der of Cilicia another rock-sculpture had come to light. Here a priest is represented adoring the Cilician Herakles, who holds in his hands a cluster of grapes and a sheaf of corn. The images of the priest and god are accompanied by hieroglyphs, the first of the kind that had been seen by European scholars.

Similar hieroglyphics, however, eventually turned up, not in Asia Minor but at Hamah, the ancient Hamath, in Syria. They were engraved in relief on blocks of basalt, and were first noticed by the great traveller Burckhardt. But it was not until 1872 that they became known in Europe, when the late Dr. William Wright took casts of them which he sent to England. It was soon recognized that the "Hamathite" characters and the hieroglyphs of Ivriz must belong to the same system of writing.

In 1879, on the eve of an exploratory journey in western Asia Minor, the identity of the art of Ivriz with that of Boghaz Keni and Karabel suddenly flashed upon me. It followed that the "Hamathite" characters were Asianic rather than Syrian, and that we might expect to find them on the Asianic monuments. As a matter of fact, the photographs of Perrot showed that an inscription in the same characters was cut on the rocks of Boghaz Keni, and hieroglyphs, supposed to be Egyptian, were said to be associated with the monument of Karabel. I prophesied in the Academy that these latter would prove to be Asianic and not Egyptian, and staked the correctness of the discovery I had just made upon their be ing so. A few weeks later, with an escort of soldiers, I visited that haunt of brigands, Karabel, and there took squeezes of the hieroglyphs in question. They turned out to be, as I had prophesied, identical with the hieroglyphs of Ivriz, of Boghaz Keni, and of Hamath.

Meanwhile the site of the old Hittite capital, Carchemish, had been found by Skene and George Smith in the mounds of Jerablûs on the Euphrates. Excavations undertaken on the spot by the British Museum, about the time that my visit to Karabel took place, resulted in the discovery of more monuments in the same peculiar style of art as that of Asia Minor, and of the same peculiar system of writing. Art and writing alike thus belonged to the Hittites, and the fact that the human heads depicted among the hieroglyphs are identical in headdress and features with the heads of the sculptured figures made it clear that the system of writing must be of Hittite origin. Other facts soon came to support the conclusion; the boot, for example, with upturned point, which appears among the hieroglyphs, is found not only in the rock-sculptures of Asia Minor, but also distinguishes the Hittites of Syria portrayed on the Egyptian monuments.

The Hittites are alluded to in two or three passages of the Old Testament, but it is only since the decipherment of the Egyptian and cuneiform inscriptions that we have learnt what an important part they once played in the history of the East. The Hittite monarch, whose southern capital was at Kadesh on the Orontes, contended on equal terms with Egypt in the plenitude of its power, and summoned to his standard not only the Lycians of Asia Minor but Mysians and Dardanians as well. The Egyptian inscriptions bear the same testimony as the sculptured warrior of Karabel to the extension of Hittite influence in the West. Northern Syria had been wrested by them from Egypt after the fall of the Eighteenth Dynasty, and when the Assyrians first became acquainted with it they were so far the dominant people in it as to cause even Palestine to be ever afterwards known

at Nineveh as "the land of the Hittites." When the Hittite empire was broken up a fragment of it, under the name of "Hittite," still continued to exist to the south of the Gulf of Antioch, and the kings who engraved the cuneiform inscriptions of Armenia found Hittites in the neighborhood of Malatiyeh.

Ever since my discovery of the origin and connections of Asianic art I have kept the problem of the decipherment of the "Hittite" hieroglyphs continually in view. I had tried, or believed I had tried, every possible and impossible clue, only to find myself confronted by a blank wall. Eight months ago I still held that the problem was insoluble without the help of a long bilingual inscription. How it has been solved without any such help I will now try to explain.

As far back as 1880 I brought to light a short bilingual text, in Hittite and cuneiform, engraved on a silver "boss," and being a royal name, the Greek form of which is Tarkondêmos. The text gave us the ideographs of "king" and "country," as well as the phonetic value of me for another character; but otherwise the reading of both the Hittite and the cuneiform texts was involved in difficulties, and, as far as I know, was necessarily misleading. What, therefore, we might have hoped to have been the Rosetta Stone of Hittite decipherment ended only in leading the decipherer astray.

At the same time I pointed out another fact. The Hittite proper names preserved in the Egyptian and Assyrian inscriptions show that the usual termination of the nominative singular was s, while an examination of the texts makes it clear that this termination was represented by the picture of a yoke. It is also clear that the grammatical forms of the language were expressed by suffixes, and that the substantive and adjective agreed with one

another as in Latin or Greek. Another discovery of mine was the ideograph or "determinative" of divinity, which is prefixed to the name of a deity, and seems to present a sacred stone wrapped in cloths. German scholars next drew attention to the use of another sign as a word-divider, words being divded by it one from the other; while it had been recognized from the outset that the inscriptions are written in boustrophedon fashion, and must be read from the direction towards which the hieroglyphs look.

