Records of Holm Cultram.

I. Physical Character.

The Holm district is a westerly extension of the plain of Carlisle. It differs in general aspect from the rest of Cumberland by the absence of stones and rocks, which creates a scenery the reverse of rugged and makes building material scarce. The rock lies at a great depth and forms the floor of a depression known to geologists as the Carlisle basin, an underground valley filled with looser soils. This extends from Maryport on the west to Hethersgill on the east, the southern boundary of the basin running through Plumbland and south of Wigton and Dalston to Brampton, and on the north it reaches beyond Annan and as far as Canobie. The late T. V. Holmes, F.G.S., in 1883 recorded borings which show its actual depth (Transactions, Cumberland Association, vol. viii, pp. 19ff.)—at Kelsick Moss near Abbeytown, 198 ft. 6 ins.; at Bowness-on-Solway, 41 ft.; at Lynehow, below Westlinton, 36 ft. 3 ins.; and at Garlands Asylum, 28 ft. At Kelsick Moss the beds traversed were sandy and gravelly to a depth of 92 ft.; below that, to the rock floor, was mainly clay. At Garlands the borings found 26 ft. of sand and gravel with 2 ft. of clay beneath.

The floor of the basin is St. Bees sandstone, overlaid in the southern and western area with gypseous shales. The whole of the north-eastern part has been planed down by denudation and further depressed by a fault with the downthrow to the north, running from Brackenbank on the Eden to about Dalston. The surface of the Holm, therefore, lies far above the rocks which are so conspicuous in the scenery to the south and east, and its only stones are glacier-borne boulders.

Further information was obtained in the making of the Silloth dock, recorded by Dr. J. Leitch (the same Transactions, 1885, vol. ix, p. 170). From the surface downwards the formations were:—(a) blown sand, about 8 ft.; (b) layers of sand and gravel, altogether 8 to 10 ft., in which was a shell-beach about 2 ft. deep at about 11 ft. from the surface; the shells were of species at present existing in British waters; (c) gravel, 10 ft. thick; and (d) to a depth of at least 30 ft. though not probed further, red and sandy clay containing great waterworn boulders from Criffel, Ennerdale and other sources. Of fossil remains were found Red deer (Cervus elaphus), the extinct Ox (Bos primigenius) and a Fin whale of a species of Balaenoptera still existing. All the mammalian remains were in the gravel, mainly at about 28 feet and just above the boulder-clay.

An ancient raised beach was traced by Messrs. R. Russell and T. V. Holmes between Workington and Bowness (the same Transactions, part. ii, p. 68; 1876–7), pointing to an elevation of the land like that which has been observed on the Scottish coast north of the Solway. From Workington it stretches almost continuously to Silloth. Between Silloth and Grune point there extends along the coast a gravel ridge, on the east of which the land is some 4 or 5 feet below its level. North of Moricambe Bay a similar ridge runs through Cardurnock and Herd Hill to a point just west of the Solway viaduct. The general elevation of this beach is 20 to 25 ft., occasionally rising to 40 ft. It is apparently the result of a very gradual rise of the land, taking place before historical times, followed by a slight depression. Messrs. Russell and Holmes, with other writers, thought that the marsh between Drumburgh and Burgh was lower in Roman times than at present, and that—as it was under water—the Wall was not carried across this stretch of country. But Professor Haverfield, after considerable exploration, decided against the view that the Wall had run to the south of the present coast-line; and as he could not believe that this part was undefended, he said that he would not reject the view that the Wall had been continuous between Burgh and Drumburgh, although its existence there had not been proved (C. & W. Antiq. Soc., Transactions, o.s. xvi, 96). This would mean that the marsh has been formed since Roman times; and it may be remarked that submerged forests from St. Bees to Skinburness, and ploughed ground close to the shore at Skinburness— not to mention the inundation of about 1303, of which later—give good reason for believing that the erosion has been continued or repeated within the historical period. Whether this is the result of actual subsidence of the land is not clear; the opinion of geologists has negatived such a theory, but there has been considerable loss of foreshore all along this coast.

