In this Lecture I shall deal with certain groups of relics which present in their forms and their decoration features which we have learned to recognise as distinctively Celtic.

About the year 1820 a singular object was found in a morass on the farm of Torrs, in the parish of Kelton, Kirkcudbrightshire. Having passed into the possession of Mr. Joseph Train, it was presented by him to Sir Walter Scott, and it still remains in the Museum at Abbotsford.^[55] It is of the form of an elongated mask (Fig. 92), somewhat resembling the frontal of a horse. It measures 10½ inches in total length, but the tip is apparently imperfect. Its breadth in a straight line across the lower margin is 3⅝ inches, and about 8½ inches on the round outside. Its greatest breadth in a straight line across the back is 6 inches, and 11 inches on the round outside, immediately above the insertion of the horns. At a height of 3 inches above the lower straight margin are placed two circular holes, one on each side, each measuring 2 inches in diameter. From between these eyelike holes, and a little above the level of their centres, two curiously curved, cylindrical, tapering horns spring close together on either side of the median line. The diameter of each of the horns at the base is 1⅜ inch, and they rise to a height of 8¾ inches to the top of the curve, the whole length of the perfect horns along the curve of the outer edge being 16½ inches. The horns are hollow, the whole object being formed of thin beaten bronze.

Its ornamentation is as peculiar as its form. It consists of a series of irregularly divergent spirals in repoussé work repeated symmetrically but not identically on either side of the median line of the front of the object. These spirals or scroll-like figures are formed of curves which are long and flattened, passing suddenly into curves of quicker motion, and ending in volutes. These curves, though proceeding in the same direction, do not proceed at parallel or regular distances from each other, but converge and diverge so as to enclose between them alternate spaces of varying extent of surface. The spaces enclosed between the curves are raised, and the spaces enclosed by their convolutions are flat, but the raised spaces are modelled so as to express the confluence of solid curves of the peculiar forms already indicated. These trumpet and spiral scrolls, as they are called, enclosing irregularly formed curvilinear spaces, and producing designs which are similar but unsymmetrical, are repeated in different varieties of pattern on the outer sides of the horns (Fig. 93). In the terminal convolutions of the scrolls the curves are sometimes arranged so as to produce a zoomorphic effect, which differs from the later zoomorphism of the metal-work of the Christian time and of the later manuscripts, in being more geometrical in form and character. The zoomorphic termination of the horns has also more of a geometric character than is usual in the Christian period.

The object being incomplete, its purpose is not obvious. But it is suggestive of the probability of its having formed part of a helmet that Diodorus Siculus, writing only a few years after the conquest of Gaul by Julius Cæsar, describes the military equipment of certain Gallic tribes as including “bronze helmets with lofty projections rising out of them, which impart a gigantic appearance to the wearers; for upon some are fixed pairs of horns, upon others the shapes of birds and beasts wrought out of the same metal.” These horned helmets are represented on some of the consular medals, and the whole description of the Gallic equipment is so similar to what we know of the habits of the Celtic tribes of Britain, that it may be concluded that in this respect their customs may not have been greatly dissimilar.^[56] And, in point of fact, there is in the British Museum a bronze headpiece found in the river Thames, near Waterloo Bridge, which, from its peculiar form, was at first considered to be a jester’s cap. But Mr. Franks has shown that it is a military helmet of native workmanship. It consists of a cap of thin bronze, with an additional plate at the back, decorated with scrolls of this peculiar character in low relief, among which are cross-hatched discs once coated with red enamel. From each side of the cap projects a conical horn terminating in a moulded button, and upon one side of the horn runs a string of small projecting studs.

It is therefore not improbable that this object at Abbotsford may have been the front part of a military helmet, or of a headpiece used for display. Such a headpiece with similarly large and curving horns, terminating in similar zoomorphic endings is seen (Fig. 94) on the head of a warrior who appears to be engaged in mimic combat with another accoutred as fantastically as himself, and whose grotesque headpiece bears a resemblance still more remarkable to another bronze object of the same character which I have next to describe. These representations occur on a bronze plaque dug up in the island of Oland, and they have therefore no necessary connection with the usages of the Celtic people. They merely show that in assigning such a purpose to these objects we are not attributing to them a purpose to which they were never applied. But the special use of the object is really of no great moment for the purpose of the present investigation. That purpose is fulfilled when we are enabled to say, from an examination of its special characteristics, that it has certain typical relations linking it with other objects, forming a distinct group and occupying a definite place in the series of types which characterise the area now termed Scotland. I therefore proceed to the description of other objects distinguished by the same characteristics.

