THE MATERIAL OF WHICH PATTERNS ARE MADE. (continued 2)

Fig. 72 is a detail taken from a plate in the second volume of the magnificent atlas by Prisse d’Avennes;[69] it is part of the offerings on an altar before Osiris, who is crowned with the solar disc. Osiris is the sun in the Lower World—i.e., during the night, and the father of Horus. Horus is sometimes depicted seated on a lotus.

The various animals which were symbolic of the sun or associated with sun-divinities are also placed in direct connection with the lotus, as if to emphasise its solar significance; for example—

The solar-bull is well recognised in Egyptian mythology, the Bull-god Apis being an incarnation of Osiris, and an offspring of the Sun-god, Ptah of Memphis. Similarly also for Assyria, Merodach, “the Bull of Light,” was originally a Sun-god; his Syrian equivalent was Baal. The Phœnician Moon-goddess, Astarte, had the bull as her symbol, and the bull of Europa was its counterpart. The Taurus of the Chaldean Zodiac commenced the year.

The lion was another sun-animal both in Egypt and in Chaldea and Assyria.

Among birds the hawk and the eagle were sun symbols, especially the former, and it is sometimes depicted standing on a lotus. The solar-goose is also important in its association with the lotus. (Fig. 129, G.)

In early Cyprian pottery we find lotus derivatives grouped with the solar cross and other symbols of the sun. (Fig. 129, F.)

The association of the lotus with the sun probably led to its other symbolic relations, and these latter have rather drawn attention away from what is here regarded as the more primitive symbol.

The lotus was a well recognised symbol of life, resurrection, and immortality. It was largely employed in funeral rites in Egypt, and is constantly associated with mummies, and also symbolised the resurrection, but this latter idea was associated in the Egyptian mind with reproductive power, and hence the relation of this also to the lotus. Professor Maspero says:[70] “The assimilation and occasional complete identity of the Supreme God with the sun being once admitted, the assimilation and complete identity of the secondary divine beings with Ra (the sun) were a matter of course. Amon, Osiris, Horus, Ptah, were regarded sometimes as the living soul of Ra, sometimes as Ra himself.” From this would result a mingling and extension of symbolism; but upon these troubled waters the lotus calmly rides supreme. Its association with the sun, its connection with reproductive energy, its descent into the grave, and its symbolism of a resurrection have given to the lotus that immortality which it symbolised.

Although lotus designs are profusely scattered up and down in Egyptian art there is no reason for believing that the Egyptians regarded it as a national emblem, but it was a universally recognised symbol. At the beginning of the year it sprouted from its slimy bed and floated beautiful and pure on the surface of the waters. At sunrise the buds opened and studded the water with white or cerulean asters, which closed when night fell. Every autumn it died its annual death only as prelude to the vernal resurrection.

The intensely religious mind of the Ancient Egyptians was permeated with the problems of death and elevated by the prospect of immortality. Resurrection and future bliss were articles of firm faith, not merely a pious hope. What wonder then, with this religious saturation of immortality, that the flower which symbolised the resurrection should be depicted in such profusion in their tombs and elsewhere!

If the reader will take the trouble to compare lotus representations in books on Egyptology it will be beyond dispute that it is the white or blue lotus (Nymphæa), and not the rose water-lily (Nelumbium), which is so ubiquitously delineated.

Fig. 73.—Lotus flower with two leaves, on a vase from the Necropolis of Memphis, Fourth to Fifth Dynasties; after Prisse d’Avennes.

A slightly conventionalised lotus with two of its leaves (Fig. 73) is drawn on a vase contemporaneous with the pyramids, from the Necropolis of Memphis (Fourth and Fifth Dynasties, 3998-3503 B.C.).

The same lotus flower (Fig. 72) appears some two thousand years later in a representation of an offering to Osiris from the Necropolis of Thebes belonging to the Twentieth Dynasty. Indeed, it was painted and carved so frequently for thousands of years that it would be impossible to describe its variations and applications. I must, however, permit myself to allude to one or two examples which are interesting from other points of view. In Plate VIII., Fig. 12, we see single lotus flowers employed in an isolated manner in a border pattern, and alternating with these is another device. The separation of the elements of a border pattern is by no means universal in Egyptian decorative art; for example, the scroll pattern (Fig. 74) from the Necropolis of Thebes is a good example of a pattern which gives an idea of flow, but even here there is a lack of continuity in the spiral band which creates a feeling of dissatisfaction when one attempts to trace out the construction of the design. It is evident that in such patterns the spiral is quite a secondary motive, and it thus has not been worked out logically; the lotus flowers and the rosettes are the essential elements of the pattern.

Fig. 74.—Lotus border; from Goodyear, after Prisse d’Avennes.

With the last figure we may compare the scroll detail (Fig. 75) from a Melian vase, the lotus flower being represented by four black marks, and the scroll has acquired that development which is so characteristic of Ægean art.

Fig. 75.—Lotus scroll detail on a Melian vase; from Goodyear, after Conze.

Various causes may lead to the evolution of a recognised scheme of decoration of certain objects, but when a new class of objects is to be decorated the artist has a chance to exhibit his originality; even so this is about the last thing which decorative artists do manifest. The constraint of custom appears to exert an influence too potent to be readily snapped, and so the Egyptian decorator, being further tied by religious sentiment, ornamented even extensive areas, such as the ceilings of tombs, with lotus designs, the main elements of which had been elaborated elsewhere.

Fig. 76.—Pattern from the ceiling of a tomb, Necropolis of Thebes, Eighteenth Dynasty; from Coffey, after Prisse d’Avennes.

In Fig. 76 we have a ceiling design in which the lotus is very apparent both in flower and bud; the rosettes, like spiders’ webs, may possibly represent the leaves of the lotus (Fig. 70), and we have the same interlocking but discontinuous spirals that occur in Fig. 74.

A different treatment of the same motive is seen in Fig. 77, but here only the lotus flowers and the interlocking scrolls are employed. Below each flower is a fan-like portion apparently tied on to the former; this may have some significance or it may be merely a convenient method of finishing off the flower.

Fig. 77.—Pattern from the ceiling of a tomb, Necropolis of Thebes, Eighteenth to Nineteenth Dynasties; from Coffey, after Prisse d’Avennes.

In these old Egyptian designs the rosette is often associated with the lotus and lotus derivatives, as in Fig. 74; and it may happen, as in Fig. 78, that the former is the most prominent motive. The lotus is here represented solely by the black triangles which occupy the angles of the quadrangular spaces which contain the rosettes; at all events there is good evidence to support this view.

The angularisation of the last pattern gives us Fig. 79, which many people would imagine to be Greek, although, as a matter of fact, it is ancient Egyptian. The rosettes and the angled scrolls alone persist. We cannot speak with certainty as to the reason for the modification of the scrolls, but it is probable that it resulted from an attempt to copy such a painted design as Fig. 78 in textiles, and the pattern metamorphosed by the new conditions was painted on the tomb ceilings along with its more flowing progenitor. For further examples of analogous transference of designs from one technique to another, and their consequent transformation, the reader is referred to p. 112.

Fig. 78.—Pattern from the ceiling of tomb No. 33, Abd-el-Kourneh, Thebes, Seventeenth to Twentieth Dynasties; from Coffey, after Prisse d’Avennes and Goodyear.

Fig. 79.—Pattern from the ceiling of a tomb from Thebes, Seventeenth to Twentieth Dynasties; from Coffey, after Prisse d’Avennes.

Professor Flinders Petrie has stated that the scroll or spiral was one of the greatest factors in the early development of ornament, and only second to the lotus in the part it played in the decorative ideas of the ancient world. What it symbolised, if symbolise anything it did, we know not. Some affect to see in it a representation of the wanderings of the soul, but why, as Professor Petrie suggests, some souls should come to the end of their wanderings in a spiral and others in an oval is not explained. Its oldest use was on the scarabs, where it was clearly used first as “filling-in” ornament. We can first trace it about 3,500 B.C. At first in loose unconnected “C” and “S” links, and afterwards in every variety of combination, continuous as well as unconnected, the scroll line winds its way for ages through the records of Egyptian decoration. Yet there is a clear margin of 1000 years at least between any Egyptian date of its use and its appearance in the art of other ancient countries. From the fact that it is generally coloured yellow in Egyptian designs, Professor Petrie infers that gold was used in these forms to enclose gems, cloisonné and coloured stones; indeed Schliemann found such work in his explorations at Mycenæ.

Mr. Arthur Evans remarks:[71]—“On the twelfth dynasty [about between 2778 and 2565 B.C.] scarabs the returning spiral motive, as is well known to Egyptologists, was developed to an extraordinary degree. These purely spiral types, like the twelfth dynasty motives, were also copied by the native Cretan engravers. From Crete, where we find these Aegean forms in actual juxtaposition with their Egyptian prototypes, we can trace them to the early cemeteries of Amorgos, and here and in other Aegean islands like Melos can see them taking before our eyes more elaborate developments. Reinforced a thousand years later by renewed intimacy of contact between the Aegean peoples and the Egypt of Amenophis III., the same system was to regain a fresh vitality as the principal motive of the Mycenæan goldsmith’s work. But though this later influence reacted on Mycenæan art [about 1500 B.C.], as can be seen by the Orchomenos ceiling, the root of its spiral decoration is to be found in the earlier ‘Aegean’ system engrafted long before, in the days of the twelfth dynasty.

“In the wake of early commerce the same spiraliform motives were to spread still further afield to the Danubian basin, and thence in turn by the valley of the Elbe to the Amber Coast of the North Sea, there to supply the Scandinavian Bronze Age population with their leading decorative designs. Adopted by the Celtic tribes in the Central European area, they took at a somewhat later date a westerly turn, reached Britain with the invading Belgae, and finally survived in Irish Art.”[72]

Among the most frequent of the decorative designs employed by the Assyrians are the knop (or bud) and flower pattern and the rosette, and usually these are found in combination. For the former design I shall employ the Greek term “Anthemion.”