II. Prehistoric Ages.

The late Mr. Steel of Southerfield owned a collection of stone and bronze relics which were dispersed after his death, and no list is known. It perhaps included the stone implements, celts, etc. and the bronze spear-head from Southerfield shown at Carlisle in 1859 at the visit of the Royal Archaeological Institute, and named by Chancellor Ferguson in his Archaeological Survey of Cumberland, 1893. Other objects mentioned by the same were stone hammers, polished celts and a stone adze from Mawbray, and polished celts from Newtown of Mawbray (Proceedings, Society of Antiquaries, N.S. viii, 493–4).

The Ordnance Map notes:—a stone hammer, found S.W. of St. Roche's in 1873; a celt, found halfway between Hayrigg Hall and Gillbank in 1890; a stone axe found 200 yards west of Mireside in 1898; a spearhead found half a mile S.S.E. of Aldoth in 1898; and an arrow-head, found half a mile S. of Aldoth in 1899. And in 1916 a stone axe-hammer, 10½ ins. long, was found south of the railway near Causeway head (C. & W. Ant. Soc., Transactions, N.S. xvii, 254) and is now in the Carlisle Museum.

There are no hut-circles or other British remains known in the parish. The nearest were three "British Settlements," now obliterated by agriculture, between West Newton and Newtonfield, close to the southern border of the Holm.

There are also no British place-names in the Holm. Rivers are usually called by the names given to them in remote antiquity, showing some continuity of tradition in spite of invasion and alien settlement; but here we have only the Waver, written Waura in a charter dating soon after 1150, and Wafyr in Earl Gospatrick's charter of about 1060. Professor Ekwall (Place-names of Lancashire, 112, 229) comes to no conclusion about its derivation; waver in place-names connected with water is wide-spread, but may be of O.E. (Anglo-Saxon) origin and cannot be claimed as one of the ancient British river-names, so far as our information goes at present.

III. Roman Remains.

Roman occupation left one important relic, the fort at Beckfoot, Mawbray. Its site was more or less conjectural until 1879, when Mr. Joseph Robinson of Maryport made a partial exploration. He found the remains lying chiefly in the third field west from Beckfoot mill. Mr. Robinson described his finds (C. and W. Ant. Soc., Transactions, o.s. v, 139) in these terms:—

"A hole six feet in depth was dug. The first obstacle was a pavement of cobbles. I preserved the first dug up and it is much worn on the surface. Underneath was a bed of black earth mixed with slate, pottery and stones bearing signs of work and fire; at 3 feet sand was reached, and 18 inches beneath this a block of welldressed freestone was brought up. We afterwards found this part of the field to consist largely of such layers as are here described . … The four corners of the camp [read fort] have all been found and uncovered, and thus we have its dimensions, viz— interior, east and west, 405 ft.; north and south on west side 283 ft., on east side 267 ft. The area of the camp is about 2¾ acres … It has no gate on the west or seaward side and the gates on the north and south sides are nearer the west gate of the camp than the east. There is a gate in the east or landward side. Two guard chambers occur at the south gate and two at the east, but only one at the north. The walls are each 2 ft. 6 ins. in thickness and the interior space nine feet square; buildings have existed outside the camp to the north-east, and probably elsewhere."

At the south-western corner of the fort were found an altar measuring 17 ins. by 7 ins. and bearing a figure of Diana (now no. 146 in the Carlisle Museum, given by Mr. Thomas Carey in 1914); a mutilated Victory; three querns; a coin of Trajan, much worn, and one of Constantine; two copper beads; fragments of copper and iron; a round stone, 11 ins. in diameter, with a hole near the edge; also Samian, Castor, Upchurch and Salopian ware. The carved stone mentioned in Hutchinson's History of Cumberland, ii, 346, Mr. Robinson re-found in 1880 built as a gatepost into a wall; it is given in Lapidarium Septentrionale as no. 903; in Huebner's Corpus vii, no. 417; and it is now at Netherhall, no. 25 in Mr. J.B. Bailey's Catalogue. The stone, broken in two, is 5ft. long, with an inscription in letters 3 ins high,—]LIA . PRAEF . COH . II . PANNON . FECIT, which is the conclusion of a statement that a certain piece of building at Beckfoot fort had been done under the care of an officer of the second cohort of Pannonians.

