This Essay of Hincks exercised a decisive influence upon the future study of Assyrian. It demonstrated that, although the language was Semitic, the mode of writing was not Semitic; and for a time it divided scholars into two opposing camps. Those who followed Hincks maintained that the language was syllabic, and that each sign expressed a consonant associated with an inherent and invariable vowel. Those, on the contrary, who sought to assimilate it to the Hebrew system were of opinion that the signs represent simple consonants that might be preceded or followed by any vowel. We have noted the gradual recognition of the syllabic nature of the Assyrian writing. The earliest opinion was that the signs were both syllabic and alphabetical, and we have seen that Grotefend in his transliteration treated them in this manner. Hincks in his first Essay followed the same method, but he found the vast majority of the signs were syllabic, and his Table shows only four that are purely alphabetical. Then came the discovery that the language itself was Semitic, and the inference naturally followed that the writing was so likewise. Under the influence of this conviction, Hincks drew up his Second Table, showing only the consonantal value of the signs, and leaving them to be associated indifferently with the vowel sounds. But he remained in this opinion for a comparatively short time, and in the end of the same year he had reverted to his original view. The effect of the present essay was to establish the absolute syllabism of the language; and in a paper read shortly afterwards before the British Association ‘On the Language and Mode of Writing of Assyria’ (August 1850), he ‘maintained, in opposition to all other writers, that the characters had all definite syllabic values, there being no consonants, and consequently no necessity or liberty of supplying vowels.’ In this opinion he then stood alone. Rawlinson, in reply, expressed his belief that the signs had a syllabic origin, but that they were ‘subsequently used to express a mere portion of a syllable.’ ‘He could,’ he says, ‘adduce numerous instances where the cuneiform signs were used as bona fide letters.’[815]

In France, the opinion Hincks expounded in his second essay took immediate root. The logical instincts of the French mind clung with desperate tenacity to the conviction that a Semitic language could only be expressed by a Semitic mode of writing. Löwenstern at first (1845) thought that the signs represented some sort of mechanical union of consonant and vowel: that is to say, that there was a fixed portion of the sign to represent the consonantal sound, and a variable portion to indicate the conjunction of the vowel. ‘The signs,’ he said, ‘reproduce in part the same forms differently combined, which suggests a syllabic union in many of the signs.’ But he subsequently became the most thorough-going champion of the alphabetical theory. Botta only just found it ‘possible to conceive that the language was syllabic,’ yet he followed Longpérier in the luminous suggestion already described.[816] Even De Saulcy was haunted by dim fears that, after all, Assyrian might turn out to be syllabic, and the consistency with which he adhered to the opposite or Semitic mode of writing rendered his subsequent studies almost valueless. Rawlinson, as we shall soon see, yielded in time, and his transliteration of the Behistun inscription shows small traces of his early heresy, which he was still ready to defend in August 1850.

Such was the progress already made in decipherment when Rawlinson at length gave to the world some of the results of his labours in the same field. It will be remembered that in the autumn of 1847 he succeeded in taking a copy of the third column of the Behistun inscription. Whatever leisure he could command during the year 1848 and the early part of 1849 he devoted to its study; and when he returned to England in the autumn of that year, he brought the translation home with him. The work of publication was one of great difficulty, in consequence of the multitude of strange characters in many languages that had to be reproduced and corrected; and although Rawlinson remained in England till 1851, he was obliged to leave before it was accomplished. Some Continental writers chose to make this delay a matter of complaint against Rawlinson, whom they accused of deliberately withholding his copies for personal and selfish motives. It would be difficult, however, to mention anyone who was at that time at all likely to profit by their possession. The special qualifications of a decipherer are by no means common, and M. de Saulcy at least gave decisive proof that he did not possess them. M. Oppert, on the other hand, was still absorbed in the Persian and Median versions. But these gentlemen, and those who then shared their feelings, write as though Major Rawlinson had appropriated the rock of Behistun as well as the copy of the inscription that covered it.[817] They seem to forget that if they were prepared to undergo the same sacrifice and overcome the same difficulties, they could in a few weeks procure copies for themselves. What would have been more natural than to give the commission to M. Flandin, whose enterprise in such matters had already been so conspicuously illustrated? Nothing, however, could be farther from their intentions. They had no notion of foregoing the luxury of feeling aggrieved with the English soldier whose energy, like his genius, so far out-soared their own. Rawlinson was, we submit, fully justified in the course he adopted. He had obtained his copy at great personal sacrifice; no one in Europe was so qualified to accomplish the task of decipherment as himself, and the eighteen months he devoted to the task was not excessive. The subsequent delay in publication was incident to the nature of the work itself, for which he was not responsible.

He, however, lost no time in placing the general results at which he had arrived before the public. On January 19 and February 16, 1850, he read papers before the Asiatic Society ‘On the Inscriptions of Assyria and Babylon,’ and these, with a few additional notes, were published in March of that year.