“That flower,” write MM. Perrot and Chipiez,[73] “has been recognised as the Egyptian lotus, but Layard believes its type to have been furnished, perhaps, by a scarlet tulip which is very common towards the beginning of spring in Mesopotamia.[74] We ourselves believe rather in the imitation of a motive from the stuffs, the jewels, the furniture, and the pottery that Mesopotamia drew from Egypt at a very early date through the intermediary of the Phœnicians. The Phœnicians themselves appropriated the same motive and introduced it with their own manufactures, not only into Mesopotamia but into every country washed by the Mediterranean. Our conjecture is to some extent confirmed by an observation of Sir H. Layard’s. This lotus flower is only to be found, he says, in the most recent of Assyrian monuments, in those, namely, that date from the eighth and seventh centuries B.C., centuries during which the Assyrian kings more than once invaded Phœnicia and occupied Egypt.[75] In the more ancient bas-reliefs, flowers with a very different aspect—copied in all probability direct from nature—are alone to be found.

“The lotus flower is to be found, moreover, in monuments much older than those of the Sargonids, but that does not in any way disprove the hypothesis of a direct plagiarism. The commercial relations between the valleys of the Nile and the Euphrates date from a much more remote epoch, and about the commencement of the eighteenth dynasty the Egyptians seem to have occupied in force the basin of the Khabour, the principal affluent of the Euphrates. Layard found many traces of their passage over and sojourn in that district, among them a series of scarabs, many of which bore the superscription of Thothmes III. [1481-1449 B.C.]. So that the points of contact were numerous enough, and the mutual intercourse sufficiently intimate and prolonged, to account for the assimilation by Mesopotamian artists of a motive taken from the flora of Egypt, and to be seen on almost every object imported from the Nile Valley. This imitation appears all the more probable as in the paintings of Theban tombs, dating from a much more remote period than the oldest Ninevite remains, the pattern with its alternate bud and flower is complete (Plate VIII., Fig. 12).

“The Assyrians borrowed their motive from Egypt, but they gave it more than Egyptian perfection. They gave it the definite shapes that even Greece did not disdain to copy. In the Egyptian frieze the cones and flowers are disjointed; their isolation is unsatisfactory both to the eye and the reason. In the Assyrian pattern they are attached to a continuous undulating stem, whose sinuous lines add greatly to the elegance of the composition.”

While admitting that the lotus motive overran Assyrian art, there is reason to believe that it did so only because there was an antecedent style upon which it could be engrafted. The pattern shown in Fig. 10, Plate VIII., is an example of an Assyrian anthemion engraved on an ivory panel in the British Museum, and of purely Assyrian workmanship. It is worth while attempting to trace this back as far as possible. In Fig. 4, Plate VIII., we have a pattern painted in red, blue, white, and yellow upon plaster, discovered by Sir Henry Layard in Nebuchadnezzar’s palace in Nineveh. In this there is a serial repetition of a disc, or sphere, which is pendant; all the pendants are connected by a single cord, which appears as if it were drawn into loops by their weight.

In Fig. 7, Plate VIII., we have a representation from a stone carving of an Assyrian pavilion, and in Fig. 2 a “tabernacle” from the famous bronze gates of Balawat, which were made for Shalmaneser II., and are now in the British Museum. Yet more simple is the tasselled canopy (Fig. 6) from an enamelled brick from Nimroud, a king who is standing under this canopy has a fringe (Fig. 5) to his robe which is composed of alternate white and yellow tassels. King Sargon (about 722 B.C.) is also represented on a relief from Khorsabad in the Louvre, with a similar fringe (Fig. 1) to the hem of his robe.

Any one who has done any plaiting in bands of two colours knows that if the intersections be truly alternate the fringe along the opposite borders will all be of the same colour as in A, Plate VIII., but if the colours run in stripes the fringe all round will be composed of alternate patches of colour. When bands composed of several threads are employed, it is necessary to knot the strands together at the edge to prevent fraying. A more pleasing border is formed by taking half the strands of one band and tying them to half the strands of the next band of the same colour, and so on (B, Plate VIII.). By this means we naturally obtain a structural root-like origin for each tassel in the fringe, which may be termed the connecting strand. This appears to have been the common method of finishing off the edge of Assyrian textiles.

There is thus no difficulty in accounting for a fringe of tassels (Figs. 1, 5, Plate VIII.). Awnings (Fig. 6) as a protection from the blazing sun were a very common feature in Assyrian life. When the king went out on warlike or hunting expeditions he took with him a large royal tent or pavilion made of “slender columns with rich capitals and a domed roof, made, no doubt, of several skins sewn together, and kept in place by metal weights.[76] The pavilion (Fig. 7) was a civil edifice, the temporary resting-place of the sovereign. The same materials were employed in the same spirit in the erection of religious tabernacles” (Fig. 2). It is, however, probable that brightly-woven rugs or mats were employed for the smaller canopies; these would even more require the employment of weights to prevent the wind from blowing about the covering. One can hardly interpret the pendants on the royal pavilion (Fig. 7) in any other manner than as weights to steady the awning. The pendants would in the case of textiles be fastened on to the tassels, probably they would sometimes be placed on alternate tassels. In the pavilion so often referred to the weight pendants are of two shapes, in this also carrying out that alternate arrangement which manifests itself structurally in most textiles, and which consequently gives rise to the feeling of expectancy in other objects. Another example of this is seen in the representation of the vine in Assyrian art, for the decorative sentiment has so possessed the artists as to cause them to depict the branches with a leaf and a bunch of grapes in regular succession.

There is no need to go further than this for the origin of the Assyrian anthemion. We find a fringe of tassels in alternate colours, we find a fringe of canopy weights of alternate design, we assume an occasional alternation of fringe and weight. In all cases these must be serially united by the “connecting strand.” How can the stone-carver or the wall-decorator represent these three alternatives? Clearly they would indicate rather than imitate them. What greater realism could we expect than that which we have?

There are many ways of making tassels—for example, each one may be allowed to splay out fan-wise, or it may be tightly tied round the middle, or bound round so as to form a kind of cone or spindle.

Whether as variously tied, or differently coloured tassels, or as alternate tassel and weight, a border of alternate members organically springing from a common base was constantly before the sight of the artists of this great textile manufacturing people. The conventionalising tendency of decorative art did the rest, and the various forms of Assyrian anthemion would easily follow.

A triple alternation (Fig. 9, Plate VIII.) occurs on an enamelled brick tile from Nimroud in the British Museum. It is characteristically Assyrian in style, but it does not give that effect of repose and satisfied expectancy which we demand from a pattern, and in this respect we cannot regard it as eminently successful.

If this hypothesis of Dr. March’s of the evolution of the Assyrian anthemion be correct, this pattern is essentially a skeuomorph, but at the same time certain local plant-forms were probably associated with it.

Let us now turn to the border pattern (Fig. 8, Plate VIII.) of the carved stone thresholds, which are occasionally found in a marvellous state of preservation. Here we have a “knop and flower pattern” which differs as much from the Assyrian style as it resembles that of Egypt. A comparison of this figure with Fig. 12, Plate VIII., will convince most people that borrowing has taken place. It is not always easy to determine how far the Assyrian anthemion has been influenced by native foliage or by conventional designs derived from the local flora. In these threshold borders, however, the Egyptian phyllomorph has grown, as Dr. March points out, like a floral parasite on a skeuomorphic basis. As introduced plants frequently overrun a new country and crowd out native forms, so the lusty lotus invaded the field of Assyrian art, and largely supplanted pre-existing phyllomorphs.

To return for a moment to the Egyptian pattern, the “proto-anthemion,” as one may term it, is characterised by the absence of a connecting strand, the buds and flowers springing from a basal line. My friend, Dr. March, with his usual ingenuity, has suggested to me a very plausible explanation of this fact. The Egyptian pattern was phyllomorphic from the beginning, originating in symbolism it was primitively a realistic representation of an erect water-plant.

Maspero says the decoration of each part of the Egyptian temple was in consonance with its position. The lower parts of the walls were adorned with long stems of lotus or papyrus—bouquets of water-plants emerging from the water.

This then is the solution of the difficulty. The Egyptian anthemion, derived from plants emerging from the water, has as a rule no connecting strand. The Assyrian variety, derived from a tassel-skeuomorph, is never without its looped base line, is primarily pendant, and consists in the earliest stage of plants that are non-aquatic.

The rosette (Plate VIII., Figs. 4, 8, 10) is usually stated to be an essentially Mesopotamian device, but it is scattered up and down in Egyptian and Mediterranean art. (Figs. 74, 78, 79, 84.) It may be characteristic of Assyria, but it is by no means peculiar to it.

The rosette in Egypt is probably mainly a lotus-motive; the upper end of the yellow-rayed seed-vessel may be regarded as the chief original, but some are undoubtedly fully expanded lotus flowers seen from above or below, or a group of buds or of flowers arranged radially. However conventionalised it may become, the rosette is most constantly associated with the lotus in Egypt, the land of its birth. Their association elsewhere is only to be expected, as there would naturally be a tendency for the rosette to accompany the knop and flower in their migrations.

According to Professor Flinders Petrie,[77] it is even doubtful whether the rosette was truly of vegetable origin. The use of leather-work seemed to have greatly modified the rosette. Its primitive form did not look floral at all, merely a circle with white dotted lines radiating across. Later, there were concentric rings of colours, with the same white dotted lines. The stitched leather theory explained a whole host of peculiar ornaments that could hardly otherwise be understood.

Goodyear[78] points out (p. 101) that no dated example of the rosette is known in Assyria or Chaldea before the twelfth century B.C.—i.e., on the dress of Merodach-idin-akhi, King of Babylon. It occurs with other lotuses in Egypt on the head-dress of Nefert, a statue of the Fourth Dynasty, 3998-3721 B.C. As previously stated, the earlier Egyptian kings of the Eighteenth Dynasty conquered Assyria. The reign of Thothmes III., who, according to a contemporary expression, “drew his frontiers where he pleased,” is placed by Professor Flinders Petrie[79] from 1481-1449 B.C. The Egyptian empire then comprised Abyssinia, the Soudan, Nubia, Syria, Mesopotamia, part of Arabia, Khurdistan, and Armenia.[80]

In answer to the question, How is it that the fact has been overlooked that the rosette is as familiar a feature of Egyptian ornament as the earliest dated remains of other ornaments? Goodyear (p. 102) says that the answer apparently is that the rosette is very abundantly known on carved slabs from Nineveh, while the architectural surface carvings in Egypt are almost absolutely destitute of rosette ornament, but it is very frequently represented in tomb paintings.