In 1908 there was found near this fort a large, narrow-necked store-vessel (olla), 10½ ins. high, now in the Carlisle Museum. Messrs. T. May and L. E. Hope (C. & W. Trans., N.S. xvii, 173) say that such vessels, of Belgic origin, ornamented with cordons, raised bands, comb-markings, etc., and with a black polish, date until near the end of the first century A. D.

In 1921 Mr. Harold Duff presented to the same museum pottery found by him in the sand-dunes near Beckfoot, including many fragments of Castor ware. One of these is part of a beaker with ornament of stems and leaves en barbotine. Another is a nearly complete cooking-pot, 8 ins. high, of brown ware, with lattice ornament; this has been said to date from the first century before Christ, but Mr. R. G. Collingwood remarks that in form it is identical with the ordinary cooking-pot of the second century A.D.

In 1922 further specimens were given by Mr. Duff to the museum, including roof-tiles, rims of mortaria and cooking-pots, Castor ware and part of a white clay flagon coated with black slip. He also found at Beckfoot iron nails, half a silver ring which had been subjected to fire, pieces of carved bone (perhaps from a knife-handle) and much charcoal mixed with the debris, suggesting cremations and burials on the sandhills. In 1925 Mr. Duff found the fragments of a Samian vessel, shape 33, with potter's stamp dagomorvs, probably made about A.D. 110–120, and a small brass of Carausius, A.D. 287–293.

The second field to northward of the fort yields large quantities of pottery and tiles, and may therefore be the site of the vicus outside the fort.

About 80 ft. from the northern end of the seaward wall, a T-shaped structure of large red-sandstone blocks was found, measuring 4 ft. 9 ins. by 5 ft. 5 ins. Its position, Mr. R. G. Collingwood says, forbids the supposition that it was the spina of a gateway, and it may have been a base for a statue or altar.

More information is needed before it can be said when the fort at Beckfoot was first established, but it certainly existed in the second century A.D. It was ruined and burnt at least once, and rebuilt to survive until the time of Constantine.

This is the only fort or great Roman station in the Holm, but there are several other places where Roman remains have been found, and such discoveries have caused some confusion in the past, when every site yielding relics was supposed to have been a 'station' and every 'station' a garrison-fort. In the next paragraphs we give the interpretation of these, as recently suggested by Mr. R. G. Collingwood.

Mr. Robinson in 1879 found remains in a field belonging to New House farm adjoining Wolsty Bank, a little under a mile from Beckfoot fort and about half a mile to the west of Wolsty Castle. The tenant had noticed that he got better crops on a small hill than on the neighbouring land, and digging there exposed the remains of a square building. The whole of the freestone had been removed, leaving foundations of cobbles, eight courses set in clay, altogether 3 ft. 3 ins. deep. The corners were towards the cardinal points. The wall facing north-east measured 20 ft. 6 ins. externally and was 4 ft. thick. The entrance appeared to be on the south-east, as a rough pavement, 6 ft. by 4 ft., was outside the wall. The interior, 12 ft. 6 ins. each way, had not been paved, but the floor was about a foot below the modern ground-level. Samian, Upchurch and Salopian pottery was found, and traces of a burial near the doorway, covered by fragments of a dish of Upchurch ware. The size of the building, Mr Robinson said, resembled others at Risehow (near Flimby; C. & W. Trans. o.s. v, 124) and Campfield near Bowness.

On this it is to be remarked that better crops would not result from ruins of stone walls, but presumably from a ditch surrounding them. This suggests that the building was a Roman signalstation, consisting of a stone tower surrounded by a ditch and palisade, resembling therefore the second-century Wachttiirme of the German Limes and intermediate in type between the small wooden signal-towers at Gask and thelarge stone towers, 30 ft. square internally, on the Yorkshire coast. This site, together with Risehow and others, indicates a hitherto unrecognised system of Roman signal-stations on the Cumberland coast.