At Liechestown, in the parish of Deskford, Banffshire, about the year 1816, a remarkable relic (Fig. 95), now in the Banff Museum, was found in a mossy piece of ground, at a depth of about 6 feet, and resting on a bed of clay at the bottom of the moss. This object, which is equally peculiar alike in respect to its form and ornamentation, is in the shape of a boar’s head of thin beaten bronze 8½ inches in length by 5½ in greatest breadth. The lower jaw is movable. The eyes are circular holes 1¼ inch in diameter. The whole head is formed of four plates of bronze, the snout, the palate, and the lower jaw (Fig. 96) having been each made separately, and attached to the posterior part of the head, which consists of an embossed plate bent to the shape. A disc-like plate, which was found with it, is now attached to the open back of the head, but does not quite fit, and it is doubtful whether it had been so placed originally. The ornamentation of this singular object is of the same character as that of the Torrs bronze, but simpler, being merely a series of trumpet-shaped ridges in repoussé work round the eyes. But this ornament, simple as it is, is quite sufficient to determine the relations of the relic to that general group of objects of which it and the bronze from Torrs are the most remarkable specimens.

It is obvious that if these objects had any relation to military equipment, we ought to find the very peculiar art which is so conspicuous in their decoration, also exhibiting itself in the decoration of the weapons and other war-gear in use among the same people. Diodorus, in fact, informs us that the Gauls used oblong shields as tall as the man, and painted after a peculiar fashion. Some of these shields, he also says, had figures of animals in relief of bronze, not merely for ornament but also for defence, and very well wrought. It has been already remarked that it is probable that the military equipment of the Gallic tribes resembled that of the British; and it is the fact that oblong shields, decorated with the peculiar patterns characteristic of the style of art exhibited by the two headpieces which have been described, having these patterns further adorned by coloured enamels, and also possessing the distinctive feature of figures of animals in relief in bronze, have been found in England.^[57]

No shields of this character have yet been discovered in Scotland, but there are other objects of a military kind which exhibit the same peculiar art in a sufficiently characteristic manner. One of these is a sword-sheath (Fig. 97) of bronze, 23½ inches in length by 1¼ inch in width, which was found at the foot of the Pentlands, near Mortonhall, and is now in the Museum. It is formed of thin beaten bronze; the ornamental cup-shaped expansions at the lower end are solid castings, and the ornamental strap carrying the loop in front is fastened on with pins. The back of the sheath is a thin slip of bronze sliding in grooves in the inner margins of the two sides. This is the only example of a sword-sheath of this style and period known to have been found in Scotland. Several sheaths of the same character have been discovered in England. Perhaps the most characteristic of these is one in the collection of Canon Greenwell at Durham, which exhibits, in a very special manner, the peculiar style of ornament of which I have given so many illustrations. The swords which these sheaths contained were of iron and have perished. One found in the Thames has the blade still within it, 3 feet 1½ inch in length, but a mere mass of oxide.^[58] These swords differ greatly in the length and form of the blade from the leaf-shaped swords of bronze which were in general use at an earlier period, and their sheaths differ still more widely in form and ornament from the sheaths of the leaf-shaped swords.

Another class of objects, which are more of the nature of harness-mountings or horse-furniture, also exhibit this peculiar style of ornamentation, in some cases combined with the remarkable feature of having their sunk spaces filled with coloured enamels.

A pair of massively-formed objects (Fig. 98), the precise use of which is not apparent, were found in a bank of clay on a spur of the Cheviots at Henshole on Cheviot. They are of cast bronze, and consist of an oblong body, hollow, rounded at one end and flattened at the other; the upper and lower surfaces inclined towards the small end, which is narrower than the width at the middle. A stout tang of about 2 inches in length is carried on a bar which crosses the open part of the small end, and the convexity of the larger end bears the mark of hammering as if to drive the tang home. They are destitute of surface decoration, but they seem to be allied by the characteristics of their form to other objects which are less indefinite in the indications of their art.