Those who have argued for the Assyrian origin of the rosette appear to have only compared the stone carvings of the two countries in question, but it is well known that no borrowing of architecture took place. There is evidence that portable objects were traded from Egypt to Mesopotamia, and there is no doubt that the purely decorative mural paintings of Egyptian tombs were analogous to the patterns on Egyptian textiles, and these were traded to the East. The thresholds from Assyria were undoubtedly carved in imitation of rugs; from the monuments we may suppose that the walls were often decorated with woven stuffs, the ornamentation of which was transferred to stone and glazed bricks. We may then come to the conclusion that the mural decoration of Assyria was affected by the designs of textiles and other portable articles of merchandise, the idiosyncrasy of this country making itself felt in the selection and adaptation of Egyptian originals.

In dealing with rosettes we must be very careful not to fall into the common error of imagining that things which are similar are necessarily the same. In the course of this book there are several examples of the facility with which such a mistake could arise, and sometimes has arisen. Patterns and designs must primarily be studied in situ, and the wandering “from Dan to Beersheba” is to be deprecated as a method. It is only when the indigenous material is insufficient, or fails in its results, that the comparative method should be employed, and then only when history, tradition, or other lines of evidence warrant its use.

Rosettes undoubtedly occur in Egyptian decoration as well as in Assyrian. Goodyear makes a special pleading for the derivation of the latter from the former. The question really is—Are all Assyrian rosettes lotus-motives which originally had their source in Egypt? Few will doubt that Egyptian rosettes may have travelled with other lotus derivatives to Assyria, but it is improbable that a wholly foreign ornament should stud itself so profusely and ubiquitously over Assyrian architecture and manufactures.

I do not profess to be able to suggest what may be the original, or originals, of the primitive Assyrian rosette; but it does seem as if its vitality was increased and its employment further perpetuated by the cross-fertilisation, to speak figuratively, of the immigrant Egyptian variety.

In studying the influence of the lotus in decorative art we have to travel far afield, as it has left its trace even in India. The art of modern India is, so to speak, a medley composed of foreign motives and influences associated with native designs and religion. Under the term “native” must be included all the artistic influences which have been afforded by the mixed races of that vast peninsula. A very brief and limited survey of some of the historical aspects of the question must suffice.

In very early days “the Chaldeans, whose cry is in their ships,” voyaged to India for commercial purposes. Proof of this is found in the discovery of teak wood among the ruins of Mugheir. It is agreed also that there are distinct traces of Assyrian influences in Indian art.

Sir George Birdwood[81] (ii. p. 162) says, “The researches of Mr. Fergusson have shown that stone architecture in India does not begin before the end of the third century B.C.;” and again (i. p. 99), “There is no known Hindu temple older than the sixth or fifth century of the Christian era; and all the earlier stone buildings are Buddhist.”

The same author has come to the conclusion (i. p. 146) “that the remarkable European character of the Buddhistic sculptures in the Panjab and Afghanistan is due, not to Byzantine, but to Greek influence. They are unmistakably Buddhistic sculptures, and may therefore date from B.C. 250 to about A.D. 700; and any of them which are later than the fourth century A.D. may have been executed under Byzantine influence.... Dr. Leitner was the first to insist on describing (the Buddhistic remains in the neighbourhood of Peshawar in the Panjab) as Græco-Buddhistic sculptures.... Their resemblance is probably due to their having been executed by Indian workmen from Greek designs or models.”

Goodyear remarks, “At a later date Hindu art became saturated with Mahommedan lotus patterns. These were all originally borrowed in the countries conquered by the Mohammedan Arabs, during the seventh century A.D., Syria, Egypt, North Africa, and Persia.” Islam swept into her net the decorative art of the countries she conquered, and as realism was denied to her owing to the Prophet’s injunction against depicting human or animal forms, she had to fall back on patterns, but, unknown to her, many of these were lotus derivatives. It was these patterns that the Arabs brought with them to India.[82]

“The history of India,” continues Goodyear, “thus explains why its apparently favourite water-lily [the Nelumbium] has had so little influence on its ornamental patterns. Although naturalistic rendering of the rose water-lily is found in ancient and modern Oriental art, it must be remembered that this has nothing to do with the dominance of a pattern, which is a matter of technical tradition. It appears that the famous Indian water-lily exercised no visible influence on the art of Egypt, and that Egyptian patterns have invaded its own home by many paths, at many times, borne by waves of historic influence which are admitted to have determined the character of Hindu art since the third century B.C., which is the first century in which this art is known to us.”

Examples of Indian forms of the anthemion will be found on Sindh pottery (Plate VIII., Fig. 11), on Delhi and Cashmere shawls, and on innumerable other objects and temple carvings. If one compares the anthemion combined with an “astragal” moulding in Fig. 80, which is from the Lât at Allahabad, with Figs. 7 and 5, Plate V., which are purely Greek, it will be evident that borrowing has taken place. One cannot follow Sir George Birdwood[83] when he says this “necking immediately below the capital represents with considerable purity the honeysuckle ornament of the Assyrians, which the Greeks borrowed from them with the Ionic order. Its form is derived originally from the Date Hom, but it really represents, conventionally, a flowery lotus, as the Bharhut sculptures enable us to determine. The ‘reel and bead’ pattern running along the lower border of the necking represents the lotus stalks.” This author does not state which lotus he refers to, probably it is the Nelumbium or Rose water-lily, but the stalked flowers added on each side of the central anthemion have no distinctive character, nor can I see that the figures he gives of the Bharhut sculptures are any more definite.

Fig. 80.—Anthemion and astragal moulding from the Lât (stone column) at Allahabad; from Birdwood, after Fergusson.

The Buddhist missionaries carried this pattern with them to China, where on some of the pottery unmistakable lotus derivatives occur, and those too of the anthemion series.

From the Orient we must retrace our steps westward. Persian art may be left on one side, as it was largely a legacy of Assyrian.

Among the Mediterranean peoples the Phœnicians claim first attention on account of their early assumed rôle of middle-men. But as Perrot and Chipiez remark, “In the true sense of the word we can hardly say that Phœnicia had a national art. She built much and sculptured much, so we cannot say she had no art at all; but if we attempt to define it, it eludes us. Like an unstable chemical compound it dissolves into its elements, and we recognise one as Egyptian, another as Assyrian, and yet another, in its later years, as purely Greek. The only thing that the Phœnicians can claim as their own is the recipe, so to speak, for the mixture.” Herodotus tells us that the Phœnicians had in their ship “Egyptian and Assyrian goods.”[84]

Not only did the Phœnicians barter in foreign objects, but they manufactured articles for trade, and were expert craftsmen. At the funeral games in honour of Patroklos “the son of Peleus set forth other prizes for fleetness of foot; a mixing-bowl of silver, chased; six measures it held, and in beauty it was far the best in all the earth, for artificers of Sidon wrought it cunningly, and men of the Phœnicians brought it over the misty sea.”[85] As their home-made goods were intended for foreign markets, they probably copied more or less exactly from Egyptian and Assyrian sources. They were artificers rather than original artists, their object was gain.

On the whole it appears that the Egyptian influence was more patent on Phœnician art than that of Assyrian, but on the other hand, the Phœnician religion was Semitic, and by this they were far more closely allied to Chaldea and Assyrian than to Egypt.

Through far wanderings and endless trafficking the “Phœnician, practised in deceit, a greedy knave,” as Homer dubs him, introduced numberless objects into the Mediterranean littoral which were ornamented with lotus designs or with patterns of lotus origin.

The great skill of the Chaldeans and Assyrians in weaving and embroidery enabled them to produce textiles which were highly valued wherever they found their way. The appropriation of “a goodly Babylonish garment” from the loot of Jericho by the unfortunate Achan shows how much these fabrics were prized. We know that the decoration of these beautiful and precious commodities reacted on the designs of Phœnician manufacturers, and directly or indirectly had some effect in guiding the nascent art of Europe.

When the Greeks were a young and growing people they, like most of their neighbours, were forced to trade with the Phœnicians they so despised, and were thus acquainted with trade goods from Mesopotamia and Egypt. The Ionic Greeks were more particularly influenced by Oriental art. The designs from early Greek tombs and the spoils recovered by the spade in recent excavations clearly show the nationality of the foster-mothers of Greek art.

The lessons learnt in childhood are hard to forget, and so, following the traditions of their fathers, the decorators continued to employ the same general patterns and designs that they saw around them and which they had inherited. For centuries we see the anthemion reproduced in architecture (Fig. 82 and Pl. V., Fig. 7), painting, pottery, varied it may be in detail, but essentially the same pattern. Rarely going direct to nature for inspiration, the Greeks were content with endless repetitions of slight variants of the one eternal and highly unconventional design. The mental unrest of the Greeks, which was always seeking something new, was in marked contrast to their decorative conservatism.

When the trade of Europe was taken up by Greeks they further disseminated this dominant motive. In less chaste form we find it in Roman art. The Renaissance gave it, with other matters classical, a new lease of life.

But Europe was not dependent on Greek and Roman influences alone for the spread of the anthemion. The Crusaders brought away with them many Oriental goods, and that, too, from the meeting-place of Europe, Asia, and Africa. Later the Moors invaded Spain, and left as the jetsam of their retreat a wealth of matchless decorative art, amongst which our old patterns may also be traced.

Fig. 81.—Saracenic Algerian detail; from Goodyear, after Ravoisié.