Further, a find of coins was made in 1894 at Cotebank, half a mile south-west of Skinburness; and the altar (Huebner, Corpus vii, 418) described in 1866 by Dr. J. Collingwood Bruce as 10 ins. high with an inscription matribvs parcis ('to the Mother Fates,' like no. 35 in the Carlisle Museum) was found among boulders on the coast near Skinburness (it is now in the British Museum). These finds have suggested a Roman fort at Skinburness; but Mr. Harold Duff has recently made careful search in that neighbourhood without finding anything to justify this belief, even allowing for coast-erosion. This may perhaps be another signalsite, one of a series to watch the coast when invasion from Ireland was expected.

As to Roman roads in the Holm, Mr. Robinson discovered one running west in the direction of Maryport, and traced it eastwards for 430 yards towards the mill. It was composed of large cobbles with smaller stones on the top, and measured 15 ft. in width. The Causewayhead, running N.E. and S.W. through the parish, used to be thought the Roman road between Beckfoot and Bowness (Whellan, p. 238), but of this we cannot offer any confirmation.

Finally, in 1920 Mr. Duff observed a rectangular enclosure on the shore, 400 yards below high-water-mark and 500 yards west of Mawbray Yard. It is made of rough granite boulders and measures 106 yards from N.W. to S.E. by at least 99 yards, an N.E. side not having been located. There is nothing to show that this is Roman. Perhaps it may be compared with the fishgarth at St. Bees described by Mr. F. W. Smith (C. & W. Trans. N.S. xxiv, 368).

IV. The Anglian Period.

In the fifth and sixth centuries A.D., after the period of direct Roman government, no doubt the Romano-British remained in Cumberland. We have, however, not only no traces in the Holm of habitations that can be ascribed to them, but place-names indicating their presence are wanting. This may have been the result of the exposed site of the Holm, liable to attack from the unquiet Irish Sea, and especially open to the first inroads of the Vikings in the ninth century. That the Angles, coming by land from Yorkshire, did not exterminate the native Britons, is seen by the presence of Brythonic place-names in north-eastern Cumberland and in the Eamont valley, where there are many, and to some extent in the country between the Holm and Carlisle, and south of the Holm as far as Eskdale (Prof. Ekwall, Scandinavians and Celts in the N.W. of England, 1918).

As to the time of the Anglian settlement, we know from Bede's account of St. Cuthbert's visit to Carlisle (Life of St Cuthbert, chap. xxvii) that in 685 there was already a monastery in that city, the abbess of which was the queen's sister. Bede says also that the saint himself had come there to dedicate the church of another monastery. This means that Carlisle was a considerable place; it must have been Anglian for some time. Now from the philology of the place-names Professor Ekwall concludes (English Place-names in -ing, 1923, p. 157) that though "we do not know for certain when Cumberland became Anglian … it is quite possible the southern, like the northern, part was added to Northumbria by Æthelfrith (593–617)." That is to say, Æthelfrith perhaps overcame any resistance offered by the natives; and after that, Angles began to settle. The names of Addingham, Hensingham and Whicham (anciently Witingham) are of a type suggesting settlement early in the seventh century, so that there must have been at least two generations of Anglian colonization round the Cumberland coast before St. Cuthbert's visit.

We have a possible trace of their presence in the Holm. In the 'Recapitulation' appended to the History by Symeon of Durham (Surtees Society, vol. 51, pp. 67–8) an entry under the year 854 mentions places belonging to the bishopric of Lindisfarne. Beginning with "Lugubalia, that is Luel, now called Carleil," it goes on to "these mansions, Carnham (Carham) and Culterham" and others. The late J. Hodgson Hinde, editor of the volume, identified the last with Holm Cultram. There is nothing we know either to support or to dispute the identification; (fn. 1) but it is, at any rate, possible that (Holm) Cultram was already in the ninth century an estate belonging to the see of Lindisfarne, as Cartmel was—not a monastery but a possession of the bishopric. This would give an explanation of the difficult name as originally the homestead (ham) of an Angle, perhaps named Ceolthryth; and the derivation is much more likely than others that have been proposed from Latin. When, in 1150, 'Holmcoltria' was given to the monks, it was already so called; it was wild forest and hunting-ground, anything but cultivated; it could hardly have been known by a Latin name as culta terra or cultura.

