Search

The implications of subdividing the lithosphere into upper elastic and lower viscoelastic layers are investigated by finite-element analysis. Application of uniform horizontal boundary stresses at the ends of a lithospheric plate leads to amplification of the stress in the elastic layer by a factor about equal to the ratio of lithospheric to elastic layer thicknesses, and the visco-elastic layer becomes nearly stress-free except near its ends. The time constant for approach to equilibrium is proportional to viscosity, being for our model 0.21 My for 10 23 N s m −2 , and there is some accompanying flexure of the lithosphere. Local variation in the thickness of the elastic layer causes inverse variation in the stress, in part explaining the stability of shield regions and the tectonic activity of hot plateau uplift regions. It is shown that stress amplification also occurs where the stresses arise from body forces such as differential loading and isostatic compensation across continental margins.

Summary. The deep structure of the Faeroe–Shetland Channel has been investigated as part of the North Atlantic Seismic Project. Shot lines were fired along and across the axis of the Channel, with recording stations both at sea and on adjacent land areas. At 61°N, 1.7 km of Tertiary sediments overlies a 3.9–4.5 km s-1 basement interpreted as the top of early Tertiary volcanics. A main 6.0–6.6 km s-1 crustal refractor interpreted as old oceanic crust occurs at about 9 km depth. The Moho (8.0 ° 0.2 km s-1) is at about 15–17 km depth. There is evidence that Pn may be anisotropic beneath the Faeroe–Shetland Channel. Arrivals recorded at land stations show characteristics best explained by scattering at an intervening boundary which may be the continent–ocean crustal contact or the edge of the volcanics. The Moho delay times at the shot points, determined by time-term analysis, show considerable variation along the axis of the Channel. They correlate with the basement topography, and the greatest delays occur over the buried extension of the Faeroe Ridge at about 60° 15′N, where they are nearly 1 s more than the delays at 61°N after correction for the sediments. The large delays are attributed to thickening of the early Tertiary volcanic layer with isostatic downsagging of the underlying crust and uppermost mantle in response to the load, rather than to thickening of the main crustal ayer. The new evidence is consistent with deeply buried oceanic crust beneath the Faeroe–Shetland Channel, forming a northern extension of Rockall Trough. The seabed morphology has been grossly modified by the thick and laterally variable pile of early Tertiary volcanic rocks which swamped the region, accounting for the anomalous shallow bathymetry, the transverse ridges and the present narrowness of the Channel.

The linear negative magnetic anomaly which stretches continuously for 110 km across the North Minch, extending slightly west of north from Loch Ewe, is interpreted in terms of an unusually wide dyke which does not reach the surface. Fourteen profiles taken from the aeromagnetic map have been interpreted using non-linear optimization, giving an excellent fit to a dyke having an average and fairly constant width of c . 1.1km. The depth to the upper surface of the interpreted dyke varies between 0.3 km and over 2 km. It is reversely magnetized in a direction consistent with a Tertiary or possibly earlier origin. The average magnetization is estimated to be 0.51 Am −1 . A similar magnetic feature on the shelf to the north-west of the Butt of Lewis is also interpreted in terms of a reversely magnetized dyke, having a poorly-resolved width of about 0.8 km, a depth of about 0.7 km and a magnetization of 1.6Am −1 . On seismic evidence, the Minch dyke appears to be post-Jurassic in age, and an early Tertiary age seems most likely for both dykes. It is suggested that they may be related either to the postulated large igneous pile of early Tertiary age which forms the Wyville-Thomson Ridge, or to the Skye centre. They appear to represent early Tertiary extension of the crust which is taken up by more numerous narrow dykes further south.

The sources of lithospheric stress and their distinctive features are briefly reviewed. It is suggested that there are two main categories of lithospheric stress: renewable stress which persists despite continuing stress relaxation and non-renewable stress which can be dissipated by relief of the initial strain. The two most important types of renewable stress arise from plate boundary forces and from isostatically compensated loads. Non-renewable stress systems include bending stresses, membrane stresses and thermal stresses. An important phenomenon generating large stresses at shallow depth is stress amplification caused by lower lithospheric creep. This applies to renewable stresses but not to the non-renewable type. It is suggested that only renewable stresses contribute significantly to tectonic activity. However, bending and thermal stresses are locally important in subducting lithosphere.