By this time it is often flamboyant. (Fig. 81.) The isolated elements of the design may have been the origin of the fleur de lis, of which the Prince of Wales’ Feathers appear to be a variant.[86]

Throughout the art of the civilised world of to-day we find repeated, again and again, the misnamed honeysuckle pattern, or the anthemion, as it is preferable to call it. Most of our modern examples can be traced to Ancient Greece, but even there it had a hoary antiquity and probably a multiple ancestry. It is not improbable that future research will demonstrate that the history of this pattern is far more complex than that which I have endeavoured to sketch out. Its amazing longevity may be due to the fact that it arose from various radicles, and when the branches met their differences were not too great to counterbalance their resemblances, and so a fusion or mingling of elements could easily and naturally result.

Fig. 82.—Ionic capital of the eastern portico of the Erechtheium.

Fig. 83.—Early form of Ionic capital from Neandreia; after Clarke.

Mr. Goodyear has an elaborate study of the evolution of the Ionic capital (Fig. 82) from the anthemion. A German architect and critic, Semper,[87] appears to have been the first to derive the Ionic capital from the volutes of the Assyrian palmette (Pl. VIII., Figs. 9, 10) by a process of gradual suppression of the leafy portion and increase of the scroll. Dr. J. T. Clarke[88] supported and elaborated this theory. At Neandreia, near Assos, in Asia Minor, he discovered an Ionic capital (Fig. 83) which is a valuable “missing link.” But, according to Mr. Goodyear, there is no need to seek an Assyrian origin for this capital when all the intermediate stages can be found in Egypt and in the Greek Islands.

Fig. 84.--Lotus design from a “geometric” vase from Cyprus; after Goodyear.

In Fig. 84 and Fig. 130, F, we have a lotus with curling sepals on pots from Cyprus; no one can dispute that these are really lotuses. The curling sepals become more spiral in Rhodian (Fig. 130, G), and especially in Melian pottery (Fig. 85). The central rosette has now become more leaf-like, but there are numerous true Egyptian examples of this, as in a compound flower (Fig. 86) from a tomb ceiling, or again (Fig. 87), on a blue-glazed lotus pendant from a necklace in the British Museum, of the Nineteenth Dynasty. In the Owens College Museum, Manchester, there is a somewhat similar enamel tomb amulet of the Twelfth Dynasty (2778-2565 B.C.). The transition from these to the stone or terra-cotta anthemion of the Parthenon (Fig. 88) is very gradual.

Fig. 85.—Lotus derivative on a vase of the seventh century B.C., from Melos; from Goodyear, after Conze.

Fig. 86.—Compound flower, based on the lotus, Thebes, Eighteenth to Twentieth Dynasties; from Goodyear, after Prisse d’Avennes.

Thus, according to this view, the volute of the Ionic capital is merely a drooping lotus sepal, which became spiral in the Grecian Archipelago. Many of the Ionic capitals, especially the earlier ones, exhibit distinct traces of the central palmette, but eventually only the spirals persisted, and the cleft between the curling sepals was gradually reduced so that their stems came to appear as a transverse band ending in volutes.

Fig. 87.—Lotus pendant from an Egyptian necklace of the Nineteenth Dynasty; after Goodyear.

Fig. 88.—Anthemion from the Parthenon.

In following this view of the history of the Ionic capital we have practically traversed that of the anthemion. The more typical examples of this pattern not only present us with the element which we have already briefly studied, but alternating with it is a trefoil. For this again there are any number of Egyptian originals in which the trefoil indicates a lotus flower; in this case all the petals have been eliminated and only the sepals persist.

Lack of time prevents me from attempting to follow the fascinating evolution of various patterns and designs which adorn Grecian temples and vases; but I must permit myself to indicate a probable origin of an exceedingly common pattern which has also overrun our own art. I refer to the so-called egg-and-dart moulding of Greek entablatures (Plate V., Fig. 5), and the same motive painted on vases or moulded on the later Samian ware (Plate V., Fig. 6). In these two figures the pattern is drawn in its usual position, but, the better to follow the argument, a typical variety is figured (Fig. 89) reversed. There are many varieties, from a series of U-shaped figures with alternating dots, as many Greek vases (Fig. 89, E), through the Samian device (Plate V., Fig. 6) and Erechtheium variety (Fig. 82 and Plate V., Fig. 5), to others in which there is greater complexity and more floral forms (Fig. 89, D).

With any given series of designs it is possible to begin at either end—in the one case there is an ascending evolution, in the other a degeneration. Students of the biological method of treating decorative art will recognise that the latter is by far the most general order in the evolution of patterns, and by adopting it in this case Professor Goodyear has been able to demonstrate the life-history of this pattern to the satisfaction of many students.

In Fig. 89, A, we have a typical lotus flower and bud pattern or Greek pattern from Rhodes; the same design occurs in a simplified form on a fragment of Greek pottery from Naukratis (Fig. 89, B),[89] in which the lotus flower is now a lotus trefoil; and in Fig. 89, C, the pattern is disrupted.

Fig. 89.—Hypothetical derivation of the “egg-and-dart” moulding from a lotus pattern; according to Goodyear.

A. Lotus anthemion on a vessel from Rhodes; after Salzmann.

B, C. Lotus anthemia on pottery from Naukratis; after Flinders Petrie.

D. Egg-and-dart moulding from the Erechtheium.

E. Degraded egg-and-dart pattern painted on a Grecian vase.

In Greek vases we usually find that decoration has been made with a fine feeling for appropriateness; thus the erect anthemion occurs when the vase is swelling, but where it is contracting an inverted anthemion is placed, because the decorative lines thus widen to correspond with the expansion of the vase. Again, in Egyptian tomb ceilings the bordering lotus pattern is inverted, as the base line of the design naturally is made to correspond with the peripheral line of the ceiling—in other words, the lotus anthemion was inverted.

We have then a painted lotus bud and trefoil pattern which was often inverted and as often a simple design. According to this view, the egg of the egg-and-dart pattern is simply a semi-oval left between two lotus trefoils, the dart being the central sepal. When this design came to be incised in stone, the new technique very slightly modified the pattern, and the flat oval areas necessarily came to be carved as rounded or leaf-shaped projections. On these latter occasionally appear reminiscences of the intervening buds, as on the Erechtheium leaf-and-dart moulding. Many variants occur in this device, especially in Roman sculpture.

Professor Goodyear points out that the egg-and-dart moulding as such is unknown to Egyptian patterns, owing to the almost entire absence in Egyptian art of carved or incised lotus borders of any kind, a preference for flat ornament in colour being the rule. Stone carved patterns of any kind in Egyptian art are quite rare before the Ptolemaic period. In Greek art the absence of patterns in projected carving is also a general rule down to the time of the Erechtheium. In Greek art also colour decoration on flat surfaces was the rule in architecture for earlier periods; for example, a leaf-and-dart pattern was painted on a Doric capital in Ægina.[90]

“The Ionic capital, the ‘honey-suckle,’ the egg-and-dart moulding, the meander, the various forms of spiral ornament, the guilloche and the rosette, and some few other motives, belong to one ornamental system, and have never been used in Europe, apart from historic connections with their original system, since the Greeks, and have never been used in Europe since prehistoric ages, without distinct dependance on the Greeks. As found with the Greeks they can all be traced back to Egyptian sources; except the guilloche, which is only the later variant of the spiral scroll. The guilloche pattern has been found in Egypt on pottery dated to the Twelfth Dynasty (2700 B.C.), which was probably made by foreigners resident in the country, but it may easily be an Egyptian pattern which has not yet been specified as such.

“The Egyptian rosette can be dated to the Fourth Dynasty, 3998-3721 B.C. Since that time its history has been continuous. Since its first transmission to Europe it has never been reinvented in Europe, for there was never an occasion or a chance to reinvent it there.

“The spiral scroll is dated to the Fifth Dynasty, and the meander (at present) to the Thirteenth Dynasty, about 2500-2000 B.C. The Egyptian Ionic capital is dated to the Eighteenth Dynasty, 1587-1327 B.C. The Egyptian anthemion (‘honey-suckle’ original) is dated to the Twelfth Dynasty (2778-2565 B.C.). A considerably higher antiquity than the given date must be assumed in all cases.”[91]

This in brief is Professor Goodyear’s theory;[92] it is ingenious, but time will show how far it will convince students of this subject. It is quite possible that the egg-and-dart pattern may have had a multiple origin. Dr. Colley March is still inclined to see in it a kind of artistic reminiscence of the ends of beams (Plate V., Fig. 1) of earlier wooden buildings; but it is highly improbable that the conclusion arrived at by Mr. Hulme is the correct one. He says: “The echinus, or horse-chestnut, is also called the egg-and-tongue or egg-and-dart moulding, a variety of names that may be taken as conclusive of the fact that it bears no great resemblance to anything at all, but is a purely arbitrary form.”[93] The variety of names is conclusive only of the ignorance of the name-givers as to what the pattern originated from. In future those who write on decorative art will have to prove that any pattern or design is a purely arbitrary form; that assumption is no longer permissible.

We have left the lotus far behind, and though it is hard to believe that the multitudinous designs of so many ages and of such diverse countries are all derived from the sacred flower of Ancient Egypt, yet it may well be that the oldest stock was a lotus derivative, and that the symbolism of that flower gave to it sufficient vitality to spread and multiply and replenish the earth.

C. Zoomorphs.

It is a matter of common observation that our children very early take delight in pictures of animals and in making delineations of them. It is further noticeable that the quality of the drawing makes no difference to children, and they are as pleased with the crudest representation of an animal as their elders are with a life-like portrait. In all this the child closely resembles the folk, whether they be the backward classes among ourselves or the less advanced peoples. All these agree in being satisfied with diagrammatic realism.

