PLATE XXV. FIGS. 1 & 2.

THE CROSS AT LLANGAN.

Llangan is a small village lying in the valley between Coychurch and Cowbridge. The church, dedicated to St. Canna, mother of St. Grallo of Coychurch, is small with the chancel distinct, and with a small two-bell gable at the west end. The interior of the church is chiefly remarkable for a doorway in the north wall, closed by an oaken door ornamented with some very excellent florid Gothic tracery leading up through the wall to another door which led to the roodloft (of which no traces remain). The font is plain and circular. Within the rails of the communion table is a sepulchral slab with a cross fleury of a very unusual style, ornamented with small stars, &c. In the churchyard is a tall fifteenthcentury cross, the head of which is elaborately carved and in excellent preservation. On the west side is the Crucifixion in high relief, the figures small, with the Virgin and St. John at the sides of the cross. On the east side is the Virgin seated, holding the dead body of Christ. On each of the two other sides is sculptured the single figure of a saint: the top of the cross is ornamented profusely with small pinnacles and figures of single saints in niches of small size.

In the churchyard of Llangan stands, half buried in the ground, the curious stone with the representation of the Crucifixion, here for the first time described and figured. The head of the cross is 38 inches in diameter and 6 inches thick, partially injured on the upper right-hand side, containing in the sunk panel on the front side an evident figure of the Crucified Redeemer, with apparently a conical beard, and with the arms awkwardly extended, wearing a short cloth or kilt round the lower part of the very long body, and with the feet separate and apparently with shoes upon them. At the sides of this principal figure are two much distorted and rude figures, representing the sponge- and spear-bearers, the latter on the left side with the spear broken off and with a long beak-like head. This grotesque, or rather monstrous mode of representing the persecutors of the Saviour was not uncommon in early art, and is especially seen in certain Irish sculptures and metalobjects, of which I have given figures in my work on the 'Miniatures and Ornaments of AngloSaxon and Irish MSS.' The representation of the Crucifixion with the figures of the spongeand spear-bearers is of very early date. The introduction of those of the Virgin Mary and St. John at the sides of the cross, as seen on the adjacent churchyard cross, was more generally prevalent at a later period. Below the feet of the Saviour appears a small figure with one arm pointing upwards and the other extended, probably intended for the maker of the cross, but reminding us of the figure beneath the feet of Christ on the great sarcophagus of Junius Bassus in the crypt of St. Peter's at Rome, and of the figure of Abacuc at the foot of the group of the Ascension of Christ in certain ivory carvings described in my 'Catalogue of the Fictile Ivories of the South Kensington Museum,' pp. 144, 4501.

1 The strange Eastern tradition that the tree which subsequently became the cross of the Saviour sprang from a seed from the Tree of Life, given to Seth by the Angel of Paradise and placed under the tongue of Adam at his burial, may also have led to the representation of a male figure or skull at the base of the cross seen in various early sculptures of the Crucifixion.

The reverse of the cross bears a Greek cross formed of equal-sized plain bars within a raised circle, the spaces between the arms of the cross forming sunk panels, each enclosing a raised circular boss.

PLATE XXV. FIG. 3.

THE FIGURED STONE AT GNOLL CASTLE, NEAR NEATH.

Adjoining to the Caratinus stone in the wall of the grotto at Gnoll Castle (ante, p. 6) is the stone here represented, destitute of inscription, but of considerable interest as an archæological relic. It is of irregular form, about 30 inches high and 20 inches wide in the middle, having its surface nearly occupied by a rudely-designed human figure, with the head round and uncovered, the arms raised, with the hands open and fingers spread out, and with a short apron or kilt reaching from the waist to the middle of the legs. Above the head is a series of short straight spokes or bars, some being longer than the rest and bent at right angles, forming a kind of canopy over the figure, which is raised, or rather the surface of the stone is cut away, leaving the figure itself in relief. The surface of the face is also cut away, leaving the sides of the cheeks, with the eye-brows, eyes, nose, and mouth also in relief. The kilt is formed of a series of longitudinal strips radiating from a waistband, and giving the appearance of a short and very thickly quilted petticoat, just as in several of the Irish figures on the shrine of St. Manchan.

