Convent garden, Lanherne, St Mawgan (SW 8721 6592)

First recorded in 1777 in 'the ruins of an Antient Chapel in a Field called the Chapel Close' in Roseworthy (note on a water-colour sketch in the Arundel archive in Cornwall Record Office). The name 'Chapel Close' no longer exists at Roseworthy; Chapel Field is a large field located at the reference given above. By 1814 in 'the garden of Lord Arundell's mansion-house, at Lanherne' (Lysons and Lysons 1814, ccxlv). Possibly brought to garden before 1794 when Lanherne given to Carmelite nuns; has remained in garden since.

Monument complete, except head is chipped, but stable; ornament clear, some lichen; text clear; situation good

Complete cross with rectangular shaft, set in a modern base (original method of socketing unknown). The cross-head is not symmetrical. The top arm is splayed at the top as in head type B6; the horizontal arms are slightly splayed at the top but have a straight bottom edge and the lower vertical arm is widely splayed and merges into the shaft. The ends of the arms are straight. The head has a slightly recessed ring, type a. On the shaft, the panels on all four faces are bordered by an incised edge-moulding. Although the patterns are not particularly well laid out, all the carving is deeply cut and neat; individual chisel marks can still be seen.

A (broad): The ring has two incised lines, dividing it into three strands. On the head is a figure of Christ, carved in high relief. The top of the head is mutilated but there is a hollow which may indicate the mouth. The arms extend out horizontally, occupying the lower edge of the cross-arms. The right hand has been chipped off but fingers and thumb are depicted on the left hand. Incised lines on the chest indicate either ribs or the folds of a garment (a faint line around the neck may be the neck-line of a garment). The bottom of the garment is clearly depicted as a double line, possibly decorated with a step pattern. The legs extend down on to the shaft and the out-turned feet, which reach almost to the edge-moulding, rest on the top of the interlace on the shaft. Decoration on the shaft comprises a single panel of interlace: simple pattern F (the Carrick Bend), terminated at the bottom with a Stafford knot. Each interlace strand is divided into three by two incised lines. The pattern is not finished properly at the top: one or two loose strands run up parallel to the legs of the figure. A double incised moulding separates this panel from the one below, which contains an inscription in four lines. The inscribed text is complete and legible and is incised in a predominantly insular script reading horizontally. The text reads:

B (narrow): The upper part of the ring is divided into four strands by three parallel incised lines, the lower part into three by two parallel incised lines. Within an incised edge-moulding on the end of the arm are two double-strand ovals interlaced diagonally. In a single panel on the shaft are three registers of double-strand half pattern A, one of simple D, one of half F, and an uncertain knot at the bottom of the panel, which has been mutilated and is not clear.

C (broad): Five symmetrically-placed bosses dominate the cross-head. The ring is divided into three by two parallel incised lines. There is no line or moulding separating the head from the shaft, on which is an irregularly laid-out pattern executed with double incised strands. The pattern comprises two half pattern F knots, with a ring interlacing the crossing strands between the two knots. Beneath this, without any moulding to separate the panels, is an inscription in two lines. The inscribed text is complete and legible and is inscribed in a predominantly insular script reading horizontally. The text reads:

D (narrow): The upper part of the ring is mutilated; the lower part is divided into three strands by two incised lines. Within an incised edge-moulding on the end of the arm is a mutilated pattern, similar to that on the end of the opposite arm, of two interlacing double-strand ovals. On the shaft is a single panel occupied by an animal with a sinuous ribbon-like body. Its head, at the bottom of the shaft, has mouth, eye and ear visible, and just above this, a small foot. The body rises up the shaft in waves and at the top returns, interlacing the body with Stafford knots (simple pattern E) in each of the spandrils. There are several mistakes in this interlace, because the carver has taken the tail over the body every time. At the bottom, the beast bites the tip of its tail.

This cross is a member of the Penwith group of pre-Norman sculpture, characterised principally by a Crucifixion on one side of the cross-head and five bosses on the other (Chapter IX, p. 88). Other features characteristic of the group are the use of simple interlace patterns in double or triple incised strands and the serpentine beast entwined by Stafford knots (Ill. 93). On no other cross in this group is the sculpture so clear, probably because Pentewan Stone was used here instead of granite.

The Gwinear 1 cross includes all of the decorative elements associated with the Penwith group, except a key pattern, yet none of these allow it to be dated more closely than the tenth or eleventh century. Despite the mistakes in the patterns, this is one of the finest members of the Penwith group, which was copied at Phillack (Phillack 1, Ills. 197–201), so it may be one of the earliest, along with St Buryan 1 (Ills. 29–32), for which a mid tenth-century or later date has been suggested on the basis of historical context (p. 127).

The text on face A probably reads + BREID [E/T I]MAH (Ill. 98). Its meaning is not altogether clear but it could be two personal names joined by et (see further Okasha 1993, 136). The text on face C reads RUŪHOL, presumably for RUNHOL (Ill. 99). This could be a personal name containing the Brittonic element run (of uncertain meaning), though, if so, the second part of the name would be obscure and unparalleled. Since the same word seems also to appear in the similar cross, Sancreed 1 (see p. 199 and Ill. 213), it has been suggested that it was the name of the person who carved the cross, though it could also be that of the person who commissioned it, or the person commemorated. Alternatively it could be a word in Old Cornish, meaning 'inherited' or 'inheritance' (see note by O.J.P. below).