Passive margins form by continental splitting which may follow an early pre-split graben stage of development. Such margins are divisible into rifted and sheared types. After formation the margins develop by predominantly vertical tectonics as the intervening ocean widens by seafloor spreading. The plate splitting mechanism is not yet understood, but evidence from aseismic ridges and associated continental volcanism is relevant. Location of the original continent-ocean contact is also problematical in many regions, especially where quiet magnetic zones occur above crust of uncertain status. Rifted margins are notable for great subsidence resulting from early graben formation and later flexural subsidence of the shelf and slope regions. Graben formation is attributed to wedge subsidence of the upper crust in response to crustal stretching, possibly associated with doming before splitting. Four factors have been recognized as contributing to flexural subsidence: gravity loading, thermal subsidence of the adjacent oceanic lithosphere, possible thermal subsidence of the continental lithosphere following heating and erosion at the time of break-up, and thinning of the continental crust by seaward creep of lower crustal material. The relative importance of these four mechanisms needs clarification.

Summary P-waves travelling in the lower crustal layer and possibly the topmost mantle beneath the Iceland-Faeroe Ridge are observed to travel through the crust of the adjoining Faeroe Block with a velocity not significantly greater than that of Pg. This result supports the earlier inference that a true continental margin of anomalous type separates the thick Icelandic type of oceanic crust beneath the Ridge from continental crust beneath the Block, the phase conversion taking place at the margin. Converted phases of the type observed may have potential for investigating other continental margins.

The East Greenland continental margin is inferred to have formed in three stages during the Tertiary evolution of the adjacent North Atlantic Ocean. The East Greenland basaltic volcanism occurred over a period of 2–3 Ma, which was just prior to the probable date of onset of seafloor spreading in the Early Eocene shortly before magnetic anomaly 24B (52 Ma). The associated coastal flexure is interpreted as marking the edge of a zone of stretched and thinned continental crust beneath the inner Denmark Strait shelf and Jan Mayen microcontinent; this represents a short-lived rifting phase just prior to break-up. According to the evolutionary model, the initial break-up produced the margins south of the Faeroe Transform Fault (68.5°N) and north of the Jan Mayen Fracture Zone (72°N), but in the intervening region the microcontinent remained attached to East Greenland. After break-up, the outer shelf of Denmark Strait formed from anomalously thick oceanic crust complementary to the Iceland-Faeroe Ridge. The Blosseville Kyst margin (68.5°N to 70°N) was initiated after anomaly 20 time (42 Ma) as the Jan Mayen microcontinent started to separate from Greenland with spreading on both sides. The new split occupied the line of the pre-existing coastal dyke swarm and flexure. The Liverpool Land margin (70°N to 72°N) was also initiated at this time, but the main separation of the northern part of the microcontinent only commenced after anomaly 7 time (26 Ma) when a single spreading axis was relocated west of the microcontinent. It is suggested that the Blosseville lavas subsided during deposition by loading and slight crustal stretching. Doming occurred near Kangerdlugssuaq at the time of break-up. Otherwise the topography of the lavas remained subdued until the relocation of the spreading axis at 26 Ma, after which the main regional uplift took place.

Summary. The continent-ocean transition near Hatton Bank was studied using a dense grid of single-ship and two-ship multichannel seismic (mcs) profiles. Extensive oceanward dipping reflectors in a sequence of igneous rocks are developed in the upper crust across the entire margin. At the landward (shallowest) end the dipping reflectors overlie continental crust, while at the seaward end they are formed above oceanic crust. Beneath the central and lower part of the margin is a mid-crustal layer approximately 5 km thick that could be either stretched and thinned continental crust or maybe newly formed igneous crust generated at the same time as the dipping reflector sequence. Beneath this mid-crustal layer and above a well defined seismic Moho which rises from 27 km (continental end) to 15 km (oceanic end) across the margin, the present lower crust comprises a 10-15 km thick lens of material with a seismic velocity of 7.3 to 7.4 kds. We interpret the present lower crustal lens as underplated igneous rocks left after extraction of the extruded basaltic lavas. A considerable quantity of new material has been added to the crust under the rifted margin. The present Moho is a new boundary formed during creation of the margin and cannot, therefore, be used to determine the amount of thinning.

Most fault-controlled basin formation within plate interiors occurs by normal faulting in response to horizontal deviatoric tension in the continental crust. It is suggested that the tension originates either from the plate boundary forces acting at trenches or as a result of isostatically compensated uplifted regions such as East Africa. The tension produced by both mechanisms is greatest in high heat-flow regions where the upper elastic part of the lithosphere is thinned and weakened. Particularly widespread tension in the continental lithosphere occurs when subduction takes place on opposite sides of a large continental mass, such as Pangaea in the early Mesozoic, where it led to widespread graben formation and, in association with hot spot activity, to continental splitting.