With these preliminary data the decipherers set to work. System after

system of interpretation has been proposed, each put forward with an equal amount of confidence, but satisfying none but its author. Before a system can be accepted it must fulfil three conditions. The phonetic values assigned to the characters must be such as to enable us to read, without forcing, the geographical names of the localities in which the several inscriptions are found-the name of Carchemish at Carchemish, of Hamath at Hamath, of Tyana at Tyana; the suffixes must reveal a consistent and coherent grammar to which parallels can be found elsewhere; and the inscriptions must yield a rational sense. Only when these conditions are fulfilled can the problem of decipherment be considered to have been solved.

What has principally stood in the way of the solution has been, not only the scantiness and imperfection of the texts we possess, but, still more, the inaccuracy and untrustworthiness of our copies of them. It is only recently that squeezes, casts, and photographs have at last given us accurate reproductions of such of the inscriptions as are not in the museums of London and Berlin. And one of the first results of a study of these was to show me that the ideograph of "country" or "district" had been confounded with that

for "king," though the bilingual "boss" of Tarkondêmos had long ago given us their distinguishing forms. The error had been committed by myself in the early days of Hittite research, and I have been followed in it by subsequent decipherers. But the error was vital. It prevented us from detecting those geographical names, through which alone, without the help of a bilingual, the decipherment of the texts was possible. As soon as I found that the native scribes have always carefully distinguished the two ideographs from one another, all the conditions were changed: I now knew in what group of characters I had to look for the geographical names.

Recent additions, moreover, to the number of texts known to us have also assisted the decipherer in another way. The same suffix is represented in them by more than one character; thus, in the case of the nominative singular, the goat's head (which must therefore have the value of 8) interchanges with the yoke. Thanks, too, to the fact that the hieroglyph of a man's head, surmounted by the priestly tiara, is attached to the figure of the high-priest at Fraklin in Cappadocia, I was able to read the group of phonetic characters accompanying the ideograph in the inscriptions of Carchemish, the native form of the Cappadocian word for "high-priest," having fortunately been given by the Greek writers Strabo and Hesychius. From this it resulted that the rabbit's head denoted the syllable ka.

Now in the inscriptions of Carchemish, and in them only, we find a geographical name, or territorial title, to which alone the determinative of "district" is attached. It consists of four characters, the last three of which are the rabbit's head, the character which the bilingual "boss" had long ago told us has the value of me, the head of a goat, while the first character is one

which is not met with elsewhere and may therefore be assumed to express, not a simple, but a closed syllable. As the last three characters read ka-me(î)s it is obvious that the first must be Kar. We thus get the name of Carchemish just where we should expect to find it. Besides the uninflected Karkames, an adjectival form of the name also occurs, which enables us to fix the values of some more characters.

There are two characters which from the frequency of their occurrence and the fact that they can be inserted or omitted at will after other characters, have long since been recognized to be vowels. For reasons, which it is needless to detail here, I have succeeded in fixing the value of one of them as a and of the other as i. The values of a few other characters have been obtained through their employment as suffixes. One or two Hittite suffixes have been made known to us through the proper names contained in the Egyptian and cuneiform inscriptions; thus, Khatti-na-s is "Hittite," Samal-i-u-s is "Samallian." The Hittite inscription on a bowl found in Babylon, again, has furnished us with the suffixes of the accusative singular, the first person of the verb and probably of the dative case. It begins with an ideograph, which Dr. Leopold Messerschmidt, has shown from a comparison of texts is the demonstrative "this"; then comes the picture of a bowl with a common suffix, denoted by the hieroglyph of a sleeve; then the name of a deity with its suffix; and finally the mason's trowel, which other texts show must have the signification of "mating" and to which a suffix is attached. The whole phrase must have some such meaning as "This bowl I have made for the god X," and the sleeve will denote the suffix of the accusative.

The decipherment of the suffixes has

disclosed an interesting fact. They agree in form and use with those of a language first made known to us by the famous cuneiform tablets of Tel el-Amarna.

Among these tablets are two in an unknown language, one of which is addressed to, or by, a certain Tarkhundarans, king of Arzawa. The name of the king is Hittite, and so raises a presumption that the language of the letters is Hittite also. The presumption has been confirmed by the excavations of M. Chantre, at Boghaz Keni. Here he has found other cuneiform tablets in a language closely allied to that of Arzawa. Thanks to the ideographs and stereotyped formulæ that occur in the letters of Arzawa, the meaning of several words and grammatical forms in them can be made out: thus, the termination of -s marks the nominative of the noun and -n the accusative. The remarkable agreement of the Hittite and Arzawan suffixes goes far to show that my reading of the Hittite characters is correct.

