After several years of field work in the Cheviot Hills, it is appropriate to offer some conclusions from our studies. We lack access to, and expertise in many of the contemporary chemical and isotope procedures, so we accept that our conclusions must remain tentative. We would welcome informed suggestions and corrections.

We have noticed the following striking and somewhat unusual aspects to the Cheviot pluton.

• There was a huge outpouring of lavas, mainly of andesite. The extent of this is now believed to reach up to 600 sq km and, before erosion, at least 2000m deep. Observed layering of tephra deposits in relation to lava confirms the traditional understanding that the outpouring of lava was preceded by an explosive phase. The depth of ignimbrite in the Knock and Brough Hill areas, even after 400 million years of erosion, suggests that the Cheviot volcanic system may have achieved a pretty high VEI rating.
• There is a very marked absence of volcanic necks or vent agglomerate. The brecciated rocks which we have found in the upper Harthope and Hawsen Burn valleys cannot be safely identified as vent material. The abundant veining of quartz, tourmaline and haematite could equally point to a hydrothermal origin.
• There is an absence of evidence for compression except at exposures close to the North side of the summit plateau of the Cheviot, and at the junction of two types of granitic rock on Shiel Cleugh Edge above High Bleakhope. There is, however, widespread evidence of thermal and metasomatic action (but not pressure) on the lavas adjacent to the pluton.
• There are some distinct boundaries within the pluton. These are especially noticeable on Dunmoor Hill. Elsewhere, exposures of bedrock are too scanty to trace boundaries.
• We have only recorded one chilled margin within the pluton, at Shiel Cleugh edge. There is no obvious chilling at the junction of granitic rock types (Marginal and Central Belt types) on Dunmoor Hill. The Quartz-monzonite of Dunmoor Hill close to the junction with the lavas does show an increasing fineness of grain consistent with chilling of the pluton against the earlier lavas.
• Recent research suggests that the pluton is at least 4km deep and 20km wide at that depth.
• Most of the plutonic rock is medium-grained. The Upper Cheviot evolved granite borders on the finest and is therefore almost a felsite. Much of the granitic rock has larger phenocrysts mainly of andesine feldspar, and sometimes of augite and biotite, in a fine-grained groundmass of alkali feldspar and quartz.
• Current dating by the BGS suggests a slightly earlier date of 395-400Ma which would make the Cheviot more or less contemporary with the Skiddaw, Shap and Criffel plutons.
• Other contemporary plutons of the early Devonian period do not have an associated volcanic phase. The Cheviot is unusual in exhibiting both plutonic intrusion and volcanic extrusion.
• There is widespread pyroxene in the ‘Marginal’ quartz-monzonite. This is especially prevalent in the Northern Marginal area.
• Evidence for cauldron subsidence which is an obvious characteristic of the almost contemporary Ben Nevis volcano, is not apparent in Cheviot. Surviving outcrops of lava above the pluton show bedding parallel to the slope of the pluton. This does not support evidence for either stoping or cauldron subsidence.

In light of these observations, we would offer the following suggestions for interpretation of the Cheviot system.

The extent of the andesite lava flows is rather remarkable. Andesite is a relatively viscous lava which normally does not flow far from source; hence the steep-sidedness of andesite composite volcanoes. To achieve the extent of the Cheviot lava flows, multiple vents scattered over a wide area, must have been present. However, the lack of any evidence for vents is problematic. Even allowing for late Devonian erosion, it is surprising that any such evidence is missing. Looking across the Southern Uplands of Scotland from the upper slopes of Cheviot, the numerous volcanic necks of Southern Scotland’s Carboniferous volcanics are clearly visible. Yet these necks survive from only about 50 million years later. Perhaps the answer is that there were no necks in the Cheviot system, and that the massive volumes of andesite were expelled from a series of rifts valleys. Could it possibly be that the network of faults which mark the main river valleys in the Cheviot Hills, are the remnants of these rifts?

The impressive volume of the Cheviot pluton rules out the traditional view that it is a laccolith. There is a real possibility that it is the remnants of the magma chamber or one of the chambers which fuelled the Cheviot volcanic system. Presumably the less silica-rich fractions of the chamber being more fluid, were expelled first as andesite, leaving the more silica rich melt behind to crystallise out into varying forms of granitic rock. Whether this theory is correct or not will depend on evidence gathered from modern chemical and isotope analysis which is beyond our own resources.