Isostatic subsidence of a downward narrowing wedge of the brittle upper crust bounded by normal faults is shown to provide a viable mechanism of graben formation. Energy balance calculations indicate that the mechanism depends on the occurrence of substantial deviatoric tension, but it requires a much smaller extension than the lithospheric stretching hypothesis. For the mechanism to be effective, a relatively high geothermal gradient is required to allow ductile flow of the lower crust, and the friction on the faults must be much less than that appropriate to dry rock. The amount of subsidence also depends on the narrowness of the graben, the strength of the tensile stress field in the upper crust, and sediment loading. Preliminary finite element analysis has been carried out in an attempt to model the subsidence process. It is suggested that the tension causing the majority of the present day rift systems is produced by the topographical loading and isostatic body forces related to the associated uparching. Present plate boundary forces may be of some importance in the vicinity of convergent plate boundaries and in forming pull-apart basins along transform faults where they bend. On the other hand, plate boundary forces may have been of much greater significance at certain periods in the past when large continental regions such as Pangaea in the early Mesozoic may have been subjected to general tension resulting from the trench suction force acting at the subduction zones on opposite sides.

Summary. The temperature distribution beneath an ocean ridge with a magma chamber solidifying to form the crustal layer 3 has been investigated by numerical methods. Assuming that water circulation rapidly cools layer 2 but does not penetrate deeper, it is found that a magma chamber forms provided that the half spreading rate exceeds about 0.45 cm y-' ; for slower rates instantaneous dyke-like solidification occurs down to the Moho beneath the ridge axis. Numerical models show that the width of the magma chamber and the thickness of the dyke complex depend on half-spreading rate. The width is about 20 km for 3 cm y-'. If there is significant crystal settling to form cumulates, the chamber width is much reduced. The resulting heat flow pattern is as follows: an outer zone where latent heat effects are negligible; a middle zone about twice as wide as the magma chamber where latent heat significantly contributes to the heat loss; a narrow axial zone where quenching of magma to form layer 2 provides the main heat loss. The computations support Cann's petrological model and suggest that the depth of water penetration at the accretion boundary defines the layer 2-layer 3 boundary.

A series of major sedimentary basins of Mesozoic to Tertiary age are associated with the Brazilian continental margin between 8° and 14° S (Fig. 1). This passive continental margin formed when Africa and South America split apart with the initiation of seafloor spreading in the Aptian/Albian3. The offshore basins which border the continental margins on both sides of the Atlantic show evidence of both rifting and crustal attenuation before break-up and thermal subsidence after it. In contrast, a combined stratigraphical and gravity study of the Brazilian onshore basins, which are separated from the offshore basins by a northward-widening strip of Precambrian, suggests that up to 7 km of non-marine sediments were deposited during the rifting stage of the South Atlantic, with no apparent extension of the lower crust beneath these basins and no subsequent thermal subsidence. A mechanism for the linked development of these onshore and offshore basins is proposed here along the lines suggested by Wernicke4,5, in which upper crustal extension in the onshore region is connected to, and balanced against, deeper lithospheric extension beneath the incipient passive margin by intracrustal detachments.

The crustal structure just south of the Iapetus suture1 along a line extending from the mid-North Sea high across northern England to the Solway Basin has been investigated as part of the Caledonian Suture Seismic Experiment of 1982. This unique sea-to-land wide-angle explosion seismology project has produced data of unprecedented quantity and quality. The main innovation was the use of both closely spaced shots (mainly at sea) and closely spaced stations (mainly on land), yielding a data set which can eventually be arranged in a variety of new ways for phase correlation and modelling, such as constant offset and common mid-point. The line was extended by Irish and German groups across Ireland to the Shannon Estuary, having a total length of over 600 km, but here we present only preliminary results for northern England and adjacent marine regions. The results for this region differ radically from those of the earlier LISPB experiment2 in that we have recognized a well-defined 6.15 km s−1 basement interpreted as pre-Caledonian metamorphic basement beneath most of the line at relatively shallow depth, and there is a well-defined mid-crustal refractor starting at approximately 16 km depth beneath the North Sea and northern England which is not recognized beneath the Irish Sea.

