The Wider Picture

The relationship between free-face, tor preservation and folding discussed above is seen to be not restricted to the Charnwood district, but occurs elsewhere in environments where tors are better preserved. If the main tor groups of England and Wales are considered in relation to the main geological structures on which they are superimposed, all are found to correlate with epeirogenic anticlines, periclinal domes and associated granitic or doleritic instrusions or monoclines. There are no tors which are found in association with synclinal folds. Table 1 lists the main tor groups and their associated geological structures.

Table 1: The main areas of tors in England and Wales.

Clearly, tors appear to have a closer association with geological structure than with environments in which periglacial conditions would have prevailed during the Devensian. There are extensive tor groups both within the known limits of periglacial activity (the Pennine tors, Prescelly Hills and Cheviot Hills) and beyond (Scilly Isles and Bodmin Moor). Thus, although congelifraction weathering and solifluction removal would have been important processes in free-face retreat and tor isolation in environments where they occurred (Palmer et al, 1961, 1962), tors are by no means defined by type of weathering. If periglacial activity is the sole determinant of tor creation then they should be found in all upland environments where denudation has occurred for example the Chiltern Hills, the Jurassic Escarpment of central and southern England and South Wales. The reason for the absence of tors in these locations lies in either the lithology of the rock (chalk and limestone, unless dolomitised (Cunningham, 1965), is unresistant to physical and chemical weathering) or in the structure of the rocks (South Wales represents a synclinal basin).

The picture is thus for tors to be found in association with geological structures which give rise steep escarpments for rapid free-face retreat and tor isolation or the development of residual hills in the central core of periclinal domes and anticlines These are all environments where the general dip of the bedding planes with respect to the ground surface slope is near perpendicular (as in anticlinal structures and cuestas) or parallel and horizontal (as on exposed batholiths where exhumation has created dilatation joints parallel with the surface). In synclinal structures the direction of dip is parallel to the ground surface and downslope precluding the maintenance of vertical free-faces and tors. Indeed synclinal structures are associated with large-scale mass movements, for example valley-side cambering as has occurred on the Jurassic escarpment of central England (Hollingworth et al, 1944; Ackerman and Cave, 1967; Kellerway and Taylor, 1953) giving rise to the characteristic tor-less and gently undulating relief of counties such as Oxfordshire and Northamptonshire.

The presence of bedding or cleavage planes against the general slope reduces the propensity for mass movements to grade the slope. Blocks which have been displaced through congelifraction have tended to pile up en echelon upslope of the most resistant bed which has acted as a sill on which tors have built up. In areas where there is marked resistance from one bed to another, as in parts of the central and southern Pennines between the Millstone Grit series, shales and Carboniferous limestone series, free-faces or 'edges' are marked giving characteristic 'echelon' tors. Only where cleavage is well developed in the rock on account of long periods of burial, exhumation and uplift, and where significant weathering has taken place are tors poorly developed, although as with Charnwood Forest, they can be discerned as significant piles of logans amongst the general clitter.

The origin of the granite-based tors of the Scilly Isles, Bodmin Moor and Dartmoor is slightly different to the tors found elsewhere. The tors in these environments are clearly not 'echelon' tors and only in a few places, for example Bowerman's Nose on Dartmoor, do they appear related to a retreating free-face. King (1948) postulated that the summital tors of Dartmoor represented residuals of pediplanation on account of their similarity with tors developing from inselbergs in semi-arid environments. These ideas were dismissed by Palmer and Neilson (1962) who claimed that there was no evidence to support Dartmoor being a pediplain or for inselbergs and tors to have survived throughout the Pleistocene. Nevertheless, the exhumation of the Dartmoor batholith would inevitably have resulted in dilatation of the surface granites and their predisposition to weathering.

Exhumation occurred firstly in Lower Cretaceous times (Brammall, 1928), followed by subsequent burial during the Eocene and final exhumation during the Miocene. Chemical weathering undoubtedly occurred during the Eocene concentrating along the dilatation joints in the manner proposed by Linton (1955) and imparting an irregular surface topography. Exhumation of the granite during the Miocene would have further widened the dilatation joints. Subsequent erosion of the chemically weathered granite took place as a generally radial drainage network evolved around the granite dome. Areas of intact granite became isolated firstly as low inselbergs and then massively jointed tors. The final phase of weathering occurred during the Pleistocene and involved congelifraction on the residual tors and removal of the weathered clitter by solifluction. The problem of tors surviving the Pleistocene glacial periods, highlighted by Palmer and Neilson (1962), has now been largely solved. Firstly, there is little evidence that Dartmoor was ever glaciated in the Pleistocene. Indeed, Bodmin Moor and the Scilly Isles even escaped the periglacial conditions associated with the Devensian. Secondly, the behaviour of ice sheets is now better understood and it is unlikely that the nature of ice should it have reached as high as the tor-bearing interfluves would have been conducive to significant erosion (Lewis, 1993). Finally, there is now considerable doubt over the Pleistocene chronology in the British Isles. Atkinson et al (1978) and Gascoyne et al (1983) have shown by analysis of the ages of speleothems in limestone cave systems in the Yorkshire Dales and the Mendip Hills that the relief of these areas was already established before 350000 years BP, i.e. before the end of the Hoxnian interglacial stage. This evidence together with that of Rose (1987) suggesting stratigraphical anomalies at the type site for the Wolstonian glaciation implies that the British Isles may have experienced only one glacial episode capable of exerting a marked influence upon relief. Indeed many have argued that more erosion would have been accomplished since the middle Tertiary by normal fluvial erosion, a mechanism ideally suited for the creation of residual hills or tors.