Although now at Lanherne Convent in the parish of St Mawgan in Pydar, the cross came originally from the site of a chapel of St Gwinear at Roseworthy in Gwinear parish. Roseworthy was a demesne manor of the king in 1086 (Thorn and Thorn 1979, 1,11). The chapel's legendary associations with the local saint and the presence here of a holy well may point to an early origin for the chapel site (Henderson, C. 1953–6b, 198–200), although it is likely that by the date the cross was carved it was serving as a manorial chapel. The cross may perhaps therefore have originated as the memorial of a (Cornish?) steward associated with the manor.

The grid reference given in this entry is for the chapel at Roseworthy, and is based on the name 'Chapel Field' given in the Gwinear parish Tithe Apportionment of 1841. This field, whose centre is located 0.5 km (0.3 miles) south-west of the present settlement at Roseworthy Barton, is on high ground and contains the cropmark of an oval enclosure. Although this would normally be taken as the site of an Iron Age or Romano-British round, the Cornwall Historic Environment Record notes (on the basis of information from Charles Thomas) that it 'may indicate the site of the enclosure which once surrounded the chapel' (HER 29560).

Old Cornish runhol (O.J.P.)

Two stones of similar decorative style, Gwinear 1 and Sancreed 1, bear closely similar words inscribed on them (Ills. 99, 213). On Gwinear 1 the word is RUnHOL, while on Sancreed 1 it was read by Langdon as RUNHŌ (Langdon, Arthur 1894, 364, and facing plate; but more recently read as R[–HO.] by Okasha 1993, 253, and see the entry for Sancreed 1, p. 198 below). Langdon's drawing shows the h with a long horizontal tail attached to its lower right-hand corner, an unusual formation which could be intended as ligatured hl, with the horizontal bar of l underslung below the following letter to save space, as found elsewhere; compare for example the ligatured and underslung AL in VITALIANI, Nevern 2, Pembrokeshire (Edwards 2007, 392–4). However, in the context of the word RUnHOL appearing on the similar Gwinear 1, the word RUNHLO (if that is the reading) would best be understood as intended for RUNHOL, with the ligature incorrectly used to express two letters to be read on either side of the following o. Such an anomalous use of a ligature might be ascribed to ignorance of the correct usage, or alternatively (or additionally) to an attempt to remedy the accidental omission of the letter, or to a lack of space at the end of the inscription.

The word has been explained as a personal name, and the coincidence of finding the same supposed personal name on two neighbouring stones of similar style and date has given rise to the suggestion that the name was that of the sculptor (references in Okasha 1993, 136 and 253). However, such a personal name is unattested in the Brittonic name-stock, and its second part would be unexplained and unparalleled. Moreover, it would be unusual at this period to find a sculptor signing his work; the few parallels incorporate some verb indicating that the sculptor 'made' the cross or inscription. [1] The variety of Brittonic personal names is such that one can never exclude a particular form as a possible personal name; but the lack of any parallel makes other possibilities worth exploring.

In fact, a Brittonic word of this form is recorded twice, in Old Welsh and Old Breton glosses of the same period as the two crosses. Tenth-century Old Welsh roenhol and ninth-century Old Breton roinol mean 'paternal, inherited', and Old Cornish runhol would be a regular equivalent of these words. Since the word occurs on two tenth-century Cornish monuments, and has not been convincingly explained as a name, it is worth asking whether a meaning 'inherited, paternal', or perhaps a noun, 'patrimony, inheritance', could be appropriate on two neighbouring tenth-century Cornish crosses.

In Old Welsh the word occurs in a tenth-century gloss to the metrical Latin version of the Gospels by Juvencus (Cambridge, University Library, MS Ff.4.42); the Latin phrase patrii pecoris 'of the paternal livestock' is glossed by two hands, one writing in Old Welsh (or Old Breton) and one in Latin: roenhol, dei patris 'inherited, of God the father' (McKee 2000a, 262–3, text and gloss, and 549–50, commentary; photograph of the passage in McKee 2000b, fol. 24r, line 9). Scribe F of this manuscript, who wrote the gloss roenhol, may perhaps have been Breton rather than Welsh (McKee 2000a, 20–3), so the word may actually be Old Breton rather than Old Welsh; at this date the ending -ol might be happier as Cornish or Breton rather than Welsh, though not necessarily so. Pierre-Yves Lambert has suggested that the word may here be a noun, 'patrimony, inheritance', rather than an adjective (Lambert 1995, 96–100; summarised by Schrijver 1997, 108; Falileyev 2000, 139; and McKee 2000a, 550).