In a large cairn on the farm of Hillock Head, in the parish of Towie, Aberdeenshire, which covered an interment placed in a cist with an urn, there were found a number of bronze objects, all of which were lost except two (Fig. 99), which are now in the Museum. They are in the form of oval hollow rings, expanding on the inferior side, and having an oval opening in the under part, which shows the remains of an iron pin fastened at each side of the opening with lead. Their general appearance is suggestive of the mountings of horse-harness, but their precise purpose is not obvious, and the articles found in association with them are undescribed. Although the testimony is singularly defective on that point, it is not probable that they had any connection with the interment in the “short cist” which contained bones and an Fig. 100.—Mounting in Cast Bronze from Dowalton Loch (2 inches in diameter).
Fig. 100.—Mounting in Cast Bronze from Dowalton Loch (2 inches in diameter). urn. A similar object in bronze, also presenting the remains of iron fastenings in the lower part, was found under a large stone on the hill of Crichie, near Kintore, along with a number of globular balls of shale each about 1¼ inch in diameter, slightly flattened on one side, and having the remains of iron loop-like fastenings in the flattened side.^[59] A number of rings and harness-mountings found at Middleby, in Annandale, in 1737, and now preserved in Penicuick House, exhibit the same style of decoration in a more pronounced and characteristic manner.^[60]

A mounting in cast bronze (Fig. 100), 2 inches diameter, the sunk spaces of which had probably been filled with enamels, was recently found in the dry bed of the loch of Dowalton, which was drained about eighteen years ago. It is formed of a combination of segmental spaces, the curves of which are those of the divergent spiral, each space being surrounded by a raised border, and the sunk surfaces roughened with a tool.

A bridle-bit (Fig. 101), found in a moss at Birrenswark, in Annandale, before 1785, and now in the National Museum, exhibits the characteristics of this peculiar phase of art in a very striking manner. It is no less peculiar in its design and construction than in the character of its ornamentation. It is a single casting of bronze. The loops of the cheek-rings have been cast within the loops of the centre-piece, an operation implying technical skill and experience of complicated processes of moulding and casting. The design, however, is the most remarkable feature of the object. It is designed as carefully as if it were a piece of jewellery. The rings, though cast in one piece with the loops, are penannular in form, grasping the neck of the loop between their expanded ends. The two rings differ slightly in size, and the loops differ greatly in form. The one is treated as a loop formed of a cylindrical rod bent to the shape of a loop, and carrying the ornamented open-work of its terminal part as between its extended ends. The other loop is treated as a solid form, and in its ornamental termination there is no open work. The two rings are similar, but not identical. The idea of openness suggested by the modelling of the one loop is carried into the construction of the terminal portions as open work, and the idea of solidity is similarly carried out in the other loop. The surface decoration of the terminal portions of the loops is of the same character in the parts of both that are similar, and is partly carried also into the parts of the one which are wanting in the other. It consists of red and yellow enamel champléve, the colours alternating in alternate rows of triangular and oval spaces. A double spiral and trumpet pattern appears in the open work of the one loop. The loops and rings are greatly the worse for wear, and have been strengthened by thin pieces riveted on.

It is certainly a peculiar feature of an art so singularly decorative that it was applied so largely to the ornamentation of objects that were appropriated to the commonest uses. Enamelled horse-trappings of the most finished and beautiful workmanship have frequently been found in England, sometimes associated with the remains of chariots.^[61] Not only is the use of enamel in the decoration of such objects unknown beyond the area of the British Isles, but the special system of design which accompanies its use is also confined within that area. And it is an interesting fact that there is historical evidence as to the nationality of these remains. The only classical author who mentions the art of enamelling is Philostratus, a Greek sophist in the household of Julia Domna, wife of the Emperor Severus. In a notice of the variegated trappings of the horses in a painting of a boar-hunt he accounts for their peculiar appearance as follows:—“They say that the barbarians who live in the Ocean pour such colours on heated brass, and that they adhere to it, become as hard as stone, and thus preserve the designs that are made in them.” It is matter of inference what people they were who are thus styled “barbarians in the ocean,” but it is matter of fact that horse-trappings of bronze (or brass) decorated with coloured enamels have hitherto been found in the British Islands alone.

