three fingers and a thumb, while the horizontal line is the palm connecting them. If this be so, the earlier character must be pronounced less like the object than the later, since it consists of four strokes only, which, according to the explanation propounded, would be the palm, the thumb, and two fingers. But it may be questioned whether the original "picture" was not a hand and wrist in profile,* in which case the four strokes, representing the wrist, the palm, the thumb, and the index-finger, would be more correct than the five.

Kaph is properly "the hollow hand," vola of the Romans. In the characters, both early and late, it is placed at the end of the fore-arm. Now here the "hollow" is certainly better represented by the open head of the earlier than the closed one of the later character.

Lamed is a "prick-stick," or "ox-goad," which is well represented by the early character, where the long upper line terminates in a point, while the lower end represents a curved handle. In the later forms of the letter the point was lost, the handle became angular, and an addition to the handle was made, which had nothing corresponding to it in nature. This in the later Hebrew form,, became the main part of the letter.

Of mem, "water," represented originally by a wavy line, like that whereby the Assyrians represent water on their sculptures, the first and second forms (see the plate) retain the original idea, which is almost wholly lost in the third and fourth.

The original form of samech is a good representation of a "prop supporting a trellis for vines. The later, which are tachygraphical abbreviations, furnish far less accurate pictures. Ultimately the idea became completely lost in the square-headed Hebrew .

If tsade means, as is probable, "a fish-spear," and if that had in early times the common later form of a trident, then we must pronounce the ancient tsade a better representation of the object than the later, though even in the most ancient form the desire to write rapidly has caused a considerable departure from the original figure.

Scholars are not agreed as to the meaning of the word koph. It has been translated, "the back of the head," "an ear," "an axe," "a pole," and "the eye of a needle." If the last is, as we believe it to be, the true meaning, then the earliest form must be pronounced the best, indeed the only good, representation. In this the circle is small in proportion to the length of the stroke below it, as the eye of even a rude needle would be, and the eye is not traversed by any part of the shaft. In the second form, which resembles a Greek (see the plate), this departure from truthfulness takes place. In the third and fourth everything is sacrificed to the desire of forming the letter by a single stroke.

* So Bunsen, following Rödiger, "Philosophy of History," vol. iii. p. 261.

Shin, we are told, means a tooth." The original form was probably the picture of a molar with two long fangs. Of this the early shin, which resembles a W, is a fair tachygraphic imitation. The later forms are, all of them, less like; the last of all being, however, less unlike than some of the intermediate ones, since the fangs are there represented. In the "square" Hebrew the fangs were once more dropped, and the letter became w.

Finally, with respect to tau, which meant simply a "mark" or "badge," and which is generally explained as originally a "brand on cattle," it is clear that the simple cross of the Moabitic Stone would be more likely to be the actual mark used than the more ornate forms of later times, where first one, and then both arms of the cross have a terminal deflection.

*

It appears, therefore, upon the whole, that the alphabetic forms of the Moabite Stone strongly favour the view which is maintained by most critics, and which the names of the letters suggest, that the original Phoenician writing was pictorial, resembling in this respect the writing of the Egyptians and (most probably) of the Babylonians. The letters were the pictures of familiar objects, which pictures underwent a gradual corruption, the great object being to simplify, by reducing the character to forms which could be traced without removing the hand from the paper. The similar corruption of the Egyptian hieroglyphics into the character known as demotic is generally admitted, and has been well illustrated by Lepsius and others.

Another interesting palæographical question, on which the Moabite Stone throws considerable light, is that of the time at which the Greeks obtained the elements of writing from the Phoenicians. It has been strongly argued by Mr. Grote, and is now believed by many, that letters were absolutely unknown to the Greeks in the time of Homer and Hesiod (about B.C. 850-776), and were first introduced into Greece about the period of the first Olympiad, or soon after. The evidence of the newly-discovered Stone favours a much earlier date for the communication. The archaic Greek alphabet, as it exists in the earliest inscriptions, resembles far more closely the alphabet of the Moabite Stone than it does that of any subsequent period. In proof of this, we would refer especially to the following characters: beta, delta, zeta, iota, mu, xi, koppa, and tau, which correspond respectively to the beth, daleth, vau, zain, yod, mem, samech, koph, and tau, of the Phoenicians.

The early beta of the Greeks is angular, not rounded, and thus resembles the earlier, rather than the later, beth. It differs from the

As Gesenius, Rödiger, Baron Bunsen, E. Twistleton, Wright, and others. Wuttke, however, and Fürst maintain the opposite view.

beth by repeating in the lower limb the form of the upper one, an alteration due apparently to Greek ideas of symmetry. But in both limbs the angle is kept as the essential idea, the rounded form being a later introduction, and the open head (see plate, line 2) being absolutely unknown in Greece.

The early Greek delta is commonly like the daleth of the Stone, a simple triangle. In a very few cases the right arm descends a little below the point of junction with the base. But in no early Greek inscription is the head open (as in plate, line 2), or the left angle rounded (as in lines 3 and 4), or the right arm much produced (as in lines 2 and 3). In other words, the early Greek delta resembles closely that on the "Stone," while it differs considerably from those of the Assyrian tablets, the Eshmunazar sarcophagus, and the monuments of the Persian time.

The earliest form of the Greek bau (Bav, pronounced vau) was either Y or V, a form evidently derived from the vau of the Moabite Stone, rather than from any later Phoenician type. The later Greek vau, F, was a corruption of this, Y having first been changed, for expedition's sake, into, and then a second horizontal stroke having been added, as a diacritic mark, to distinguish vau from gamma.