Such representations of the ancient Britons on the sculptured stones of Wales are extremely rare, differing in this respect from those of Scotland, Ireland, and the Isle of Man. With the exception of the stone at Llandevailog, Brecknockshire, in which an uncovered figure holds a short club in each hand, and who appears to be clothed in a similar short kilt (figured in Arch. Camb., 1858, p. 306, and post, pl. XXXIII. fig. 2), the small stone at Llanfrynach near Brecon, on which is sculptured a diminutive figure with uplifted arms and outspread hands (also figured by me in Arch. Camb., 1856, p. 141, and post, pl. XXXIX), and the Llanhammlech stone (post, pl. XXXVIII. fig. 3), I recollect no other single figure of an ancient Briton represented on a stone monument. Much space might be occupied in discussing the two peculiarities observable in these figures, namely, the upraised hands and the dress. I shall only observe, however, that the attitude of this and the other figures agrees with that which is found repeatedly in the Catacombs of Rome, and which is generally interpreted as representing the act of prayer or worship, and which seems especially suitable for the tomb of a Christian, whilst the peculiar character of the dress seems especially Celtic, as it is seen in many of the early sculptured remains in other parts of the kingdom, as well as in early Irish metal-work, as in the small full-length figures of the shrine of St. Manchan, above referred to.

PLATE XXVI. FIGS. 1 & 2.

THE ROMAN MILIARY STONE AT PORT TALBOT.

This stone originally stood upon the road from Nidum (Neath) to the southern Bovium or Bomium (either Boverton or Ewenny). Thirty years ago I found it lying in the harbourmaster's office at Port Talbot, broken into five or six pieces. It was about 5 feet long and 1 feet wide, tapering at the lower end, and was doubtless originally inserted upright in the ground. On its face it bears the Roman inscription, written in large capital letters across the stone:

IMPC

FLAVA

(?)DMAXI

MINO

INVIC

TOAV
GVS

Owing to the crack across the third line there is a little difficulty about the first letter, which has been given as CL in the Gentleman's Magazine, 1840, p. 296, and simply as L in Col. Grant Francis's work on Neath Abbey and its Charters (Swansea, 1845, p.8). It appears to me to be a D, unless, as is possible, the top of the curved second stroke of the letter should be only part of the crack of the stone. The oblique elongation of the top of the second stroke of the A without a cross bar and the form of the G in the last line are palæographical peculiarities worthy of notice.

Col. Francis thus speaks of this stone, which was found in 1839 at Port Talbot, between Neath and the Roman station of Bovium. The older inscription, if we may depend on the fidelity of the engraving, is dedicated, not, as some have pretended, to Maximian, but to Maximin, who bears upon his coins the title of Invictus (as upon the stone), which is not found on those of Maximian. The inscription as it now stands may be read thus-Imp. C. Fla. Val. Maximino Invicto Augus(to). But on looking closer we are led to believe that the first part belongs to a still older inscription, as the C. Fla. Val. (Caius Flavius Valerius) were names which never belonged to Maximin, whose name was Galerius Valerius. (The names of the two Maximians were "Marcus Aurelius Valerius Maximianus" and "Galerius Valerius Maximianus.") The appellation of C. Flavius Valerius belongs to the second Severus, and it is probable that to him this stone was originally inscribed, but that subsequently his name had been erased to give place to that of Maximin by some person who carelessly or ignorantly overlooked the prenomens.'

The name of the Emperor Maximinus here recorded (with the title INVICTUS) also occurs in an inscription found in Cumberland, given by Horsley (Brit. Rom. p. 192, No. 40; and conf. Hübner, Inscr. Lat. vii. 1158).

On the other side of the stone, and evidently of a somewhat more recent date than the Roman inscription, is the following, written in debased Roman capitals (except the h):

hIC IACIT CANTVSVS PATER PAVL(?)NVS.

The crack of the stone across the latter word has damaged part of the inscription, but the stroke following the conjoined AV is extended below the line, and was evidently an L, doubtless followed by I, preceding the terminal NVS. The inscription, differing as it does from the ordinary formula of Welsh stones, is therefore, I presume, to be interpreted, 'Here lies Cantusus, whose father was Paulinus'; or may it not mean, Here lies Cantusus, who was the father of Paulinus'?

6

In addition to the notices above given, this stone has been recorded in the Journal of the Archæological Institute,' iii. 1846, p. 274; the Journal of the Archæological Association,' ii. 1847, p. 287; Thomas Wright's The Celt and the Saxon,' p. 456; Haddan and Stubbs, Councils, i. 168; and by myself in the Arch. Camb., 1856, pp. 251, 320. The name Cantusus appears to be unrecorded by the Welsh historians, but Paulinus is repeatedly; and besides the Merthyr Mawr stone above described (p. 8), the stone dedicated to him, now at Dolocauthy House (post, pl. XLIV. fig. 1), is the most interesting of all the Welsh inscriptions. The irregularity in the size of the letters, varying from 2 to 4 inches in height, is their most noticeable peculiarity, the Latinity of the Welsh stones being somewhat defective.