The other place in the Holm which seems to betray Anglian occupation is Mawbray, in 1150 'Mayburg.' This may be 'mægburg' (found in Beowulf, line 2887) meaning a tribe or clan, and in this case used for a family settlement.

The Holm in Anglian days must have been a poor place. No remains of carved stones are known to show the burial of wealthy persons, as at Carlisle, Irton, Waberthwaite and elsewhere.

V. The Viking Settlement and the Place-Names.

The Danish conquest of eastern Northumbria did not touch this district, which continued as it was for another fifty years until, in the first quarter of the tenth century, the Vikings, who had often raided the coast, began to settle on it. Their earlier raids are shown by the pagan burial of a chief at Beacon Hill, Aspatria, in the barrow explored in 1789, and in another at Hesket, of which the relics are in the Carlisle Museum. The silver fibula found at Brayton before 1790, when Pennant published it (Tour, ii, 44), is further evidence of the Vikings in this neighbourhood. When they settled, at first already half-christianized, they left one interesting monument at Aspatria, in the rude grave-slab with a swastika, the form of cross they had learnt from contact with Eastern Christianity. Later, and in the second half of the tenth century, they set up their monuments at Aspatria, Crosscanonby, Bromfield, Dearham, Plumbland, Bridekirk, Brigham and elsewhere, round about the Holm, though none have been found in the Holm itself. Across the Solway their traces are plentiful; so that the Holm was the centre of a great area of Viking occupation, and could hardly escape their presence.

They were of Norse origin, but much mixed with Gaelic Celts, from Ireland, Man and the Hebrides; and their language was Old Norse. Where they settled, they left their place-names; and in the Holm we find them. Only a few show evidence of the original early settlements as indicated by purity of the Norse forms; for during the twelfth century the language became changed by mixture with English, and developed into the dialect of Cumberland, from which most of the names in the Holm are given. We can pick out some that appear to be of the tenth or eleventh century:—Raby (Rabi in 1150) is obviously of this type; rá means a boundary or landmark, and Lindqvist and Ekwall, who are both authoritative, agree that this is the best interpretation of a not uncommon word. Skinburness is the promontory of the 'Skinburg,' so named in 1175, and this in Old Norse might mean the 'shining fort,' suggesting a beacon, which was not impossible in the eleventh century as shown by instances in Orkney (Orkneyinga-saga, capp. 71, 74) though it is hardly possible that any tradition of a Roman signal-station remained; the site, however, would be still one where history could repeat itself. Edderside looks like the usual form in -side for O.N. sætr, a shieling, though the name does not occur before 1537; Professor Sedgefield (Place-names of C. & W., 46) suggests the Anglian personal name Eadhere, and it is conceivable that an Englishman kept the Norse lord's herds there. Holm for 'island' is Norse, adopted late into Anglo-Saxon; in this case it describes the island, always in Latin insula, cut off by the Holm and Black Dubs from the mainland. Arlosh (in 1185 Arlosk) was a waterlogged district; just possibly from some such phrase as O.N. ár-löskr, the sluggish (reach) of the river (Wampool).

The condition of the Holm in the twelfth century is hinted by the charter of 1150, which mentions Holmcoltria and Rabi as forest, i.e. uncultivated. In 25 years more, the monks had established granges at sites with names suggesting former settlement. They were the Old Grange, most probably at Sandenhouse near the abbey, Mawbray, Skinburg, Rabi, and the grange de Ternis, of the Tarns (O.N. tjörn). Arlosh is not named until 1185. In 1189 the charter of Richard Cæur-de-Lion gives the names of Cocklayc on the Wampool ('haunt of wild fowl'), Cromboc or Crombroc, i.e. Crummock beck (from crum, crooked, with alternatively O.N. bekkr or O.E. bröc, brook; Eyntrepot or Antrepot, 'one tree pool' on the Waver; St. Laurence holm (now Lawrenceholme) on an island in Wedholme marshes, perhaps a hermitage; Midelrigg (now Mealrigg); Polneuton, the stream running through Westnewton; Waytheholm (now Wedholme) which might be O.N. veidi-hólmr,' island (preserved) for hunting'; Wytheskeld, O.N. vídis-kelda, 'spring of the willow,' the source of a tributary of the Crummock beck. All these show the generally uncultivated condition of the country in the twelfth century, as well as the survival of Norse as the basis of Cumbrian dialect.

