The signs themselves are chosen from a conventional field. Limited as is the number of stones that we have to draw from, it will be found that certain symbols are continually recurring as certain letters or syllables or words would recur in any form of writing. Thus the human eye appears four times and on as many different stones, the 'broad arrow' seven times, and another uncertain instrument (No. 16 of the list given in the succeeding section) as much as eleven times. The choice of symbols is evidently restricted by some practical consideration, and while some objects are of frequent occurrence, others equally obvious are conspicuous by their absence. But an engraver filling the space on the seals for merely decorative purposes would not thus have been trammelled in his selection.

Two other characteristics of hieroglyphic script are also to be noted. The first is the frequent use of abbreviated symbols, such as the head for the whole animal, the flower or spray for the plant. The second is the appearance of gesture-language in graphic form—an invaluable resource of early pictography for the expression of ideas and emotions. Amongst such may be noted the human figure with arms held down (Fig. 36b), the crossed arms with open palms and thumbs turned back (Fig. 31b), and, closely allied to this, the bent single arm with open palm (Fig. 35d). Such features, again, as the wolf's head with protruding tongue—also found on Hittite ornaments—or the dove pluming its wing, have probably a significance beyond the mere indication of the animal or bird.

The symbols occur almost exclusively in groups of from two to seven; the most frequent however are of two or of three, which seems to show that the characters thus appearing had a syllabic value. Certain fixed prin

p.32

ciples, also, are traceable in the arrangements of the symbols in the several groups. Some signs are almost exclusively found at the beginning or the end of a line. The human eye appears thus three times out of four; the instrument No, 16 of the list below occupies the extremity of the group in seven, or perhaps eight, cases where it occurs. The same two symbols moreover are seen on different stones in the same collocation. Thus the horns and four-raved star occur in close proximity on the stones (Fig. 23b and 325) from Crete. The instrument (No. 16) above referred to occurs five times on as many different stones in collocation with the 'broad arrow.' The arrowhead, again, is twice placed beside the -like sign No. 54 (Figs. 235 and 35a. In four cases where the bent leg makes its appearance (Figs. 225, 2.5a. 345 and Supplement, p. 136), it is in immediate contiguity with a symbol that seems to stand for a door or gate. Such collocations in the small number of instances at our disposal are alone sufficient to exclude the supposition that the signs on these stones were engraved haphazard for decorative purposes.

It further appears, when we come to file the several columns, as on the Babylonian principle they would follow one another in the impression of a seal, that in several cases a boustrophedon arrangement has been adopted which recalls that of early Greek writing. This is specially noticeable iii Figs. 22, 23, 33, as well as in Fig. 34, where by the analogy of other Mycenaean gems from Crete representing ships the vessel must be taken as going in the direction in which the oars slope. It seems usual to begin from right to left.

That these seals were designed to convey information regarding their owners in a primitive form of writing is clearly brought out by another phenomenon with which we have to deal. On Fig. 36d the place of the pictographic symbols is taken by linear characters which no one will deny represent actual letters, and which fit on in fact to an Aegean or Mycenaean syllabary the existence of which can be demonstrated from independent sources. This phenomenon must certainly be taken to throw a retrospective light on the hieroglyphic forms that replace the letters on the bulk of these stones. It will be further shown in the course of this inquiry that a certain proportion of these pictographic signs reduced to linear forms actually live on in this Aegean syllabary.

In a succeeding section ^19a attention will be called to a still earlier class of Cretan seal-stones presenting for the most part the same typical triangular form as those of Class I. already described. These more primitive stones, which cannot in fact be separated by any definite line of demarcation from the later series, throw a valuable light on the original elements out of which the more formalized pictographic system finally grew. In some cases the same symbols are actually seen in a more primitive stage of development. But on this earlier class the more purely pictorial and ideographic elements

^19a See p. 824 seqq. The stones, Figs. 21, 87, 89, 40, might perhape with greater propriety have been grouped with this earlier series.

p.33

are naturally preponderant arid the personal relation in which the seals stand to their owners is clearly revealed. They seem indeed to be descriptive of his individual character as an owner of flocks and herds, a merchant, a huntsman or a warrior.

These more naive delineations, of a ruder stage of culture, supply a welcome clue to the interpretation of such ideographic elements as survive in the more conventional forms with which we are at present dealing. Here too we may often see a reference to the avocation or profession of the owner of the seal and may venture to conclude that the more purely symbolic characters have a personal application. Thus for example Fig. 34, exhibiting at the beginning of one column a ship with two crescent moons above it, may be reasonably supposed to have been the signet of one who undertook long voyages. Fig. 24, with the pig and door, would have belonged to some one who owned herds of swine: in which ease the two figures of the axe and kid on the other face may contain the elements of the owner's own name. The fish at the head of Fig. 33 may indicate a fisherman. The seal-stone represented in Fig. 23, with the adze and other implements—including one in which I have ventured to recognize the template of a decorative artist,—probably belonged to a member of a masons' guild. The harp on Fig. 31 suggests a musician. It is possible that the individual element of ownership, which on the earlier class is brought out by the complete human figure, may be elsewhere indicated by the human eye alone, which is of frequent occurrence in these stones.