It is also clear from previous research and from the BGS past surveys that the pluton is also an intrusive body. The sharp contact between the pluton and the lavas, and the obvious chilling of the quartz-monzonite at Cunyon Crags on Dunmoor Hill, suggest that the lavas had already been consolidated before the intrusion of the granitic rock. The exception is in the North near Goldscleugh where tongues of granite penetrate the andesite, suggesting a perhaps more fluid environment there. The lava shows evidence of thermal and chemical alteration but not pressure from the granites. This suggests that the intrusion took place in circumstances where the lavas were relatively easily pushed aside. The relatively fine-grained structure of the plutonic rocks suggests that consolidation took place close to the surface. There is, however, evidence that, close to the summit plateau of Cheviot, there was some compression of granite and andesite. This evidence taken with the composition of the Upper Cheviot rock as the only unequivocally true granite in terms of Kspar and SiO2 content, together with its relatively fine grain, suggests that it may have formed an incipient but failed rhyolitic lava dome which nevertheless failed to break through the andesite crust. It should be remembered that we can only record data from what is apparently the topmost layer of this large pluton. Different but unobserved conditions may well prevail at greater depth.

The overwhelming presence of glacial drift and blanket bog obscures most of the Cheviot pluton’s bedrock. However, in the few places where contacts can be seen between different types of plutonic rock, they appear clearly defined. Just above Linhope Spout there is a sharp contact between the darker Marginal rock (Quartz-monzonite), and the much pinker Central Belt rock (borderline monzonite/granite). The boundary between the two types can be traced across much of Dunmoor Hill although it is somewhat more fuzzy there with tongues of either type penetrating the other for several metres suggesting that neither were fully consolidated when contact took place. On Shiel Cleugh Edge, there is clear evidence that one phase of Central Belt rock has been chilled against another. All this evidence indicates that multiple intrusions of granitic magma into the top reaches of the pluton must have occurred.

Our final concern is with the abundant presence of pyroxene, both clino- and ortho-, in some parts of the pluton, especially in the Marginal rocks. We have not been able to find any satisfactory explanation how pyroxenes can form in a subductional melt without the injection of mantle material. Pyroxenes form at high temperatures, whereas subduction hydrous melts achieve lower temperatures and are likely to produce the lower temperature hydrous minerals such as amphibole and biotite. Biotite is a universal constituent of the Cheviot pluton with primary amphibole (rather than as an alteration product) more scarce. This is as would be expected in a calc-alkaline subduction hydrous melt. So where did the high temperature pyroxenes come from? The only reasonable explanation that we can find is that much hotter and more basic mantle melt became available in the Cheviot pluton possibly as a result of contemporary transtensional tectonic activity. This could also offer an explanation why the Cheviot pluton out of all its contemporary colleagues in Northern England and Southern Scotland, produced an extensive volcanic outpouring.

On Tuesday 29th November we set out to make a more detailed study of the rocks in the upper Harthope Valley above Harthope Linn. We believe from preliminary samples that these may differ significantly from the granitic rocks of the Cheviot pluton. The day is bright and cold with a hard frost promising to make boggy ground easier to walk over. The furthest permitted parking is at the Hawsen Burn just below Langleeford. From there it is a long but easy farm track to Langleefordhope, and then about a quarter of a mile to the lower Harthope Linn close to the stell (circular sheepfold). After that, the going becomes much rougher with several potentially difficult burn crossings.

The rocks outcropping above the stell prove to be granitic Central Belt group although rather more mafic than usual. Similarly the dyke-like outcrop beside the path about 200 yards above the upper Harthope Linn, is also granitic. This rock seems to give rise to a fairly even rolling hillsides.

Looking down the Harthope Valley from near Harthope Linn.
The granitic rocks of granite/quartz-monzonite give smooth rounded slopes.

From this point onwards, the type of rock changes consistently. We checked this most of the way to the watershed. The rock appears to be a form of breccia. In many places it has obviously been severely shattered. In other places it seems more stratified and less disturbed. Because of the shattered nature of the rock, it erodes into gullies more readily giving a more irregular pattern to the slopes of the hills.