Summary. Regions of plateau uplift such as western USA and East Africa are known to be in isostatic equilibrium as a result of hot, low density regions in the underlying upper mantle. The combined effect of the excess topography and upthrust of the low density region below it causes deviatoric stresses to occur in the intervening region. These stresses have been investigated by finite element analysis. Assuming a purely elastic model, the horizontal deviatoric tension is about 20 MPa (200 bar) in the crust beneath a compensated plateau uplift of 2 km, with stress differences in the mantle dying off towards the base of the low density region. If the lithosphere is more realistically modelled by an upper elastic layer above a viscoelastic region, then the equilibrium deviatoric stresses are much larger in the elastic part of the crust, reaching about 100 MPa (1 kbar) if the elastic layer is 10 km thick. A smalI finite strength in the viscoelastic region has little effect on the equilibrium stresses. The value of the viscosity does not much affect the equilibrium. Such a stress system may explain the Occurrence of graben formation, dyke intrusion, and possibly continental splitting, within the setting of a hot upper mantle with associated domal uplift.

Bott, M.H.P., 1980. Crustal doming and the mechanism of continental rifting. In: J.H. Illies (Editor), Mechanism of Graben Formation. Tectonophysics, 73: 1–8. A mechanical explanation of the relationship between crustal uparching and graben formation is suggested, based on the geologically supported assumption that doming precedes rifting. Rifting and graben formation can occur under conditions of crustal tension by the Vening Meinesz wedge subsidence mechanism modified to apply to the uppermost 20 km or thereabouts of the continental crust. If sediment-filled troughs of around 5 km depth are to be formed by this mechanism, then a persistent tension of about 200 MPa (2 kbar) is required. Such a stress system may result from the combined effect of the topographic load of the uplifted region and the upthrust caused by the underlying low-density upper mantle region below. If the upper 10 to 20 km of the crust is elastic but the underlying region can deform slowly by visco-elastic creep, then stress differences of the order of 200 MPa can occur in the upper elastic crust by this process. It is suggested that such a stress system, rather than one arising from bending of the uparched crust or from plate boundary stresses, may be the primary cause of the rifting.

Summary This paper presents those results from the sea-to-land North Atlantic Seismic Project of 1972 which apply to the North Scottish shelf region. Time-term analysis based on first arrivals yielded P, of 6.1050.15 km s-', P* of 6.4850.06 km s-' and P, of 7.9950.02 km s-'. The presence of a P* refractor yielding first arrivals distinguishes the northern British crust from that south of the Caledonian belt. Beneath the Caledonian foreland, the P" refractor varies in estimated depth between 2 and 16 km and its depth correlates with the Bouguer anomalies; it most closely approaches the surface beneath a 90 mgal Bouguer anomaly high west of the Shetland Islands, and is interpreted in terms of widespread occurrence of Lewisian granulites beneath the region. The Moho beneath the foreland occurs at a remarkably uniform estimated depth of 26+2 km. The P* refractor has been identified beneath the western part of the Caledonian fold belt and beneath the Moray Firth, and may occur more extensively. The crust beneath the fold belt is of the order of 20 per cent thicker than beneath the foreland, except beneath Moray Firth where the crustal thinning to the present estimate of about 23 km may have occurred during the post-Caledonian formation of the Moray Firth basin.

A mechanism for causing graben-like subsidence by crustal stretching in response to tension is suggested, based partly on previous hypotheses of Vening Meinesz, Artemjev and Artyushkov, Bott and Fuchs. The mechanism requires rheological subdivision of the crust into a brittle upper layer about 10–20 km thick overlying a ductile lower crust. The brittle layer responds to tension by normal faulting and wedge subsidence; the ductile layer responds by local or regional thinning and by lateral flow of material from beneath the subsiding wedge causing complementary uplift by horst formation or elastic upbending. A graben width of between 30 and 60 km is predicted in absence of basement inhomogeneity. Computations of the energy budget indicate that sedimentary basins of more than 5 km thickness can form by the mechanism provided that water pressure reduces the friction on the faults. The mechanism can explain relatively rapid beginning and end of subsidence, and spasmodic sinking may occur. A wide variety of observed graben-like basins can be explained by the hypothesis, including classical rift valleys and the Mesozoic basins of UK and the North Sea, but it is inapplicable to broad unfaulted cratonic or shelf subsidence.