PLATE XXVI. FIG. 3.

THE ROMAN ALTAR AT LOUGHOR.

We are indebted to the late Rev. H. L. Jones for the first notice of this Ogham inscribed altar (Arch. Camb., 1869, pp. 258, 344). Loughor, the mediaval Castell Llychwr, stands on the extreme western boundary of the county of Glamorgan, where the river of that name, coming down from the hills near Llandeilofawr, divides it from the county of Caermarthen. Just where the present decayed town stands near the stream, or rather on the sandy and marshy ground south of it, was the Roman station of Leucarum, and coins, pottery, and other traces of a Roman station have been found here, and tradition points to the site of part of it as just opposite the modern railway station. A small hill rises here immediately from the water's edge, and on it are the earth-works and a square tower of the Norman castle, as well as the parish church, the former having probably been erected with its mound in the strongest part of the station. Mr. H. L. Jones continues: 'It is stated in Lewis's "Topographical Dictionary" that a Roman miliary stone was found at Loughor, where it formed one of the steps leading up to the rectory house. It was not a miliary stone but an altar, and it now stands on the lawn in front of the rectory, in the upper part of Loughor, on the slope of the hill. It is rather plain, but its shape, as will be perceived by the accompanying engraving, plainly betokens its origin and purpose. The material is a fine-grained, white sandstone of the carboniferous series, so close in texture and so light in colour that at first sight it appears to be marble. It is entirely devoid of any sculptured ornaments or inscription, but on examining it closely and under a favourable light in 1857, the author found on the south-west angle of the lower portion certain Oghams, the edge which serves for the fleasg or basal line on which they are cut being chipped in two places, and the oghamic inscription is so far imperfect. Owing to the very light colour of the stone, and

the total absence of shadow cast by these cuttings, the Oghams were extremely difficult to be made out, but considerable time was spent over the monument, and the accuracy of the delineation may be fully depended upon. The Oghams begin from the bottom, and read upwards from left to right, as is usual in similar cases.'

These letters are represented in Mr. H. L. Jones's figure here copied ;-two oblique lines between the chippings of the edge of the stone, a dot above the upper chipping, followed by three oblique strokes to the right, then five dots, and at the top, in the curved space below the square cornice or abacus of the top of the altar, four oblique lines to the left of the middle line. These marks Mr. Jones considers to represent L(?)... ASIC, and he is further induced to consider that they were cut on the edge of the stone after it had ceased to be used as an altar, and when it served for a commemorative purpose; and hence that it is to be considered, not as of pre-Christian times, but of a date later than the departure of the Roman garrison from the adjoining station. The stone is about 4 feet 6 inches high, and the width of its flat sides about 1 foot 7 inches.

Prof. Rhys (Arch. Camb., 1873, p. 198) maintains that the reading is L(?)VIC, which, if the drawing be correct, should be completed by inserting E, which makes it LEVIC, that is, according to the Irish method of reading, Lefic. The former reminds one strongly of Leucarum, the name of the Roman station in the neighbourhood. But which are we to trust, Mr. H. L. Jones's reading or his drawing?' In his Notes (p. 7) he, however, gives a different reading, observing, 'After it (the stone) had been cut for an altar, an inscription in Celtic was made on it, which is now very hard to read-it may be Lehoric or even Vehomagic. Were the former correct, it might stand for Lehori C., i. e. Lehori Castra, meaning Cas Llychwr, which is the name by which Loughor is known in modern Welsh.' On which I may observe that it is not at all usual on these Celtic or Romano-British stones to use initials alone, and that it is also not common to inscribe localities on them.

Mr. Brash, whose experience of the Ogham inscriptions in Ireland was very considerable, states (Arch. Camb., 1873, p. 286) that he had examined and copied the Loughor stone. He says, "The inscription is much damaged. Only two letters are determinable, IC. Before the I are two scores across the angle, which, if a letter, would be G, but as there is a flake off the angle before it, it may have formed portion of an R. There is neither an L nor an F on the stone. Farther down is one score, but as the angle before and after it is damaged, it cannot be determined whether it is an M or a portion of another letter.'

PLATE XXVII. FIG. 1.

THE ROMAN MILIARY STONE IN THE SWANSEA MUSEUM.

In Col. G. Grant Francis's work on Neath and its Abbey we find a notice of another Roman miliary stone, found on the Via Julia Maritima, between Nidum (Neath) and Bovium (Boverton) near Pyle, which having been rescued from destruction by that gentleman, has