A view looking up the Harthope Valley from above Harthope Linn.
The erosion gullies formed from the breccia rocks can be clearly seen.

An outcrop of typical breccia.

An outcrop of stratified breccia.

The breccia has been cemented together by silica and red haematite. The silica occasionally has space to crystallise out into attractive quartz crystals, mostly rock crystal but a few specimens show a hint of mauve. Occasionally, veinlets of black tourmaline appear with the quartz.

The Harthope valley marks the line of a SW-NE fault almost bisecting the pluton. Lateral displacement can be detected from the plutonic/lava margins on either side of the fault, and can be measured to about a quarter of a mile. Vertical movement, if any, is unknown. The faulting begs the question whether the breccia is a product of a crush zone, or evidence for volcanic vent activity. We remain uncertain about this. However, certain conclusions can be drawn. The breccia must have been cemented together under high temperatures for crystalline silica and tourmaline to have been deposited. The flatter layers imply more stable conditions for at least some of the hydrothermal activity. The presence of haematite (ferric oxide) indicates oxidising conditions but whether these were the result of iron reacting with high temperature water vapour inside the magma chamber, or exposure to the air at a vent, is uncertain.

Severely shattered breccia outcrop

We buy a 10.00 GBP day permit from Savills, Glendale Road, Wooler and drive up the College Valley to Dunsdale where we start our walk up Bellyside Hill. The purpose of the expedition is to examine the granitic rock at the northern end of the pluton at Bellyside Crag which is classified by Al-Hafdh and by previous writers as ‘Marginal’ like the rock of Cunyon Crag and Dunmoor Hill on the south east side of the pluton. We also hope to find and examine the andesite outlier mapped at the head of Goldscleugh to see if it gives any clues as to its relationship with the pluton.

View from Bellyside Hill looking up towards Goldscleugh Sike
The ground behind is the summit plateau of the Cheviot.

Bellyside Hill is Cheviot’s closest approximation to a classic ridge walk. Until about 600m there is a good path passing a series of granitic tors on the way. These are composed of a granitic but in appearance mafic rock with occasional felsic intrusions of pink rock.
At about 600m the path disappears and the only way forward is through tiresome blanket bog vegetation.

View looking north down Bellyside Hill showing the outcropping tors

A felsic phase or dyke in one of the tors on Bellyside Hill

Upper Goldscleugh reveals some rather heavily altered rock which may possibly be andesite, but it occurs as scattered blocks and pebbles, so it is impossible to determine any relationship with the pluton. If it is indeed andesite, its altered state indicates that it must have originated from close to the pluton contact.

We now tramp over to Bellyside Crag. This turns out to be composed of a mafic granitic rock similar to that found on the tors of Bellyside Hill.
The screes around the crag have a large colony of clubmosses.

Thin section of rock from Bellyside Crag viewed with crossed polars at X40
Granophyric texture can be seen in several places. There is a large crystal of clinopyroxene at the bottom of the picture.

Subsequent thin section analysis reveals some interesting facts. The rocks of Bellyside Hill and Bellyside Crag are very similar to each other although the Bellyside Hill specimens may have a slightly higher proportion of alkali feldspar. However, neither are anything like the ‘Marginal’ rock of Dunmoor Hill and Cunyon Crag.
The Bellyside Hill specimens have a lower proportion of plagioclase to alkali feldspar (identifiable feldspars show a ratio of about 1:2, plagioclase to alkali feldspar); there is very little biotite, the main mafic mineral being pyroxene (10-15%); quartz content is similar at about 20%. It also has a more marked fine matrix than the southern Marginal with frequent granophyric texture.
It therefore seems rather doubtful that it has the same origin as the southern Marginal rock.

As regards the “andesite” of upper Goldscleugh Sike, if we are correct in identifying this as a granite/andesite contact, both the andesite and granite show clear signs of banding which may indicate contact compression.
Apart from the contact on Shiel Cleugh Edge, this is the only evidence for kinematic action which we have found on Cheviot. Where it exists it appears to be very localised. However, we remain rather uncertain whether the rock from upper Goldscleugh Sike was andesite or just an altered granitic rock.