The Greek seta, from the earliest times to a period later than Pericles, was always I, not Z. This form is only found in Phoenicia in the earliest period, being replaced by Z, uniformly on the Assyrian tablets, as well as in all the later inscriptions. The exclusive use of the perpendicular zeta by the early Greeks is an especially strong argument in favour of their having got their alphabet from the Phoenicians very considerably before the time of Tiglath-pileser II.

The most ancient form of the Greek iota was a zed placed diagonally (Z), the upper and lower arms being shorter than the line connecting them. This form will be seen by the plate closely to resemble the more ancient of the Phoenician types, only differing from them in the absence of a second line projecting towards the left below the upper arm of the letter. As, however, the Phoenicians continued to use this form as late as B.c. 650, no important argument can be drawn from this letter.

Mu, on the contrary, furnishes a strong argument in favour of the early derivation of the alphabet. It is only in the primitive Phonician alphabet that the angular or zigzag form of this letter obtains. In the later types curved lines replace the zigzag, or forms still more remote from the primitive ones. But the early Greek mu, which is spread out and has the last arm short (M), is exactly the mem of the Moabite Stone, except that the last line of the zigzag has been omitted.

The correspondency of the Greek xi with the Phonician samech,

whose place it occupies in the alphabet, had long been suspected; but the absolute identity of the two was first proved by our "Stone," which uses for samech the exact form*-a perpendicular line, crossed by three vertical bars -found to express xi, where it first occurs in Greek inscriptions.† As this form was superseded by a simpler one before the period of the Assyrian tablets and gems, we have here again an evidence favouring the early passage of the Phoenician. letters into Greece.

A similar result ensues from a consideration of the Moabite koph. The Greek koppa (Q), the original of the Latin Q, had never previously been found in a Phoenician inscription; and its form seemed so remote from the ordinary Phoenician types, that it was difficult to regard them as having any real connection. We now find that the original Phoenician letter was identical with the Greek, or differed from it only by having the vertical line somewhat longer. By the time of Tiglath-pileser II. (B.c. 745), the vertical line had been further lengthened, being carried to the top of the circle (P). As this is never the form of koppa among the Greeks, we must conclude that they obtained this letter in the first, rather than in the second, Phoenician period.

The Greek tau seems to be derived from the earlier, rather than the later, tau of the Phoenicians, from the fact that its arms are straight, and not deflected. The Phoenician tau is, in every case, a cross; and the only important difference between the earliest and the later forms is in the deflection of the arms at their extremities. The Greek tau, which is sometimes a cross with the upper limb but slightly developed (†), sometimes a mere T, with that limb wholly suppressed, has in every case the arms perfectly straight, with no sign of the terminal ornamentation which we observe in the later Phoenician.

The only Greek letters whose archaic forms resemble the later Phoenician types more than they do the earlier are lambda and pi. The early lambda, whether it takes the form of or, is always angular; the lamed of the "Stone," and of the Assyrian tablets and gems, is rounded. Contrariwise, the early pi is either round or square topped, and has never the angular head which marks it on the Moabitic monument. Now this angular head was laid aside

It was argued at first by some that the occurrence of this form upon the Stone was fatal to its pretensions to a great antiquity, since there was, it was thought, no precedent for its early use, and classical writers (Plin. H. N. vii. 56; Euseb. Chron. i. 13) ascribed its invention to Simonides, who flourished about B.c. 530. But as the Assyrian gems of the seventh or eighth century exhibit a very similar type (see the plate), and one manifestly derived from it, there appears to be no reasonable doubt that the form is really extremely ancient.

The zi is rare in early Greek inscriptions; but still it occurs occasionally. (See Böckh, Corpus Inser. Gr., vol. i. pp. 53 and 55; Rose, Inscr. Gr. Vetust., pl. 8, and p. 71.)

before B.C. 750, and the round head, which thenceforth continued in use, was adopted. So far, therefore, as the evidence of these two letters goes, the alphabet might have been communicated to the Greeks about B.C. 750-700. But the point in question has to be determined by the balance of evidence; and the balance of evidence is as nine to two in favour of the alphabet having passed into Greece in the course of the first, rather than the second, of its known stages -about B.C. 900, rather than about B.C. 750.

Such are the chief palæographical results which the Moabite Stone appears to us to have established. It favours, we think, the view that the "Cadmean" characters were originally pictorial-imitations, i.e., of familiar objects; and it helps us sometimes to determine the objects which the characters represented. It also lends important support to the view that the Greeks obtained the elements of writing from the Phoenicians at an early date, either before the time of the inscription of Mesha, or, at any rate, very shortly afterwards. These results may be, and no doubt will be, disputed; but we have little doubt that palæographical science will eventually accept them as established.

It is probable that the inscription may involve other important palæographical results which have escaped our notice. Certainly it must always be, unless superseded by some more ancient document, the final court of appeal in controversies as to the original shape and power of the characters by which Western thought has expressed itself since the dawn of civilization. It is at least a century and a half earlier than any other inscription that we possess in the same species of writing; and it is three centuries earlier than any other such inscription of any considerable length. By means of it we are informed what the characters were in which a contemporary and neighbour of Ahab and Elijah was in the habit of writing. As all the evidence goes to prove that one and the same system of written characters was spread ever Western Asia, from the borders of Egypt to Assyria, we may fairly assume that the books of the earlier prophets, the correspondence between Hiram and Solomon, Solomon's Proverbs and Songs, David's Psalms, Samuel's history, were originally thus written. Further, perhaps, it may be over-bold to go; but a suspicion forces itself on us that, in the characters of which the photographic traces are before us to-day, we see the forms of the letters in which, more than three thousand years ago, the Pentateuch itself was penned, and which "the finger of God" impressed upon the Two Tables.

G. RAWLINSON.