Savages, however, vary greatly in their power of representing animal forms. In Fig. 3 we have a number of outlines of animals which were etched on bamboo pipes or carved on wooden drums by the Papuan natives of the islands of Torres Straits or of the adjacent coast of New Guinea. The figures are all reduced to the same scale by photography from tracings of the original delineations, and are therefore faithful copies of the originals. A glance at the figure will show that the animals are drawn with a very fair degree of accuracy, so that in most cases it is perfectly easy to identify the genus of the animal intended. There are numerous little touches which appeal to the eye of the naturalist as indicating keen observation on the part of the artists; for example, the sharks (C, D) are always drawn with unequally lobed tails, the tail of the dugong (M) is accurately rendered; several characteristic details are, as a rule, well brought out in the drawings of the cassowaries (K). On the other hand, there are several anatomical mistakes, as for instance, giving shark-like gill-slits to a bony-fish, or even to a crocodile. The mouth is represented in a sucker-fish (F) as being on the upper side of the head, whereas it should be underneath, and the view of that fish’s tail would be impossible from that particular point of view; but these and numerous other similar examples which I could name are merely due to a desire to express several salient features, without regard to the possibility of their being all seen at once. The artists’ aim was to give a recognisable representation of animals, and in this they have as a rule succeeded perfectly; it is captious to expect more from them.

On other parts of the mainland of New Guinea one rarely meets with representations so life-like as these,[94] and nowhere else on that largest of islands are so many kinds of animals drawn. Animals are often depicted by the Australians, but usually these are very poor as works of art; they are also employed in pictography.

Although animals are so frequently drawn by the Torres Straits Islanders, they never arrange them in groups or in series. They are pictures of individuals, drawn for decorative effect, but they have no story to tell. The only exception to this rule occurs in the case of certain animals, two of which are sometimes placed symmetrically on the decorated object.

Representations of animals are not uncommon in Melanesia, but they are distinctly of rare occurrence in Polynesia. They occur in great profusion in America from north to south, but here they are predominantly religious or pictographic in significance. Animal forms are not characteristic of African art, except among the Bushmen, and there we find pictures of animals which are comparable with those of the Eskimo or the natives of Torres Straits.

As far back as the time when men hunted the reindeer and wild horse in Western Europe do we find drawings of animals. This was at the time period when the glacial cold was abating and when men lived in caves, used chipped, unpolished stone implements, and were unacquainted with pottery. In archæological nomenclature this is known as the Epoque Magdalénienne of the Cave Period in the Palæolithic Age. The figures of the mammoth, reindeer, horse (Fig. 90), etc., are usually cleverly etched on bone or ivory, and sometimes they are wonderfully life-like and accurate; the representation of human beings are as a rule very weak indeed.

Fig. 90.—Horses etched on an antler from La Madelaine; from Taylor.

“The wild horse roamed in immense herds over Europe, and formed the chief food of the palæolithic hunters. In some of the caverns in France the remains of the horse are more abundant than those of any other animal, more even than those of the wild ox. Thus at Solutré, near Macon, the bones of horses, which had formed the food of the inhabitants of this station, form a deposit nearly 10 feet in depth and more than 300 feet in length, the number of skeletons represented being estimated at from 20,000 to 40,000. This primitive horse was a diminutive animal, not much larger than an ass, standing about 13 hands high, the largest specimens not exceeding 14 hands. But the head was of disproportionate size, and the teeth were very powerful. He resembles the tarpan or wild horse of the Caspian steppes. A spirited representation of two of these wild horses is engraved on an antler found at the station of La Madelaine in the Department of the Dordogne.”[95]

It is impossible for me to do more than just touch on the subject of the relation of animals to decorative and pictorial art; the few examples I can offer will, however, demonstrate its importance.

Wherever it occurs the crocodile or the alligator, as the case may be, almost invariably finds its way into the decorative art of the district. From north to south the crocodile asserts itself in the decorative art of New Guinea; for further information the reader is referred to Dr. Uhle,[96] who has made an elaborate study of the crocodile in Malayo-Papuan art, has noted the strange metamorphoses to which it is subjected in north-west New Guinea; he also draws attention to the cult of the reptile in these parts. The belief of a relationship between the crocodile and man occurs among the Malays of Sumatra, Batta, Java, Makassar and the Bugis, Tagals, in Banka, Timor, Bouru, Aru, and the south-western islands. The Javanese have no fear of the crocodile when bathing, they believe that their “grandfathers” and “fathers” could do them no harm. The crocodile is reverenced in Borneo and killed only when the blood-revenge demands it; their teeth are used as talismans all over the island. The inhabitants of Kupang and Timor have an unconquerable fear of the killing of crocodiles and pray by dead ones. Even the Malays (Hovas) of Madagascar are afraid to kill crocodiles, since they would revenge themselves.

From Melanesia we will pass to Central America and take advantage of Mr. W. H. Holmes’s masterly study of the ancient art of the province of Chiriqui in the Isthmus of Panama.[97]

Wherever it occurs, the crocodile or the alligator, as the case may be, almost invariably finds its way into the decorative art of the district. From north to south the crocodile asserts itself in the decorative art of New Guinea; and, although associated with other animals, the alligator predominates among the zoomorphs of the Chiriqui.

Fig. 91.—Conventional alligator from the “lost colour” ware of Chiriqui; after Holmes.

In Fig. 91, we have a highly conventionalised representation of an alligator. The scutes (or scales) are represented by spotted triangles and run along the entire length of the back; a row of dashes in the mouth indicates the teeth.

In another class of ware the treatment is quite different, more clumsy, but prominence is given to a number of corresponding features; the strong curve of the back, the triangles, dots, the muzzle, and mouth. In Fig. 92 all the leading features are recognisable, but are very much simplified, and the body is without indication of scales, the head is without eyes, the jaws are without teeth, and the upward curve of the tip of the upper jaw in the last figure is greatly exaggerated, but this is a common feature in these representations.

Fig. 92.—Simplified figure of an alligator from the “alligator” ware of Chiriqui; after Holmes.

The spaces to be decorated also largely determine the lines of modification. In Fig. 93 we have an example crowding an elongated figure into a short rectangular space. The head is turned back over the body, the sunken curve of the back is enormously exaggerated, and the tail is thrown down along the side of the panel.

Fig. 93.—Alligator design, Chiriqui; after Holmes.

It often happens that the animal form, literally rendered, does not fill the panels satisfactorily. The head and tail do not correspond, and there is a lack of balance. In such cases, as Mr. Holmes points out, two heads have been preferred. The body is given a uniform double curve and the heads are turned down, as in Fig. 94. This figure “is extremely interesting on account of its complexity and the novel treatment of the various features. The two feet are placed close together near the middle of the curved body, and on either side of these are the under jaws turned back and armed with dental projections for teeth. The characteristic scale symbols occur at intervals along the back; and very curiously at one place, where there is scant room, simple dots are employed, showing the identity of these two characters. Some curious auxiliary devices, the origin of which is obscure, are used to fill in marginal spaces.” Judging from some of the figures in Fig. 100 we may regard the upper supplementary device as another alligator derivative.

Fig. 94.—Alligator delineation, greatly modified, Chiriqui; after Holmes.

Fig. 95 is an extreme form of conventionalised alligator which has become metamorphosed into an apparently meaningless design which is intended to be symmetrical.

Fig. 95.—Highly conventionalised alligator derivative, Chiriqui; after Holmes.

In Fig. 96 we have a series showing the degeneration of the alligator into a curved line and a spot. The series shown in Fig. 97 illustrate the tendency of linear bands not only to cramp the original in a vertical direction, but to force it into a serial pattern. Fig. 97, A, is a simplification of such a two-headed form as Fig. 94. One might be tempted to regard it as a doubly tailed form, but such do not appear to have been recognised by Mr. Holmes. The transition from this undoubted alligator derivative to the broad chevron of Fig. 97, E, is quite obvious, the conventional scutes, dotted triangles, together with the zigzag body alone forming the pattern, and in Fig. 97, F, the latter has disappeared. Mr. Holmes states “there is little doubt that the series continues further, ending with simple curved lines and even with straight lines unaccompanied by auxiliary devices.”

Fig. 96.—Series of derivatives of the alligator, showing stages of simplification, Chiriqui; after Holmes.

Fig. 97.—Series of alligator derivatives, showing modification through use in narrow zones, Chiriqui; after Holmes.

Mr. Holmes also points out that the Chiriqui have arrived at the scroll and fret by way of the alligator. I can here illustrate only two of these (Figs. 98, 99); in these the body of the reptile is the element of the design. In other cases Mr. Holmes finds that parts of the creature, such as head, feet, eye, or scales, assume the role of radicles, and pass through a series of modification ending in purely geometrical devices.

Fig. 98.—Scroll derived from the body-line of the alligator, Chiriqui; after Holmes.

Fig. 99.—Fret derived from the body-line of the alligator, Chiriqui; after Holmes.

The designs in Fig. 100 are painted upon low rounded prominences on vases, and hence are enclosed in circles. In Fig. 100, A, the alligator is coiled up, but still preserves some of the well-known characters of that reptile. In B, we have the double hook modification of the alligator’s body, but the triangles are placed separately against the encircling line. In the next figure the body-line is omitted, and three dotted scutes alone represent the animal. The four scutes of the next designs assume a symmetrical position, and the central crossed line may represent the alligator’s body. In the last figure of this series the cross has become the predominating feature, and the spots have migrated into it, so that the triangles have become mere interspaces.

Fig. 100.—Series of alligator derivatives, showing modification through use within a circular area, Chiriqui; after Holmes.

Finally, Fig. 101 is a zone pattern, painted on an earthen drum, the central zigzag line represents the body of the alligator, and the notched hooks its extremities; these are here arranged with perfect regularity, but sometimes only the latter occur in patterns, and then they are often somewhat irregularly disposed.

Fig. 101.—Pattern composed of alligator derivatives, from a clay drum painted in the style of the “lost colour group,” Chiriqui; after Holmes.

From his prolonged study of ancient American art, Mr. Holmes formulates the following generalisation:—“The agencies of modification inherent in the art in its practice are such that any particular animal form extensively employed in decoration is capable of changing into or giving rise to any or to all of the highly conventional decorative devices upon which our leading ornaments, such as the meander, the scroll, the fret, the chevron, and the guilloche, are based” (p. 187). The importance of the following conclusion is obvious:—“We are absolutely certain that no race, no art, no motive or element in nature or in art can claim the exclusive origination of any one of the well-known or standard conventional devices, and that any race, art, or individual motive is capable of giving rise to any and to all such devices. Nothing can be more absurd than to suppose that the signification or symbolism attaching to a given form is uniform the world over, as the ideas associated with each must vary with the channels through which they were developed” (p. 183).