The tors associated with periclinal domes, such as the Weald in Kent, the Stiperstones in Shropshire and the southern Pennines, would have formed following the same general sequence of epeirogenic uplift and denudation as Dartmoor and often were accompanied by batholithic intrusions, for example the Weardale granite of the central Pennines, although the latter were not exposed at the surface. In many of these periclinal domes the residual hills of the central core have mostly disappeared, although still remain as tors in domes recently uplifted such as the Weald. The remaining tors associated with periclinal domes are confined to the retreating cuestas. Since the cores of all of the older domes are now extensively eroded, the tors found linked with retreating free-faces on the cuestas form the most significant group.

Figure 6 shows the main geological structures with which tors are associated. As the figure shows, tors can be divided into three groups: 'echelon' tors linked to escarpments associated mostly with monoclines, 'residual' tors associated with the denudation of granitic intrusions such as Dartmoor, and 'outlying' tors representing former positions of an actively retreating free-face in younger rocks forming a ridge along an anticline or around a periclinal dome. The tors of Bodmin Moor and Dartmoor are characteristically larger and more consolidated that the other types. It is thought that the greater vertical continuity of granite and the small amount horizontal dilatation jointing is responsible for this. Tors developed from sedimentary rocks have undergone more folding, burial, metamorphism and exhumation which have caused significant dilatation joints and cleaving. They have therefore been subjected to greater weathering and tend to be stouter and generally smaller than their Dartmoor counterparts. Indeed tors developing in the oldest rocks which have undergone the longest periods of earth movements have such well developed cleavage that they have weathered to no more than mounds of clitter. A spectrum of tors can thus be detected ranging from the most intact granite-based tors of Dartmoor through to the general rubble or cairns of Charnwood Forest.

Figure 6: The evolution of tors in different geological environments.

Summary and Conclusions

Tors are believed to have formed a significant feature of hilltops in Charnwood Forest. Evidence has been presented which shows that the area's history of burial and exhumation and the dilatation joints which resulted would have predisposed the rocks to both extensive chemical decomposition and removal through a combined process of congelifraction acting on exposed free-faces and solifluction.

These two processes, being the theories for tor formation of Linton (1955) and Palmer et al (1961, 1962), would not have occurred synchronously. For chemical weathering to be effective warm tropical conditions are needed and these appear to have been experienced during the early part of the Cenozoic. Chemical weathering would have occurred rapidly where the jointing was most extensive, i.e. the apex of anticlines and periclinal domes such as Beacon Hill or the contraction joints of intruded igneous rocks such as found near High Sharpley. The rejuvenation which occurred during the subsequent Pliocene would have facilitated the downslope removal of much of the weathered material.

It was not until the Pleistocene had been reached that periglacial processes began to modify the hilltops of Charnwood. It is probable that many tors were well developed before the Pleistocene and that periglacial activity acted destructively rather than constructively except where the combined process of congelifraction and solifluction led to the formation of tors from the retreating freeface on the flanks of Beacon Hill and other summits. Where solifluction was limited, such as on the level ground of the summits, removal of weathered material would not have been facilitated and most of the summital tors were rendered little more than piles of rubble or cairns. Only where the rock was highly resistent, such as in the areas of intruded igneous rocks like High Sharpley, did the rock remain consolidated albeit highly weathered superficially.

The removal of solifluction as a transport agent on the flanks since the Pleistocene has inevitably aided in the destruction of the tors associated with the free-faces. Contemporary mass wasting is rapidly grading the free-faces throughout the Charnwood area and it is only where substantial free-faces are still present, such as on the north-western flank of Beacon Hill, that traces of former tors can be discerned. The rapid grading which is taking place, in contrast with other tor supporting geologies such as the Millstone Grit sandstone of the Pennines, is in part attributable to the greater degree of jointing and cleavage within the rock and the longer time- scale for weathering. Examination of the relic tors of Charnwood Forest, in relation to its geological history, has challenged some of the more traditional interpretations of tor formation which have tended to over-focus on weathering processes and lithologies capable of producing characteristic tower-form tors.

The view that granite is known to break down chemically to produce tors in tropical climates (Linton, 1955) led to chemical weathering being proposed for the Dartmoor tors. Furthermore, knowledge of congelifraction and solifluction working on and undermining a free-face (Palmer et al, 1961, 1962) placed the tors of the Pennines as former free-faces of the characteristic Millstone Grit 'edges' before the onset of periglacial conditions in the Pleistocene. However, problems have been raised by transferring Linton's ideas directly to the Pennines or Palmer et al's ideas to Dartmoor. Many have suggested that tors form by a combination of both chemical and physical weathering. This has been the view of Gerrard (1974) and it is a view that fits well in the context of Charnwood. However, there has been an over-restriction on viewing the tors of Dartmoor/Bodmin Moor and the Pennines as being the only examples. Tors are far more widespread, not only in the British Isles but elsewhere, and reveal evidence of their formation under a great many climatic environments from the polar regions through to the hot deserts.

Superficial inspection of the geological structures associated with British tors seems to reveal that tors are best developed where folding has placed the dip of bedding and cleavage planes perpendicular to the main direction of mass wasting on hillslopes or where tors have evolved from inselbergs during the pediplanation of uplifted periclinal domes or granitic intrusions. It would thus appear, now that the destructive nature of Pleistocene ice sheets on interfluves can be largely discounted, that tors can result from a variety of weathering processes, chemical or physical, so long as the general geological structure is conducive to their preservation.

A fuller version of this paper has been published by the author in Trent Geographer (1996).