But this peculiar style of art was not confined to the decoration of such objects as parts of military equipments or harness of horses. It was largely employed in the decoration of personal ornaments and objects of personal use.

In the parish of Balmaclellan, in Kirkcudbright, a number of bronze articles were found in draining a bog. It is stated that they were found about 3 feet under the surface in four parcels, each wrapped in coarse linen cloth. Close by them the upper stone of a quern was also found. The quern stone (Fig. 102) is ornamented, but the ornament possesses none of the distinctive features of the decoration of the bronzes. They consisted of a circular mirror with handle, and a number of thin plates of bronze, some being long narrow bands, others curved and cut into various shapes. The mirror (Fig. 103) is of the form so commonly seen on the sculptured monuments of the Celtic Christian time in Scotland. The circular part is 8 inches in diameter, and the handle 5 inches in projection from the circumference of the circular part. Fig. 103.—Bronze Mirror found at Balmaclellan (8 inches in diameter)
Fig. 103.—Bronze Mirror found at Balmaclellan (8 inches in diameter) The body of the mirror is a thin plate of bronze, surrounded by a plain-rolled edging. The handle, which is also a thin plate of bronze similarly edged, is attached to the circular plate by rivets, and the junction is concealed by a finely-ornamented plate (Fig. 104), presenting a pattern composed of those peculiar raised surfaces formed by the meeting of curves rising from the flat at different angles, and traversing the ground also in curves, which converge and diverge in a manner pleasing to the eye, but difficult to describe. The upper part of this ornamental plate is tri-lobate, the lobes bounded by curves of peculiar form, and bordered by an edging of studs embossed on the metal. Fig. 104.—Ornamental Plate of thin bronze, embossed, at the junction of the mirror with its handle (actual size).
Fig. 104.—Ornamental Plate of thin bronze, embossed, at the junction of the mirror with its handle (actual size). The central ornament of each lobe is a circular device, with a central boss surrounded by a circle of oval-raised surfaces, and presenting a nearer approach to the effect of a floral decoration than is usually seen in this style of ornament. The handle of the mirror is pierced with three segmental openings formed of the curves of the divergent spiral. A crescentic collar-shaped plate of bronze (Fig. 105), 13 inches in diameter, and 2 inches in the width of the band, is decorated with a chased pattern of similarly convergent and divergent curves, the spaces enclosed by the curves being hatched with parallel lines. Fig. 105.—Half of the Crescentic Collar-like Plate of Bronze found with the Mirror at Balmaclellan.
Fig. 105.—Half of the Crescentic Collar-like Plate of Bronze found with the Mirror at Balmaclellan. The remaining plates (Fig. 106), of which there are a considerable number, are of various forms. Some have straight outer edges, and the interior edges cut into curves, meeting each other with long and short points; others are triangular pieces, with one convex and two concave edges, while others again are long narrow bands with straight edges. They are all bordered with an edging of thin metal doubled over and pinned on, and they seem themselves to have been attached by pins to some object of a more perishable nature. What their precise purpose was—whether they were mountings on wood or leather, or whether they formed parts of some object constructed wholly of thin plates of metal (as the two objects previously described are constructed)—it is not necessary to conjecture since the form and condition of the objects themselves give no definite indications on these points. Their being wrapped in cloth in separate parcels may imply that they are not all parts of the same object, and their local association with objects of such incongruous purposes, as a mirror and a quern, may imply that they were not necessarily even associated with each other when in use. There is no evidence that the deposit was in any way connected with sepulture, although the mirror of this form, and bearing precisely the same kind of ornamentation, has been found associated with interments of Pagan time in Britain.

At Mount Batten, near Plymouth, a series of graves were discovered in 1865,^[62] which presented phenomena of a very peculiar character. They were pits 4 or 4½ feet in depth, one foot of which only was soil, the remaining three feet being sunk in the disintegrated surface of the underlying rock. They were very numerous, sometimes close together and irregular in form, and had mostly been refilled with the materials removed in making them. They contained fragments of pottery of black and yellow ware, and wheel-made. Some fragments of glass vessels, portions of iron implements, among which were a pair of shears, bronze rings and fibulæ, and jointed armlets of bronze, with a knife or dagger in a sheath of thin bronze, were also found. But the most interesting part of the discovery was the circular plate of a bronze mirror (Fig. 107), 8 inches in diameter, which lay on its face at the bottom of one of the graves. It is a very thin plate of bronze, with a rolled edging. The back is ornamented with three circular engraved patterns of spirals formed of the same peculiar curves, converging and diverging, the spaces between the lines forming the curves being filled with hatching. So closely do the patterns resemble those on the collar-like object from Balmaclellan, and so similar is the style of the work, that the conclusion is unavoidable that the two objects belong to the same school of art, and cannot be very far apart in time.