The investigations of Dr. P. Ehrenreich and Professor Karl von den Steinen on the decorative art of various tribes in Central Brazil have led to results which may, without exaggeration, be termed startling. The patterns employed by these people typically belong to the class which is popularly described as geometrical. On page 176 I have selected examples of these patterns which will give a fair idea of the style of design.

Dr. Ehrenreich[98] informs us that in the Bakaïri chiefs’ huts a frieze of blackened bark tablets run along the wall which are painted in white clay with very characteristic figures and patterns of fish. All the geometric figures are in reality diagrammatic representations of concrete objects, mostly animals. “Thus a wavy line with alternating spots denotes a large, dark-spotted colossal snake, the Anaconda (Eunectes murinus); a rhomboidal mark signifies a lagoon-fish, whereas a triangle does not by any means indicate that simple geometrical figure, but the small, three-cornered article of women’s clothing” (p. 98).

The following quotation is also translated from Dr. Ehrenreich[99]:—“The ornaments of the Karaya consist of patterns of zigzag lines, crosses, dots, lozenges, and peculiar interrupted meanders, whereas the quadrate and triangle occur only incidentally (that is, owing to the filling up of other figures), and circles are entirely absent. As in the ornamentation of the Xingus tribes, so also here occur those apparently entirely arbitrary geometrical combinations fundamentally of wholly defined concrete presentments, of which the most characteristic traits are therein reproduced. Unfortunately it is not always possible to correctly ascertain the respective natural objects. The frequently occurring cross (Fig. 102, A), which in America has so often given occasion for amusing hypotheses, is here nothing but a kind of lizard.... Also peculiarly characteristic are the extensive wings of a bat (Fig. 102, B), as well as the frequently occurring snake pattern, such as Fig. 102, C, which represents the rattlesnake, while another snake is represented in Fig. 102, D. Accurate representations of men and animals, as we know them to be done so excellently by the Bushmen and Eskimo, do not appear to be forthcoming among the Karaya.”

Fig. 102.—Patterns of the Karaya, Central Brazil; after Ehrenreich. A. Lizards; B. Flying bats; C. A rattlesnake; D. A snake. A. Incised on a grave-post; B, C, D. Plaited on the handles of combs.

Professor von den Steinen[100] describes the above-mentioned frieze more fully. The pieces of bark, which were from 15 cm. to 40 cm. (6 to 16 inches) broad, were blackened with soot, and the white or yellowish lime applied with the fingers. The frieze itself was over 56 m. (over 184 feet) in length.

I would ask the reader to refer back to Fig. 52, p. 97, although this motive is not a zoomorph, in order to show that triangular designs, or resulting zigzags, may have various origins.

Only one tablet represented a plant. (Fig. 69.) It indicates the leaves of a small “cabbage”-bearing wild palm.

The bulk of the motives for the decorative art of these people, the Schingú tribes (the Xingu tribes of Ehrenreich), are drawn from the animal world; Fig. 103 A, H, I, K, are Bakaïri patterns, and Figs. 103 B-F those of the Auetö.

The pattern to the right in Fig. 103, H, indicates a kind of ray, the characteristic rings and dots which ornament the skin of this fish are here represented.

Fig. 103.—Patterns from Central Brazil, after Von den Steinen. A. Bakaïri paddle; B-E. Mereschu (fish) patterns of the Auetö; F. Locust design, Bakaïri; G. Fish-shaped bull-roarer, Nahuquá; H. Sukuri (snake) and ray patterns; I. Jiboya (snake); K. Agau (snake); H-I. Bakaïri tribe.

Common to all the tribes of the Schingú stock is the employment of conventionalised representations of the mereschu. This is a small compressed lagoon-fish, about 19 cm. (7½ inches) long, and 9.5 cm. (3¾ inches) deep; its colour is silver-grey with brown spots. The mereschu belongs to the genus Serrasalmo or Myletes; the figure on p. 260, given by Von den Steinen, looks as if it were drawn from a badly-preserved spirit specimen, and one fails to see how Fig. 103, C, for example, could by any stretch of the imagination be considered to suggest that fish. On p. 613 of Dr. Günther’s Introduction to the Study of Fishes (Edinburgh, 1880) is an outline figure of Serrasalmo scapularis; the contour of this fish is approximately rhomboidal, the head, the dorsal fin, and the tail fin occupy three of its angles, and the anal fin practically runs up to the fourth angle. Von den Steinen points out that in most cases representations of these animal-forms are incisions, not paintings, and the diagrammatic rendering of curved lines by angles is due to this fact. The patterns which I am about to describe are common to numerous allied tribes, and everywhere these patterns bear the name by which this kind of fish is locally known.

Sometimes the mereschu fish is employed singly, but most frequently a number of them are evenly distributed over the decorated surface, and between the fishes single, double, or even several lines may be drawn, as in Fig. 103, B, C, E; these latter represent the net by means of which these fish are caught. Thus we may have a fish-pattern or a fishes-in-net pattern. These patterns are delineated on masks, posts, spinning-whorls, and other objects. Fig. 103, B, is a pattern of the mereschu fishes-in-net group, but the fishes themselves are entirely filled up with black, and not their angles only.

The Auetö pattern drawn in Fig. 103, F, is intended for a mailed- or armadillo-fish.

On a Bakaïri paddle (Fig. 103, A) are incised four circles, which are the ring-markings of a ray, pinukái, on the other side of a transverse line follow two mereschu in the meshes of a net, then a pakú, and finally several kuómi fish. Professor von den Steinen believes that the object of this decoration is simply to bring fish close to the paddle. “But it is extremely instructive to see,” he continues,[101] “that concerning these scribblings, though they certainly do not denote anything in their order of arrangement, consequently are not picture-writing; however, every single one is by no means a casual flourish, but the diagram of a well-defined object, and consequently, in fact, represents the element of a picture-writing.”

Zigzags and waved lines are snakes. Fig. 103, K, represents common land-snake, the agau, or cobra of Brazil; to the left is the tail, the head is simply rendered, and as the skin of the snake is marked the artist characterised it by adding spots. Very similar is the sukuri water-snake or anaconda (Boa scytale), drawn to the left of Fig. 103, H. A boa-constrictor is indicated in Fig. 103, I; the row of diamonds left on the dark background, between the two rows of triangles, represents the marking of the snake’s skin. The larger terminal diamond to the left is probably the boa’s head. A snake is also painted on a Nahuquá bull-roarer (Fig. 103, G).

Fig. 104.—Patterns derived from bats; after Von den Steinen. A. Bakaïri; B, C. Auetö.

We have seen that rows of horizontal triangles are uluris, women’s triangles, but when they are margined above by a line, as in Fig. 104, B, they are bats; but rows of triangles vertically disposed, as in Fig. 104, C, are hanging bats; Fig. 104, A, is also a bat device.

Another triangular ornament (Fig. 105) represents small birds, called by the Bakaïri natives yaritamáze, that is, they are a particular kind of bird, not birds in general.

Finally, one would naturally consider that the ornament engraved on the post, Fig. 103, D, is simply the favourite mereschu pattern; but Von den Steinen assures us that the central design is not composed of mereschu, in which the angles are only slightly filled up, but that it is a locust, the lines arising from the angles of the lozenge being the legs. This locust pattern is, however, associated with true mereschus, which may be seen between the legs of the locust.

Fig. 105.—Bird design, Bakaïri, Central Brazil; after Von den Steinen.

In Europe and in our own country we can study analogous transformations.

More or less recognisable animals break out, as it were into scrolls and floral devices, as on Samian vases (Plate VI., Fig. 1), on Gaulish swords (Fig. 2), on Pompeian walls (Fig. 3), and on the gold ornaments of Tuscany (Fig. 5). In Fig. 4, Plate VI., we have on an ancient pot from New Mexico a decorative treatment of birds which recalls that of the mural paintings of Pompeii.

Often in Greece and Italy symmetrical scrolls are associated with a head. (Plate VI., Fig. 6.) The scrolls themselves may, in some cases, be an animal form which has ended in a flourish, as is taking place in Plate VI., Fig. 5; or in others they may be the remnants of plant motive.

Dr. Colley March calls attention to old bench-ends of English churches, notably those in Cornwall, which are frequently surmounted by a crouching quadruped; at a later period this appears to be converted into a single scroll like that which adorned the old pews in Ormskirk Church. (Plate VI., Fig. 7.)

An ancient silver plate (Plate VI., Fig. 8), found in a tumulus at Largo, Fifeshire, is decorated with the distorted fore-half of an animal. The transformation is advanced to flamboyant curves in the zoomorph of the Dunnichen Stone (Plate VI., Fig. 9); but the head and ear and legs can still be distinguished. It is not quite certain what animal this is intended to represent. Earl Southesk[102] believes it to be the horse, which was sacred to Frey, and is a special symbol of the sun. The second figure is very remarkable, but it seems to be an extreme and foliated form of the same zoomorph.

There are numerous examples of linear series of animals in the early art of Egypt, Assyria, Greece, and other artistic centres, but these do not appear to have developed into patterns, possibly because the units were readily recognisable, on the other hand, serially repeated conventionalised zoomorphs frequently metamorphose into patterns. These patterns by repeated copying tend to become simplified till finally not only is all trace of the original long lost, but the resultant pattern may so resemble other simple patterns as to be indistinguishable from them. This may easily lead to confusion and cause the designs to be classed as one. We thus come to the conclusion that before any pattern can be termed the same as another, its life-history must be studied, otherwise analogy may be confused with homology, and false relationships erected. Things which are similar are not necessarily the same.