Woolhope Crag lies on the north east flank of the Cheviot just above Goldscleugh. It is one of the few convincing outcrops on the Cheviot itself. When we first visited it in 2013, we were looking for, and thought we had found, evidence that the outcrops occur because of the presence of harder, more mafic rock.
Subsequently, we read Al-Hafdh’s thesis which uses Woolhope Crag as the type location for his ‘Woolhope’ variety which is the least mafic, and therefore the most felsic and silica-rich, of his classifications.

Woolhope Crag from the East
The break on the right is where the felsic rock (left) becomes more mafic.

I returned here on19 August 2016 to investigate this apparent contradiction.
The answer is relatively simple. The southern end (higher end) of the crag is felsic; the northern (lower) end is more mafic in appearance. The join occurs at the broken mid-point of the crag. It is not an abrupt contact but a gradual change over about 20 cm and appears at first sight to be an example of fractionation – the first we have discovered in the Cheviot pluton.

Thin section of the upper (more felsic) rock at Woolhope Crag viewed with crossed polars at X40
There are some large phenocrysts of alkali feldspar showing Carlsbad twinning.

This theory breaks down when the two types are compared in thin section.
There seems to be very little difference in content between the two. Thin section analysis suggests that the difference between them is more apparent to the eye in hand specimen than actual. There is a slight increase in plagioclase with the more mafic rock but no increase in mafic minerals (both around 7%). The darker colour is probably due to the smaller grain size of the lower darker rock. Identifiable feldspars in both types show a ratio of about 1:3, plagioclase to alkali. Both types are rich enough in quartz (25-30%) and alkali feldspar to be classified as granite, albeit a relatively fine-grained variety.
An interesting feature of the upper more felsic rock is the presence of frequent chlorite.

Thin section of the lower (‘mafic’) rock at Woolhope Crag viewed with crossed polars at X40
There is no increase in mafic content compared with the more felsic rock. However, the grain size of the matrix is much smaller which must account for the darker appearance.

Quartz and haematite veining

At NT 91826 20984 the rock looks brecciated. The situation is confused at NT 92020 21066 and NT 91789 20966 by the appearance of a much more mafic fine-grained rock with the Evolved type. We are uncertain whether these represent a more mafic differentiation within the Evolved type or are xenoliths of andesite which have collapsed into the roof of the pluton. Thin section analysis should answer this problem.

Andesite xenolith or fine-grained plutonic rock?

The summit plateau of Cheviot is magnificent with a large extents of blanket bog containing cotton-grass, clubmosses and cloudberry some of which was fruiting quite prolifically. We abandon any plans to cross over the bog to Bellyside Crag because there is no path and we do not wish to damage the fragile ecology. We continue along the paved footpath to Cairn Hill and descend to the head of the Harthope Burn. What exposures exist indicate that the rock type is Evolved over this whole area.
We begin to see a change back to the typical coarser pink porphyritic ‘granite’ in the upper reaches of the Harthope Burn at NT 90671 19035. From then onwards down to Harthope Linn there are some exposures of very weathered rock. We lacked time to make a detailed examination of this, and plan another expedition specifically to this area.

From left to right: the succession of rock=types down the Shiel Cleugh Burn
We see increasing mafic content as the burn is descended to the River Breamish.

However, there is little sign of clear boundaries. The rocks which are fairly well exposed in the Shiel Cleugh burn which flows off the southern slope of the Edge into the Breamish, show a gradual change from coarse to finer pink ‘granite’ and then a gradual increase in colour index towards Marginal rock.
There however some rather exciting dicoveries. At NT 91922 16767 we do find a distinct boundary between coarser and finer rock.

Contact between the coarser and finer-grained rock

At NT 91847 16893 and NT 91825 16916 we find further junctions between the two types. At these locations, there is evidence for compression of one against the other together with some flow structure. Significantly there appears to be a chilled margin, and later thin section analysis confirms this. However it is the finer (‘Dunmoor’) which is chilled against the coarser (‘Standrop’). This suggests that the coarser rock cannot have been intruded into the finer.

Thin section with crossed polars showing the chilling of the finer-grained rock against the coarser-grained