So, too, does the fact that the right geographical names occur in the inscription in which we should expect to find them. A stela, for instance, has been discovered on the site of the ancient Tyana which begins with the name of a priest-king. This is followed by his territorial title, to which the determinative of "district" is attached. The title, according to the values I have obtained for the characters, reads *-a-na-a-na-a-s. Nas is the suffix of a gentilic adjective with the nominative termination; the same suffix is found not only in the name Khattinas, which I have quoted above, but also in the Arzawa letters. Stripping the title, therefore, of its suffix, there remains *-a-na-a. What else can this be except Tu-a-na-a?

What I have said will, I hope, explain my method of decipherment. But it is usually only the proper names

and suffixes that are written phonetically. The roots or stems of the nouns. and verbs are more commonly expressed by ideographs. The pictorial nature of Hittite writing, however, not unfrequently gives us a clue to the meaning of the latter. And when once the texts are broken up into their constitutional parts so that we know where the name of an individual or of a country is found, and where we may look for the verb with its subject and object, the translation of the ideographs is comparatively simple.

But it must be understood that the decipherment of the inscriptions is still only in its initial stage. If it took half a century to complete the decipherment of the Persian cuneiform texts we must not expect to decipher the Hittite hieroglyphs in a day. All I can claim to have done is to have made a start and pointed out the road that others may follow.

Meanwhile such Hittite inscriptions as we possess have yielded little that is interesting. The three shorter inscriptions of Hamath record the restoration of a temple. The most perfect of the Carchemish texts is a long list of the titles of the priest-king. Two facts, however, have resulted from the decipherment which, to me at least,. were unexpected and surprising. On the one hand the name of "Hittite" is. confined to the inscriptions of Syria and the districts eastward of the passes of the Taurus; in the inscriptions of Cilicia and Cappadocia it does not occur. On the other hand, the language that has been revealed to us is,. on the grammatical side, extraordinarily like Greek. Thus the priest-king who is commemorated on the rocks of Bulyar Mader calls himself Sandanyas, "of the city of Sandes," the Cilician Herakles. The same perplexing similarity recurs in the case of Lycian grammar: how it is to be explained r do not know. Apart from its gram

matical forms I see nothing in Lycian that is Indo-European; and Hittite seems equally to be an Asianic tongue. Can it be that Greek is really a mixed The Monthly Review.

language, the product of early contact on the part of an Indo-European dialect with the native languages of the coast of Asia Minor? A. H. Sayce.

Keep a fairy or two for your children. -Ruskin.

A deeper import Lurks in the legend told my youthful years

Than lie upon the truth, we live to learn.-Coleridge.

The old Florentines of the Middle Ages had a noble conception of the uses of literature in the training of painters as well as children. "And to the old painter," writes Ruskin, "with his wild, weird, mysterious Etruscan instincts and ancestors, literature meant the Bible, legend, poetry, myth -it meant essentially imaginative literature-fairy tales an it please you." Literature to Plato meant pretty much the same thing; for he, too, would teach children by fables, which he says are "fictions, though there are in them some elements of truth." And by fairy tales Plato would open up the child's mind, for these half reveal and half conceal the truth, for the little child is as yet too tender to look upon Truth unveiled.

Classic fairy tales, myths, legends, and sagas, born in the childhood of the world, are the true food for little children; because the little child is psychically near to the childlike races of the early world, and the same things which appealed to the credulous barbarian appeal to him. The Puritan divines, Rousseau, as well as some modern writers on Education, would forbid the fairy tale in the schoolroom. "As for romances and

idle tales," writes Richard Baxter, "I have already shown in my book of Self-Denial how pernicious they are, especially to youth." "When thou canst read," counsels Thomas White, "read no ballads and foolish books, but the Bible, and the Plaine Man's Pathway to Heaven." Rugged enough was this pathway, as spelled out by the little Puritan child; but, happily for him, no one denied him his David and Goliath, his Daniel and the den of lions, his Joseph and his brethren; for the Bible always remained to him.

Rousseau sternly forbade fairy tales. The child was to learn only from real things within his experience, and his emotional nature was to be left severely alone. He dared to present the naked truth to the little child, which Plato would spare him. "Men may learn from fables," he writes, "but children must be told the bare truth; if it be veiled they do not trouble to lift the veil." So in his passion for realities, Rousseau would clip the wings of the child's imagination, and thus maim him for life; for the folk-lore and fairy tales not read in childhood miss their effect for ever.

Rousseau set many brains and pens at work on educational theories. There was Madame de Genlis, with her Adèle et Théodore, her amazing vanity, and her many followers; and in England he ushered in the didactic literature of the Aiken, Day and Edgeworth type, and instructive stories for children with a moral lurking behind