Summary. A seismic refraction experiment along am across the Lesser Antilles island arc in the eastern Caribbean has shown that the crust beneath the arc contains two refractors with average seismic velocities of 6.2 and 6.9 km/s. The upper, 6.2 km/s refractor varies in depth and shows large fluctuations in its seismic velocity. The layer betweeen the refractors has an average thickness of about 10 km, and varies between 2 and 20 km. The total thickness of the crust of the arc is about 30 km beneath Martinique and about 35 km beneath St Vincent. The approximate depths of the Moho beneath the centres of Grenada and Tobago Troughs either side of the arc are 21 and 25 km respectively. Both these troughs, although of different origin, are underlain by crust of modified oceanic type.

Viscoelastic finite element modelling has been used to study the state of stress in the overriding and subducting plates meeting at a subduction zone. The subduction fault is included using the dual node technique. It is demonstrated that substantial horizontal deviatoric compressive stress occurs in the trench-arc region as a result of the downpull of the dense slab and the associated surface depression including the trench and other downflexing of the plates. This may be masked at the trench by bending stress. It is the lack of significant shearing stress along an unlocked subduction fault in the presence of this compressive stress that gives rise to the slab pull and trench suction plate boundary forces. Slab pull and trench suction were found to be of comparable magnitude within the range 1.0 to 4.0 × 1012 N/m in models studied with vertical subduction, and there are indications that this may also apply when the slab dips at 45° as a result of viscous flow induced by rollback. When the slab dips beneath the arc-backarc region, it is shown that horizontal deviatoric compression can occur in this region contemporaneous with plate interior tension produced by trench suction. This suggests that backarc tension associated with Marianas type trench-arch systems may be related to the nearly vertical slab whereas backarc compression in the Chilean type may result from the low dip and small downpull of the slab. It is also shown that successive locking and unlocking of the subduction fault may give rise to large variations of stress in plate interiors.

Bott, M.H.P., Armour, A.R., Himsworth, E.M., Murphy, T. and Wylie, G., 1979. An explosion seismology investigation of the continental margin west of the Hebrides, Scotland, at 58°N. In: C.E. Keen (Editor), Crustal Properties across Passive Margins. Tectonophysics, 59: 217–231. During summer 1975, a line of large shots was fired across the continental margin between the Rockall Trough and the Hebridean shelf along 58°N. Arrivals were observed at temporary seismic stations set up across Scotland and in northwestern Ireland. A clear Pn phase was observed to cross the margin and a converted phase P1 also seen on the records is interpreted as travelling in the sub-sedimentary oceanic crust of Rockall Trough and in the upper continental crust beneath the shelf. The continental crust beneath the Hebridean shelf is estimated to be 27 ± 2 km thick, with Pg = 6.22 ± 0.03 km/s and Pn = 8.01 ± 0.04 km/s as determined by time-term analysis. Pg delays on the outer shelf are interpreted in terms of a seaward thickening wedge of Mesozoic sediments which pre-date the split. Pn beneath the Rockall Trough was poorly determined at 8.20 ±0.17 km/s and the Moho is estimated to be 18 ± 2 km deep at 58° N. This and other seismic and gravity work indicates a northward thickening of the crust along the Rockall Trough, accounting for the northward decrease in the height of the slope. Our results, and those of gravity interpretations, indicate a relatively abrupt transition between continental and oceanic crust, possibly correlating with the lack of major shelf subsidence. This is attributed to a relatively cool origin for this margin. The main thinning of the continental crust beneath the slope is attributed to outslip of continental crustal material into and beneath the newly forming oceanic crust during the first few million years after the split, possibly enhanced by pre-split stretching.

Continental splitting is the response of the continental lithosphere, under favourable conditions, to deviatoric tension and a hot, partially fused upper mantle beneath. It has commonly been suggested that rift systems such as that of East Africa may possibly represent splitting in the process of occurring. Here, the hot upper mantle has probably formed by convective upwelling from deep in the mantle, allowing the escape of heat from the Earth's interior by hotspot activity. This causes volcanism and isostatic uplift of the heated and thinned lithosphere, and the crustal tension subsequently develops in response to the surface load caused by the uplift and the upthrust of the isostatic compensation at depth. It is suggested here that the crustal tension causing continental splitting does not normally originate as in East Africa, but rather by the suction force acting on the overriding plates at trenches at periods in Earth history when subduction occurs on the opposite sides of a large continental mass such as Pangaea. The critical evidence on the splitting mechanism is mostly deeply buried at passive margins and inaccessible to study. Some margins such as the north Biscay one were apparently formed under much quieter volcanic and tectonic conditions than continental rift zones, whereas others such as those of the Icelandic region underwent extensive continental volcanism indicating a hot upper mantle beneath. The Icelandic types of margin probably acted as foci for continental splitting by dyke emplacement, with dykes propagating laterally from them through the more quietly developed parts of the margin to link with existing plate boundaries. Normal plate motions occur in response to the steady escape of heat from the deep interior through formation and cooling of the oceanic lithosphere which can be recycled, whereas continental splitting is seen as the result of the spasmodic escape of heat from the deep interior through the continental lithosphere.