Another mirror, which is almost precisely similar in form and ornamentation, was found in 1833 at Trelan Bahow,^[63] in the parish of St. Keverne, Cornwall. In the course of the construction of a new road a group of graves was discovered. Each grave was formed of six slabs set on edge, two forming each side of the grave, and one at each end. They were from two to three feet under the surface, and covered with large stones. In one of the cists, apparently with the remains of a female, there were found the bronze mirror almost perfect, some rings of bronze or brass, fragments of fibulæ, and other personal ornaments, and several beads of variegated glass. The mirror is circular, 6 inches in diameter, with a looped handle 2½ inches in length. The back of the mirror plate has a marginal ornament of triangular spaces alternately plain, and filled with short parallel lines struck by a punch. Across the central line of the mirror are two circles enclosing smaller circles and curvilinear spaces alternately plain, and filled with punched lines in a style similar to that of the ornament on the collar-like object from Balmaclellan.

Another mirror of the same character, found at Birdlip on the edge of the Cottiswold Hills, near Gloucester, in 1879,^[64] exhibits the same style of ornamentation. Three cists were discovered in a group, containing skeletons placed with their feet to the south. The first and third were apparently adult males, and with them no manufactured objects were found. The second was apparently a female. On the face of the skeleton was placed a large bronze bowl, 9 inches in diameter, inverted; and among the other contents of the cist were a smaller bowl of bronze, 4 inches in diameter, a harp-shaped fibula of silver plated with gold, a bracelet and four rings of brass, a key-handle, a knife-handle terminating in the head of an animal, a string of large beads of jet and amber, and a mirror made of a massive bronze plate, weighing 38¼ ounces. The back of the mirror (Fig. 108), which is of a slightly oval form, measures 10⅝ inches in its greatest, and 9¾ inches in its least, diameter, and is beautifully ornamented with a triple scroll-like pattern of flowing curvilinear spaces filled with hatchings of short lines in chequers, or groups disposed at right angles to each other. The pattern is so managed that the hatched spaces and the plain spaces alternate and form symmetrical arrangements, producing a pleasing effect. At the lower part, where the handle supports the mirror, is a triple arrangement of trumpet-shaped scrolls in relief, enclosing spaces which are similarly decorated. The handle is elegantly formed from a prolongation of the marginal beading of the mirror, which gradually thickens towards the lower margin to trumpet-shaped endings on either side of the handle, which takes the form of a double-loop, drawn out from the marginal bead, and terminating in a ring partly filled by an ornamented disc.

These mirrors all differ in their form and in the composition of the metal from Roman mirrors, and they differ in certain characteristics of their ornament still more widely from the Roman style. But the peculiar characteristics which form the special features of their decoration are identical with those of a large class of objects which we have now learned to recognise by the character of their art as distinctively Celtic.

The same character is exhibited by the ornamentation of a series of spoon-like objects^[65] found in England and Ireland, of which Fig. 109 is a characteristic example. Four of these are in the National Museum, and though no specimens have yet been met with in Scotland, I notice them here, because their decoration is so nearly related to that of the Scottish school. In the case of the pair of these peculiar objects found in excavating for a quarry at Weston, near Bath, the backs of the circular projections or handles (Fig. 110) are ornamented with patterns of this character in relief. The front of the disc is ornamented with a series of circular concentric mouldings, and the bowl of the spoon is quartered by incised lines. It is a peculiarity of these objects, that though found in pairs, the two members of the pair, though similar, are not identical. In some cases it is apparent that they have even been cast in different moulds. Usually one of the pair has its bowl quartered by incised lines, while the other has a small hole pierced near the edge of the bowl. Another pair, also in the Museum, were found in 1861 in a railway cutting in Llanfair parish, Denbighshire. They are slightly smaller in size, and differ in the ornamentation of the front of their discs. One of them (Fig. 111, No. 2) is here shown along with the second of the Weston specimens.