Wolsty is mentioned 1348, originally perhaps O.N. Ulfs-stigi, the 'path (?) of one Wolf,' like Swinsty in the Holm, 'of Swein,' and Thorphinsty in Cartmel fell, 'of Thorfinn.' Kingside hill was Kyngesete in 1292, before the visit of Edward I, to which it is popularly referred; Professor Sedgefield compares Kinniside in West Cumberland and explains it as the sætr or summer dairy of one Kenneth, O.N. Kinadr. Swaby was Swaleby in 1268, which looks like O.N. Svö-býr, the 'house of Swallow,' used as a woman's name. Ellerby, 1292, might be the 'farm of alders.'

In the Dissolution Surveys occur places which seem to be of medieval origin (the list in Nicolson & Burn ii, 177–8). Abbey Cowbier and New Cowper (Cow-byre); Acredale (the commonfields); Aldoth (printed Adlath by N. & B.) is perhaps the Aldelathe, or 'old barn' of about 1230 in Register no. 234 (Rev. W. Baxter, Trans. N.S. xiv, 276); Blatterlees or Blitterlees, possibly from 'blitter,' the bittern; Brownrigg; Calfehow (Calvo); Coats; Dubmylne (Dubmill); Fowlesyke (Foulsyke); Hayrigg (the ridge with the hedge or fence round it); Hielaws; Mireside; Moss side; New parke (parrock or close); Pollathow, 'pool-lathe-how,' now Pelutho; Plasket lands, in early Court rolls Plassegaytt, 'path through a wash'; Saltcoates; Sevehill or Sivill (Seaville), 'hill of sieves or rushes'; Selathe (1299), Silleth, Silloth, perhaps 'sea-lathe' or barn by the sea; Sowterfield (Southerfield), possibly 'shoemaker's field,' for 'field' is English, and the O.N. sauda, 'of sheep' is unlikely here. Sandenhouse cannot be from 'sand,' of which there is none; an explanation has still to be found.

Grune point is mentioned in 1567 as 'Groyne.' The Oxford English Dictionary quotes a statement that in 1367 Corunna in Spain was called "le Groyne, like a swine's snout [sticking out] into the sea, where they entered the land." This seems to explain the name. Goody hills was known in 1580 as Guddihills; the neighbourhood of God's house law (see chapter XII) makes one suspect a corruption.

A few names may be ancient, but we have no forms old enough to be certain about them. Aikshaw would be good Norse, eikskógr, for oak-wood. Angerton in Arlosh, like the place of the same name in Furness, looks like O.N. (but not Icelandic) angr, as in Hardanger, Stavanger, with tún, 'farmstead.' Cunninggarth is said by some to represent O.N. Konungs-gardr, 'king's court,' but it has no s and this very common word is usually from M.E. conyng-erthe, 'rabbit-warren' (O.E.D.). Professor Sedgefield interprets Lowsay as 'Laghi's island.' Overby was Outhby in 1580; possibly 'Aud's farm.' Slightholme might be O.N. slêtt-hôlmr, 'meadow-island,' and Troddersyke perhaps O.N. tradar-sik, 'sike of the cattle-pen.' Waitefield may be named from a person, but the medieval Watelands in Salkeld (Prescott, Wetherhal, p. 373) and Waitewra in Whitbeck (c. 1200, Cockersand Chartulary) are to be considered, and they seem to suggest 'wheat,' in spite of the lost aspirate.

The rest of the place-names in the parish are either modern or to be understood at their face-value. What we learn from the survey is that none are British, unless very doubtfully the rivername Waver; two are Anglian; a few are relics of the Norse settlement. In the twelfth century we find the country something like a wilderness, which was only gradually brought into cultivation by the monks, under whose management it became habitable and profitable.

Footnotes