 s-1. The densities of the snow samples tested range from 186 to 335 kg m-3 Ultimate-strength data and relaxation curves arc also presented.