The same characteristic style of art is seen in the decoration of a massive collar of cast bronze (Fig. 112), which was found in digging a well at Stitchell, in Roxburghshire, in 1747, and is now in the National Museum. Like the armlets found in the Plymouth graves, this collar is jointed, opening on a hinge in the centre, and fastening in front by a pin and socket. It is a very massive and heavy ornament, the width of the opening being 6 inches by 5, and the breadth of the flattened ring varying from 1¾ inch to ¼ inch. The character of the ornament is simple, but highly peculiar, and bearing a strong family likeness to the double escaping and divergent spirals of the later Celtic art. All the patterns are in relief and cast in the solid, except those on the two panels on either side of the central opening, which are in repoussé on a thin plate of bronze fastened to the collar by pins at the four corners.

Closely akin to this jointed collar in the idea of its construction and the form of its ornament is an elegant armlet of thin bronze (Fig. 113), found in 1826 near Plunton Castle, in the parish of Borgue, in Kirkcudbright. It is of thin beaten bronze, 1½ inch wide and 2½ inches in diameter, and, like the collar, it is made to open on a hinge in the centre, and close by a pin and loops. It is ornamented by three raised mouldings, beaten up from the back, which pass round it horizontally, but these are concealed on either side of the hinges by two plates of thin bronze of quadrangular form, ornamented in repoussé by trumpet-shaped ornaments connected by peculiar curves, and having studs placed in the concavities of the curves. These plates are fastened to the armlet at the four corners by pins, and bordered by a single row of small studs.

In the month of March 1806 a herd boy, passing along the side of the Shaw Hill, near the House of New Cairnmuir, in the parish of Kirkurd, Peeblesshire, saw something glitter in the ground, and on scraping the place with his foot he unearthed a hoard of gold objects, consisting of two twisted arm-rings, each weighing 8 oz. 12 dwt., a broken ring of the same form weighing 8 oz. 10 dwt., forty small studs, each weighing about half a sovereign, and a hollow spherical ornament weighing 4 oz. 5 dwt.—the bullion value of the whole being about £110. One of the twisted arm-rings passed into the possession of Sir George Montgomery of Macbiehill; the spherical ornament and two of the small studs were obtained by Mr. John Lawson of Cairnmuir and placed in the National Museum; the rest of the hoard is believed to have been melted. The three arm-rings are spirally twisted rods of gold, with flat circular ends bent round to encircle the arm. The studs or pellets are nearly spherical, about the size of a large pea, and marked on the surface with a cruciform ornament in relief. The spherical ornament (Fig. 114) has some resemblance to the pommel of a sword, although its form gives no obvious indication of its purpose. It is 2½ inches in length by 2 inches in width, and about 1¼ inch in thickness. It has been cast hollow, with an opening through the centre of the rounded part, and must have been made by a very skilful workman. One side of it is plain, the other ornamented in repoussé work of great beauty. The style of the ornament is simple, elegant, and highly effective. The surface to be decorated is broken up into irregular spaces by a system of the peculiar curves, which are so characteristic of the style of art of the bronzes which have been already described. Some of these spaces are further ornamented by a peculiar pitting of the surface seen in some of the decorated stone balls (Fig. 146); others are raised in solid curves of the same peculiar form, while the interspaces follow the form of the object itself. Studs and prominences, with spirals in relief, are introduced to give emphasis to the general design, which commends itself at once to the eye of taste as one of the most fitly beautiful and unaffected forms of surface-decoration which could be applied to such a purpose.

In this group of objects in bronze and gold we have characteristic examples of the work of this early school of decorative art, which in some of its features bears certain relations to the work of the later school of Celtic art of the Christian time. But the elements of its decoration are fewer. It has no interlaced work and no fret—nothing but curves and spirals. It does not systematically break up its surfaces in panels, but distributes its decorative effects in spaces that are circular or oval, or bounded by intersecting curves. Its prevailing features are not the production of intricately symmetrical and geometrically regulated patterns, but the production of effects of balance and beauty by the rhythmic recurrence and variation of curves and spaces with solid forms which, though not symmetrical, are similar. Their characteristic curves, as seen in the outlines of their figures and the sections of their solid forms, are specially peculiar, while the marked preference for relief in metal-work is in striking contrast to the general prevalence of chased and engraved designs in the later school.