At the extreme south-east end of New Guinea and in the adjacent archipelago the most frequent designs are beautiful scroll patterns, which are subject to many variations. I have already[103] described many of these, and so there is no need to again repeat what I have said, except to remind the reader that all these patterns are variations of serially repeated conventionalised heads of the frigate-bird. I shall again allude to this bird when I deal with the relation of religion to art.

In the same district one occasionally meets with a pattern (Fig. 106) which in some respects resembles the former and appears in some cases to have been confounded with it. This one clearly arises from the serial repetition of conventionalised heads of crocodiles. The illustration is part of the carved rim of a wooden bowl in my possession, which probably came from the Trobriands or the Woodlarks. The triangles above the crocodiles’ snouts are coloured black, those bounded by their jaws are painted red.

There is yet another method of representing animals which consists in grouping them so as to tell a story, or, in other words, to make a picture.

Fig. 106.—Rubbing of part of the carved rim of a wooden bowl in the author’s collection. Probably from the Woodlarks or Trobriands, British New Guinea. One third natural size.

Grouped animals rarely occur by themselves in decorative art; men, houses, implements, and even vegetation are frequently associated with them. The Arctic peoples, such as the Lapps, Eskimo, etc., greatly affect this form of art. The bulk of these pictures are representations of hunting scenes, and many incidents in the lives of these hyperboreans are depicted on bone and ivory. There is reason for regarding these as records of particular events (cf. p. 207); but they are also very useful to us as illustrations of native life and industry. Animals are sometimes drawn foreshortened, and confused herds of reindeer are often figured; but the grouping is mainly linear, without effects of perspective being attempted.

This kind of art is extremely rare amongst savage peoples, in fact its presence may be regarded as one of the proofs that the people practising it have passed from a purely savage condition, and have made some advance towards civilisation. It has reached its highest point in the works of the great animal painters of the present day, and thus has been one of the last forms of graphic art to be perfected.

As a general rule the inferior representations of animals in groups, and of animal pictures generally, are not due to the process of decay. They are the bad workmanship of inferior craftsmen. It is the imperfection of immaturity, not the symptom of decadence.

The last stage of the life-cycle of this class of zoomorphs occurs when incompetent draughtsmen copy the work of a master; when, for example, we see on the walls of country inns cheap and badly-drawn copies of Landseer’s pictures.

Animals also play a large part in mythology, and it is often very difficult to determine the limits of totemism in this direction. There are, however, numberless instances of legendary communications and relationships, of friendliness and enmity between animals and men, which have no connection with totemism, and these often form the subject of decorative art. Sometimes the animal alone is represented, at other times both man and animal are depicted, and according to their artistic treatment we may have pictures, or should the zoomorph and anthropomorph be rendered schematically, heteromorphism may result. At present we have to deal with representations of animals which illustrate some belief, myth, or folk-tale. The sacred art of the Hebrews was almost free from zoomorphs, and that of Islam totally so; with these exceptions there has scarcely been a religion in which zoomorphs have not played a greater or less part.

I need only remind the reader of the numerous examples in which animals are depicted in illustration of, or as a kind of mnemonic of a folk-tale, a legend, or myth, and of some sacred tradition or belief. There are so many intermediate stages between these different phases that it is often impossible to draw the line between them. The religious belief, with its sacred tradition of one age, becomes the myth or the legend of a later period, subsequently it is perpetuated as a folk-tale; later it may serve to amuse children, and lastly it becomes the object of scientific study.

What I have termed the æsthetic life-history may occur to the zoomorph at any or all of these stages of religious decadence. There is no correlation between an extreme or medium phase in the æsthetic cycle and a corresponding stage in the religious series. To take a homely example, the illustrations of the most recently published fairy-tales are as a whole of greater artistic merit than has been the average illustration of sacred narratives during any period of the world’s history.

D. Anthropomorphs.

As a general rule, savages are less skilful in the delineation of the human form than they are with representations of animals, nor is it usually employed so frequently as might be expected.

It is for religious purposes that the human form is most frequently represented, and I refer the reader to the section in which religion is dealt with for illustrations of this fact. I employ the term “human form” advisedly, as this includes the images of both gods and men. At one stage of its evolution in the human mind, deity, like the Spectre of Brocken, is the shadowy image of man projected on the clouds. So the gods are most naturally represented as men, but often with special attributes. Now, these attributes are worthy of special study as being the milestones which indicate the distance which any given religious conception has traversed.

In the distant vista of time we can dimly perceive the transformation of the totem animal into the god. In the highest period of Greek sculpture the evolution was, for example, perfected in “ox-eyed lady Hera,” consort of Olympian Zeus, and in the Cnidian statue of Demeter, “Mother-Earth,” whose archaic representation was a wooden image of a woman with a mare’s head and mane. For thousands of years the Egyptian pantheon was peopled by gods arrested in the process—gorgonised tadpoles of divinity. Still earlier stages may even now be noted among savage peoples.

I know of no example of the preponderating employment of the human face for decorative purposes to be compared with what I have established for the natives of the Papuan Gulf. Illustrations of this will be found in Figs. 10-19, and in my Memoir on Papuan Art, but only an examination of a large number of objects from this district of British New Guinea will bring home to the student the remarkable ubiquity of the motive. We have no information concerning the reason for copying human faces; my impression is that it is related to the initiation ceremonies, which we know from the accounts of the Rev. James Chalmers to be very prolonged and important. One would expect to find more animal representations among these people than appear on objects in our ethnographical collections. Possibly these people are passing from the totemistic into the anthropomorphic phase of religion, and the latter finds most expression in their art. However, such speculations are futile until we obtain far more detailed and extended information of their religion than we at present possess.

Human beings are comparatively rarely represented merely for decorative purposes. In pictographs they have no predominating position. But when we come to portraiture the matter is very different; here we have an adequate motive for the delineation of the human form and face; it is, however, very noteworthy that portraiture, as such, only occurs amongst civilised communities. Possibly the explanation of this may be found in the widespread savage philosophy of sympathetic magic. According to this system a portrait has a very vital connection with the subject, and any damage done to the counterfeit would be experienced by the original. Portraiture then would be too hazardous to health, or even life, to be lightly undertaken.

What we have seen happening to plants and animals is also the fate of men in decorative art. A few examples here will suffice.

New Zealand is one of the places where anthropomorphs abound, due in this case to ancestor cult. The short series of three clubs (Plate VI., Figs. 10-12) illustrates the metamorphosis of the limbs into curvilinear forms. In dealing with the religion of Polynesia I give examples (Figs. 124-128) of the degradation of the human form into “geometrical” patterns.

In the various illustrations which have been given representations of the human form may be isolated, as in Melanesia (Fig. 3, O), Mangaia (Fig. 124), and New Zealand (Plate VI., Figs. 10, 12), or they may be double; for example, one frequently finds in Polynesia two god-figures placed back to back, and these may strangely degenerate, as in the examples given by Stolpe[104] and Read.[105] Human forms placed in linear series are frequent in Mangaian wood-carving (Figs. 127 and 125, A). Fig. 126 illustrates the decoration of a broader area.

We get examples of the selection of one portion of the man in the face patterns of the Papuan Gulf. (Figs. 10-19.)

These are undoubtedly conscious selections from the very commencement, but we find various parts of the body come to be perpetuated, with the elimination of the remainder, owing to differing causes.

The reason for the simplification of the body and the disappearance of the head in the Mangaian art is probably partly due to the fact that savage peoples are usually quite content with suggestions of objects, they do not demand what we term realism. By conventionalising their representations the Mangaians were better able to multiply them, and at the same time to appropriately decorate the object with which they were concerned. It could not be with a view of economising time or labour. “Time,” as Stolpe says, “is for them of no importance, they have plenty of it, and usually they are not able even to reckon it.” Judging by the skill exhibited by these clever carvers in wood, we cannot put down the simplification of the human body to careless copying.

We have seen that the face may be represented to the exclusion of any other part of the body, but there are examples of parts of the face becoming predominant.

Professor Moseley[106] was, I believe, the first to indicate the evolution which occurred in the images of gods in the Hawaian group. In some instances the hollow crescent form, which came to represent a face, seems to have been arrived at by an enormous increase in the size of the mouth; in others, as in the case of some wicker images, by a hollowing out of the face altogether; the mouth in the latter, though large, not being widened so as to encroach upon the whole area of the face. Since, in the worship of the gods, food was placed in the mouths, the mouths may have been gradually enlarged as the development of the religion proceeded, in order to contain larger and larger offerings, and the head in the wicker-work image may have been hollowed out for a similar purpose. Moseley traced the degeneration of the human (or god’s) face down to a hook-shaped ornament cut out of a sperm whale’s tooth.

Some of the carvings of the human face from New Zealand bear a general resemblance to those from Hawaii; but a very noticeable feature in the art of the former island is the protruding tongue. The most interesting development of this member occurs in the Maori hani, or staff of office. At the upper end is what appears, at first sight, to be a spear-point. “This portion, however, does not serve the purpose of offence, but is simply a conventional representation of the human tongue, which, when thrust forth to its utmost conveys, according to Maori ideas, the most bitter insult and defiance. When the chief wishes to make war against any tribe, he calls his own people together, makes a fiery oration, and repeatedly thrusts his hani in the direction of the enemy, each such thrust being accepted as a putting forth of the tongue in defiance. In order to show that the point of the hani is really intended to represent the human tongue, the remainder of it is carved into a grotesque and far-fetched resemblance of the human face, the chief features of which are two enormous circular eyes made of haliotis shell.”[107]

My friend, S. Tsuboi, has made a special study[108] of the protruding tongue in New Zealand art. He gives illustrations of thirty-one specimens, and with characteristic Japanese ingenuity he has drawn figures of half-a-dozen models which he has constructed which illustrate the various possible variations, and the lines they may have taken. He has also made numerical tables of possible varieties. I allude to, this paper in order to draw the attention of students to graphic methods. I regret that my ignorance of the Japanese language precludes my giving the results of this investigation.