Summary. The paper describes the interpretation of geophysical observations across the Southeastern Greenland continental margin between 58" and 65' N. South of 63", magnetic anomaly 24 is the earliest recognizable oceanic anomaly. North of 63", anomalies 23 and 24 cut out against the margin and there is a complementary widening of ocean floor of this age on the opposite Rockall Margin north of Hatton Bank; this is attributed to local westward migration of the spreading axis north of 63" shortly after the split. Airgun and sparker profiles indicate the presence of three major sediment groups. Two groups of Tertiary age are separated by an erosional unconformity beneath the rise north of 62". The overlying group is interpreted as comprising mainly contour current deposits of later Tertiary age, and the underlying group as lithified oozes of about Eocene age. Below the Tertiary sediments, older seaward dipping reflectors occur between anomaly 24 and the slope. These are interpreted as Mesozoic sediments overlying subsided continental crust. The oceanic-continental crustal boundary is recognized from magnetic anomalies and seismic profiles as occurring to the east of this subsided region, and lies up to 80km seaward of the present scarp which is interpreted as an erosional feature cut by contour currents. Gravity prqfiles indicate that the main change in crustal thickness associated with the margin occurs landward of the scarp north of 63" but corresponds more nearly to the scarp further south.

The Lake District is dominated by an east-west belt of relatively low Bouguer anomaly which attains individual minima over the exposed Eskdale, Shap and Skiddaw granites. The negative anomaly is attributed to a composite granite batholith which underlies the central and northern parts of the Lake District, connecting the exposed granites at depth. Interpreting on the basis of surface density measurements, the granites appear to extend to a depth of about 7 to 10 km and the contacts with the country rocks generally slope outwards. There are substantial variations in density within the composite granite body. The roof region of the granite includes a series of shallow granite 9ridges,9 one of these connecting the Eskdale and Shap granites, and another connecting the Ennerdale, St. John9s, Threl-keld and Skiddaw granites. The Shap granite is probably connected to the Weardale granite by a deep granite ridge. Present and past uplift of the Lake District may be attributed to the granite mass deficiency, which is estimated to be 1.1 × 10 18 g and is approximately equal to the present elevation of the Lake District above a 270 ft (82 m) datum. The low gravity values along the Vale of Eden suggest that the Permo-Triassic rocks reach a maximum thickness of at least 1 km northeast of Penrith, and that these rocks formed during contemporaneous subsidence. The detailed gravity interpretation of the structure of the Vale of Eden allows a new assessment to be made of the structural history of the Pennine line which reconciles the Hercynian structures with the occurrence of Whin Sill or dyke pebbles in the Upper Brockram.

As a result of more advanced techniques now in use, the previously estimated depth of some 5,000 feet to the Weardale granite in the northern Pennines, England, is revised as follows--beneath Weardale, less than 3,500 feet; beneath Rookhope, less than 4,000 feet, probably between 2,000-3,000 feet.

Summary A vertical field magnetic survey of Durham and South Northumberland supplements the gravity survey described in the previous paper. The field survey and its reduction are described and some of the problems of interpretation are mentioned. A negative magnetic anomaly over the Alston Block is interpreted as another manifestation of a concealed granite under Weardale. Similar anomalies occur over exposed mineralized granites elsewhere, and arise from granite with low polarization existing in country rocks with higher polarization. The magnetic minima are displaced to the south-east of the corresponding gravity minima, and it is thought that this is caused by permanent polarization directed westwards in the country rocks. A violent positive anomaly lies immediately to the north at Blanchland and is tentatively interpreted as a marginal feature of the assumed granite. A magnetic anomaly of Caledonian strike in Northumberland probably marks the southern limit of low-polarization basement rocks. This may well represent the south-eastern limit of the sub-Carboniferous outcrop of the Southern Uplands Upper Ordovician and Silurian rocks.