It is to this characteristic treatment of the decoration of their metal-work by this early school of Celtic art that Mr. Kemble refers in the following remarks:—“When, as is often the case in metal, this principle of the diverging spiral line is carried out in repoussé—when you have those singularly beautiful curves, more beautiful perhaps in the parts that are not seen than in those that meet the eye, and whose beauty is revealed in shadow more than in form—you have a peculiar characteristic, a form of beauty which belongs to no nation but our own, and to no portion of our nation but the Celtic portion. It deals with curves which are not arcs of a circle; its figures are not of the class we usually designate by the term geometrical; and above all it calls in the aid of enamel to perfect its work—not cloisonné like the enamel of the East; not mosaic work of tesseræ like so many so-called enamels of the Romans, but enamel champléve as Philostratus has described the island barbarians to have invented it. The engraved spiral line, with double winding, is found from America to the Baltic, from Greece to Norway, but the divergent spiral repoussé in metal and ornamented with champléve enamel, is found in these British Islands alone.”alone.”

I now proceed to notice another group of objects in metal possessing peculiar features still more strongly marked, but exhibiting also the distinctive characteristics of the same style of art.

A pair of these objects were found imbedded in the earth over the entrance to a curious underground structure in the garden at Castle Newe in Aberdeenshire. The structure was a long narrow curved subterranean gallery about 50 feet in length and 7 feet wide on the floor. What remained of the walls was only 4½ feet high, but showed that it had been roofed over by bringing the walls gradually towards each other as they increased in height, till the space could be covered with flat stones of moderate length. Fig. 115.—Bronze Armlet, with enamelled ornaments (one of a pair), found at Castle Newe, Aberdeenshire. Front view (5¾ inches in diameter).
Fig. 115.—Bronze Armlet, with enamelled ornaments (one of a pair), found at Castle Newe, Aberdeenshire. Front view (5¾ inches in diameter). This form of structure, as we shall see in a subsequent Lecture, is typical, and extends over the Celtic area. The pair of objects found in association with this typically Celtic structure are of quite a remarkable character. They are massively formed, but highly decorated objects of cast bronze. It is obvious from their form and decoration that they are designed for an ornamental purpose. It is impossible that they could have been worn as personal ornaments either with comfort or convenience, but that impossibility does not necessarily invalidate the conclusion that they were personal ornaments, because such things have been worn in all ages, although they have entailed discomfort and inconvenience to the wearers. The special form of the objects and the circumstance that a pair of them were found together are suggestive of their use as armlets. Their form, as shown in Fig. 115, is the typically Celtic form—penannular, with rounded and slightly-expanded ends. These terminal expansions have circular spaces in the centre, bordered by a double raised edging, and filled with plaques of bronze Fig. 116.—Bronze Armlet, found at Castle Newe, Aberdeenshire. Back view (5¾ inches in diameter).
Fig. 116.—Bronze Armlet, found at Castle Newe, Aberdeenshire. Back view (5¾ inches in diameter). ornamented with chequered patterns of red and yellow enamels. These bronze plaques are fixed in their places by iron pins. The body of the armlet (Fig. 116) is divided longitudinally into three distinct ridges or bands with convex surfaces, separated by narrow bands of a tooled chevrony ornament, which lie along the furrows between the ridges. At intervals there rise from the ridges solid, flattened, and curvilinear projections of about ¾ inch in length, placed obliquely across the ridges, and standing in rows from side to side of the armlet. These are connected longitudinally by less highly raised trumpet-shaped scrolls, slightly curved, and passing obliquely across till they meet in the centre. The median ridge stops short at the circular spaces in the terminal expansions, while the exterior ridges on either side pass round to form the border of the expansion on which the projecting ornaments are continued in a less pronounced form. The general contour of the armlets is that of an oval slightly compressed from front to back. Their greatest diameter is 5¾ inches, their greatest depth 4½ inches, and the weight of each is 3¾ lbs. They do not commend themselves to our notions of elegance and comfort as articles of personal decoration, but they possess a strong individuality of character, combined with an ingenious and highly-effective style of decoration which is not met with on any other class of objects in metal.^[66]