In Old Breton the word roinol occurs in a gloss (seemingly ninth-century) to Priscian's Latin Grammar, occurring in a ninth-century manuscript (Paris, Bibliothèque nationale, MS Lat. 10290). The Latin phrase loco patronimicorum 'in place of patronymic [names]' is glossed loco roinolenuen, with roinol-enuen 'patronymic names' glossing patronimicorum. Old Breton enuen is 'names', and roinol again means 'paternal, inherited' (Lambert 1982, 185 and 199; see also Lambert 1995, 96–100). The derivation of the word roinol 'paternal, inherited' seems not to have been clearly established. Lambert has linked the word with Old Breton roi(a)nt 'kingdom'; McKee is open-minded, as is Patrick Sims-Williams (2003, 68–9 and 108).

A parallel Old Irish gloss to the same phrase in Priscian's Grammar appears in the St Gall manuscript of the same text (Sg), where loco patronimicorum is glossed as Old Irish i n-engraicc anmnae aitherrachtaigthi 'in place of a patronymic name' (Stokes and Strachan 1901–10, ii, 82–3, printing Sg. 30b16; and ii, 84, note g, commenting upon another instance of Old Irish aitherrechtaigthe 'patronymic (name)' at Sg. 32a1).

If 'inherited' or perhaps 'inheritance' is really the meaning of the word appearing on the two Cornish crosses, its significance remains a matter for speculation. As a noun meaning 'patrimony, inheritance', the word might possibly refer to a tract of land, though necessarily not the same tract at the two locations of the crosses; or might it be an allusion to the inheritance of the kingdom of heaven? However, it would seem improbable to name land as 'inherited' without giving any name of a person from whom, or by whom, it was inherited; so this suggestion remains very tentative. It has been suggested that some inscriptions and other monuments in Wales had a function relating to donation or ownership of land.[2] Or some other significance not obvious today is also possible. The possibility of an unattested personal name remains, but in this instance an alternative also exists.

Mid tenth to eleventh century

Lysons and Lysons 1814, ccxlv and figs.; (—) 1817, 73; Gilbert, C. 1820, 700 ; Hitchins 1824, ii, 459; Allom et al. 1831, 31–2; Penaluna 1838, ii, 75; Hingston 1850, no. 18, fig. 18; (—) 1851, 190; Collins 1851, 149; Blight 1858, 31 and figs.; Polsue 1868, 150; Polsue 1870, 294 and fig.; Iago 1870–3, 484–5 and figs.; (—) 1871–3, xxix; Iago et al. 1871–3, xlvi; Rhys 1875, 366–7; Huebner 1876, 2–3, no. 6, and figs.; Iago 1878–81, 399; (—) 1886–9, 110–11; Langdon, Arthur and Allen, J. R. 1888, 311–14, 317, 323–4; Allen, J. R. 1889, 128–9, 217, 222; Langdon, Arthur 1889a, 319–21, 330, 334, 340–3 and figs.; Langdon, Arthur 1889c, 239–40 and figs.; Langdon, Arthur 1890–1, 35, 39, 43, 50–1, 53, 81, 84, 86, 89n and figs.; Borlase, W. C. 1893, 107, 183–5; Langdon, Arthur 1894, 315 ; Langdon, Arthur and Allen, J. R. 1895, 51, 57, 60, no. 32 and figs.; Langdon, Arthur 1896, 357–9, 365, passim and figs.; Allen 1904, 292n, 293; Daniell 1906, 244; Langdon, Arthur 1906, 409, 412, 419, 424, 438, 443n, pl. V, fig. 34; (—) 1910–11a, 37; Henderson, C. 1925, 106; Macalister 1929, 184; Hencken 1932, 270, 276, 303 ; Jenkin 1934, 32; Dexter and Dexter 1938, 223–4, 232–3 and figs.; Macalister 1949, 178–9, 186, no. 1047, and figs.; Henderson, C. 1953–6b, 200; Henderson, C. 1957–60a, 342; Ellis, G. 1959–61b, 197; Thomas, A. C. 1966a, 69; Thomas, A. C. 1967a, 90, 104–5; Pevsner 1970, 17, 116, 207, pl. 35; Laing, L. 1975, 140; Pearce 1978, 108, 178; Sheppard 1978a, 122; Thomas, A. C. 1978, 78–9, fig. 4.3; Rickard 1978–81, 168–9; Pearce 1981, 216, 270 and fig.; Weatherhill 1981, 73; Higgitt 1986b, 141; Preston-Jones and Rose 1986, 159; Todd 1987, 296 and figs.; Okasha 1993, 133–7 and passim, no. 20, figs. II.20(i–ii); Thomas, A. C. 1994, 329, no. 1047; Langdon, Andrew 1996a, 1, fig. on front cover; Okasha 1996, 23, 34, fig. 2.2; Langdon, Andrew 2002, 57, no. 70, and fig.; Pearce 2004, 315; Henderson, M. unpub. 1985, p. 749, entry 391, and figs.

[1] The Pillar of Eliseg, north Wales (Edwards 2009, 167); Andreas 99 (73) and Kirk Michael 101 (74) in the Isle of Man (Kermode 1907, 148, 150; Page 1995, 231).
[2] See Edwards 2001, especially pp. 19, 23, 28, 31-2, 33, 35, 37 and 38.