In Ancient Egypt the eye was symbolic, and numberless amulets are found which exhibit one, two, or numerous eyes in varying stages of degeneracy, or in strange modifications. These, too, have been studied and described by Tsuboi.[109]

E. Biomorphic Pottery.

In the description of the primitive methods of pottery manufacture, allusion was made to the fact that vegetable and animal forms were copied by the early artificers.

Although the immediate originals of many kinds of clay vessels were baskets of various kinds, we must not forget that these also were often textile imitations of natural objects. Gourds which are of almost ubiquitous occurrence undoubtedly were early and independently utilised as vessels. For the more convenient porterage of them they would be enclosed in netting or basketry. The better the accessories became, the less need for the original foundations, especially as the latter were brittle. From the fact that the shape of certain baskets in a district resemble those of the gourds of that district, we may assume that this process of evolution has operated spontaneously in diverse places. Clay vessels which were modelled from the suggestion of such baskets would thus remotely be phyllomorphs but having an intermediate skeuomorphic stage.

Instead of this indirect mode of origin a more direct one has often occurred. Messrs. Squier and Davis[110] record: “In some of the southern states (of North America), it is said, the kilns, in which the ancient pottery was baked, are now occasionally to be met with. Some are represented still to contain the ware, partially burned, and retaining the rinds of the gourds, etc., over which they were modelled, and which had not been entirely removed by the fire.” They also state that the Indians along the Gulf moulded their vessels “over gourds and other models and baked them in ovens.”

It is not necessary to believe that this has everywhere been the original ceramic gourd-derivatives, even among savage peoples. Once the power of working in clay was acquired, intentional copying of gourds (Figs. 107, 108), or other vegetable vessels, may very well have occurred. This is rendered all the more probable from the fact that animal forms are modelled as earthen vessels. I am not here alluding to figures of men or of totem, sacred, or familiar animals which may belong to a somewhat higher stage of culture than that which we are now more particularly considering; but to clay utensils which are copied from receptacles which are the shells or other parts of animals.

Fig. 107.—Gourd; after Holmes.

Fig. 108.—Clay vessel, made in imitation of a gourd, from a mound in South-eastern Missouri; after Holmes.

Wherever shells of sufficient size are found they are utilised as food and water vessels, and there are numerous instances in various parts of the world of vessels being modelled so as to represent the ancient and familiar utensils.

Fig. 109.—Clay vessels imitated from shells, from the mounds and graves of the Mississippi Valley; after Holmes.

Clay vessels imitating both marine and fresh-water shells are occasionally obtained from the mounds and graves of the Mississippi Valley. The conch-shell appears to have been a favourite model (Fig. 109, A and B). A clam shell is imitated in C and D. The more conventional forms of these vessels are exceedingly interesting, as they point out the tendencies and possibilities of modification. The bowl (E) has four rosettes, each consisting of a large central boss with four or five smaller ones surrounding it. The central boss, as in A, is derived from the spire of the conch shell, and the encircling knobs from the nodulated rim of the outer whorl of the shell. Mr. Holmes suggests that in this case the conception is that of four conch shells united in one vessel, the spouts being turned inwards and the spires outwards. With all possible respect to Mr. Holmes, I venture to demur to this interpretation. The fusion of elements which are essentially isolated is rare amongst primitive peoples; it is difficult to imagine how they could conceive of the structural union and fusion of four conch shells. This is very different from the amalgamation of the clay imitations of such vessels as gourds or coco-nuts, for these are frequently fastened in pairs or in small groups to a common string handle, and there is already the idea of multiplicity and the apposition of the vessels. Again, Mr. Holmes does not present us with any intermediate stages of this or similar clay vessels; until such evidence is forthcoming it would be safer to regard this as an example of transference. According to my interpretation, the rosette derived from the spire of a conch shell was a pleasing motive, and it was applied to and repeated upon a circular bowl, which may, as Mr. Holmes elsewhere[111] suggests, be derived from the lower half of a gourd. A single conch derivative would be entitled to one rosette only, and the association of ideas would operate in favour of only one being moulded, at all events until a very extreme stage of degeneration had been attained; but in the case of transference there would be no continuity of custom to control the potter, and consequently more scope could be given to his fancy.

A highly conventionalised form is shown in F (Fig. 109). The cup is unsymmetrical in outline, and has a few imperfect bosses near one corner, but its resemblance to a shell would hardly be recognised by one unacquainted with more realistic renderings of similar subjects. In G we have an imitation of a shell cup placed within a plain cup.

The skins, bladders, and stomachs of animals are very frequently employed as water-carriers. The characteristic forms of these may often be traced in the pottery of the same districts, odd details of form or of surface marking usually persist to a surprising degree.

In Fiji and elsewhere the image of a turtle has been modelled in clay, doubtless because the carapace is often used as a vessel.

While the use of an animal or the part of an animal as a vessel has often led to the imitation of that animal in clay or other material, owing to an association of ideas, we must be very careful not to run to the extreme and to say that there was a primitively utilitarian origin for all zoomorphic vessels. Sympathetic magic and religion are responsible for many, and we must admit that mere fancy must sometimes come into play, and when this is the case theorising is necessarily at fault.

3. Heteromorphs.

As previously stated, I propose to adopt the term Heteromorph for a confusion with one another of two or more different skeuomorphs, or with the amalgamation of any two or more biomorphs, or with the combination of any skeuomorph with any biomorph. We may thus have (1) Heteromorphs of skeuomorphs, (2) Heteromorphs of biomorphs, and (3) Heteromorphs of skeuo-biomorphs.

To speak somewhat figuratively, heteromorphism is a sort of disease that may attack the skeuomorph or the biomorph. Whereas the final term of the life-history of the biomorph is, so to speak, senile decay, the result of heteromorphism is a teratological transformation. Accepting this view of the subject, the present section might be entitled “The Pathology of Decorative Art.”

Any stage of the life-history of a biomorph, whether it is the expression of decorative or religious art, is liable to be infected by heteromorphism. The only section of graphic art which must from the nature of the case be free from it is pictorial art. Where heteromorphs are introduced into pictures they form one of the subjects of those pictures, the picture itself is not subject to this modifying influence; for example, the introduction of the representation of a sphinx or a gryphon into a picture does not constitute the latter a heteromorph.

A. Heteromorphs of Skeuomorphs.

The combination of two different kinds of skeuomorphs does not appear to be of very frequent occurrence, or, at all events, we have not yet trained ourselves to appreciate them.

In Fig. 50 we have an example, which, however, is not particularly satisfactory. It will be noticed that various kinds of plaiting are indicated on this Tongan club; as a matter of fact, if it had really been covered with plaited work, the latter would have been uniform in its character, although diverse patterns might have been worked into it. If this club had been decorated in a consistent manner the simple in-and-out plaiting of the broad band, as in the middle of the figure to the left, could not occur along with the finer oblique plaiting in other parts of the object.

B. Heteromorphs of Biomorphs.

Wherever two or more animals or plants are represented in association there is a tendency for them to amalgamate in process of time. I have shown numerous examples of this in the bird and crocodile motive in Papuan art, and it would be easy to multiply illustrations.

Heteromorphism is especially characteristic of that style of decoration which we call arabesque, or grotesque. This is said to have been the invention of a painter named Ludius in the reign of the Emperor Augustus. That sovereign is said by Pliny to have been the first who thought of covering whole walls with pictures and landscapes. The fashion for the grotesque spread rapidly, for all the buildings of about that date which have been found in good preservation afford numerous and beautiful examples of it. Vitruvius was entirely out of conceit with this sort of ornament, and declares that such fanciful paintings as are not founded in truth cannot be beautiful; but the general voice, both in ancient and modern times, has pronounced a very different opinion. It was from the paintings found in the baths of Rome that Raphael derived the idea of those famous frescoes in the gallery of the Vatican. His example was immediately followed by other distinguished artists. This style derived its name grotesque from the subterranean rooms (grotte) in which the originals were usually found—rooms not built below the surface of the ground, but buried by the gradual accumulation of soil and ruined buildings.

A typical example of Pompeian treatment is seen in Plate VI., Fig. 3, where a bird’s tail passes into a floral scroll.

The representations of such mythical monsters of antiquity as the Sphinx, Chimæra, the Harpies, and so forth, are familiar to all. Originally these embodied distinct conceptions which were familiar to the initiated, if not to all. They were symbols and their origin in art was religious; their retention was due to their decorative quality.

C. Complex Heteromorphs.

We have now to consider the complications arising from a combination of skeuomorphs and biomorphs.

Again I have recourse to Dr. Colley March’s suggestive essay. He points out that in the north of Europe animals were strangled by the withy-band, as occurs on an incised stone from Gosforth (Plate VII., Fig. 3). Mr. Hildebrand endeavours to show that the so-called Scandinavian sun-snake was produced by the breaking down into curves of the figure of a lion rampant, copied by a succession of artificers, all ignorant of the appearance of a lion. But in the first place, points out Dr. March, the Norse Wurm is found long ago in prehistoric rock-sculptures. In the next place, the serpent of the north was symbolic of the sea and not of the sun. And then, it was not the unfamiliar lion that alone broke up into serpentine forms; the skeuomorph assailed the stag, as on King Gorm’s stone in Denmark (Plate VII., Fig. 2). Eikthysnir, the stag of the sun, who was an attendant and attribute of Frey, is here seen being strangled by the “laidly worm” of Scandinavia. Dr. March suggests that perhaps we may recognise the walrus in rock-sculptures at Crichie in Scotland (Plate VII., Figs. 6, 7). That the walrus was well known to the Northmen, and highly prized both for its hide, from which ships’ ropes were made (Plate IV., Fig. 4), and for its tusks, which were a source of ivory, is proved by the Orosian story (I. Orosius, i. 14). “He went thither chiefly for walruses, because they have noble bone in their teeth, and their skin is very good for ships’ ropes.” The Earl of Southesk,[112] however, brings forward a considerable body of evidence in favour of the view that this “elephant” symbol, as it has been absurdly termed, is the sun-boar—a symbol of Frey. No animal held a higher place in Scandinavia, and at an early period it was adopted as the national emblem in Denmark, and borne on the standard.