He tells us he had found more than eighty proper names in the trilingual inscriptions, including those in the Behistun; and ‘by a careful comparison of the duplicate forms of writing’ them in the Persian and Babylonian columns he had been able, by means of the former, which were known, to determine the values of about a hundred Babylonian characters. The next step was by a collation of the inscriptions to ascertain ‘the homophones of each known alphabetical power.’ By this means he ‘added nearly fifty characters to those previously known through the Persian key.’ He confessed that his knowledge of the Babylonian characters was at present limited to these one hundred and fifty characters.[818] From the direction of his studies we may infer that these signs were chiefly taken from the trilingual inscriptions; and in that case they would be practically exhaustive; but they would amount to less than one half of those in general use in the Assyrian text.[819] The same process of comparison with the Persian translation enabled him to draw up ‘a list of about two hundred Babylonian words of which we know the sound approximately and the meaning certainly.’ But in addition to these, he was able, by ‘an extensive comparison of similar or cognate phrases, to add about two hundred meanings certainly, and a hundred more, probably, to the vocabulary already obtained through the Babylonian translation.’ He was thus acquainted with the meaning of about five hundred out of a vocabulary which he estimated as containing five thousand words.[820] These words, he explains, ‘are almost all found either in their full integrity or subjected to some slight modification in Assyrian’; and they enabled him ‘to arrive at a pretty correct notion of the general purport of the phrases in which they occur.’ Although his vocabulary was still limited to one-tenth of the vocabulary, it embraced ‘all the most important terms in the language’; and he found it sufficient for the interpretation of the historical inscriptions.

The present Memoir was intended simply as introductory to the subject, and he did not give a list of the one hundred and fifty signs with their values attached. We cannot, therefore, institute a comparison as yet with the Syllabarium already drawn up by Hincks. It is sufficiently clear, however, that he had not, at the time of writing the Memoir, realised the essentially syllabic character of the language. There are, he says, ‘cases where a single alphabetical power appertains to the sign,’ and, he adds, ‘it cannot certainly be maintained that the phonetic portion of the alphabet is altogether syllabic.’ ‘There is,’ he observes, ‘an extensive syllabarium; but at the same time many of the characters can only be explained as single consonants.’ There is no indication that he had as yet apprehended the principle that governs the combination of consonant and vowel, as recently expounded by Hincks, and which is interwoven with the whole structure of the language. Indeed he says distinctly: ‘I have neither adopted, nor do I conceive it possible to adopt, any system with regard to the employment of the vowels in Assyrian and Babylonian.’ In some other respects also he was still behind the great Irish scholar. Hincks, for example, had laid down that the distinction between the consonantal sounds is uniformly maintained; and the truth of this statement has been since confirmed. Rawlinson was, however, still of opinion that ‘the gutturals and sibilants everywhere interchange.’ ‘There is the greatest possible difficulty in distinguishing between k, d, and t. L and v interchange.’ It is evident also that he had still much to learn from his rival on the subject of the ideograms. It may be doubted indeed how far he had as yet apprehended the important place they occupy. ‘The names of the gods,’ he says, ‘are represented by signs which appear in some cases to be arbitrary monograms, but which are more generally either the dominant sound of the name or its initial phonetic power.’ He thought, for example, that the monogram for Bel was simply the letter ‘B,’ an idea that is wholly unfounded. He is of course aware that there are many other ideograms besides those used for the gods, but he gives them no sort of prominence. He, however, attributes an ideographic origin to the syllables. ‘When a sign represents a syllable,’ meaning apparently a compound syllable, ‘I conjecture that the syllable in question may have been the specific name of the object which the sign was supposed to depict; whilst in cases where a single alphabetical power appertains to the sign it would seem as if that power had been the dominant sound in the name of the object.’ But this is a purely academical question. The important point lay in precisely the opposite direction, and attention had been already called to it by Hincks. The peculiarity most necessary to emphasise is that in a vast majority of cases the pronunciation of the ideogram has no relation whatever to the name of the object it represents, nor, when it has a syllabic power, to the phonetic value of the syllable. Rawlinson, however, did good service in the present Memoir by laying down the first rudiments of the grammar, a branch of the subject that Hincks subsequently did much to elucidate; and he was also the first to bring into prominence the polyphonic character of the language. Hincks had indeed remarked that ‘many characters admit of two or more kindred values’; but Rawlinson farther shows that ‘certain characters represent two entirely dissimilar sounds—sounds so dissimilar that they cannot be brought into relation with each other.’ He gives as an example the sign for the vowel a, which also conveys the sound of ‘bar.’

But the great distinction of Rawlinson lay in his unequalled power of translation. Large numbers of Assyrian inscriptions were now before the world. The ‘Monument de Khorsabad’ had appeared in 1848; Layard’s collection followed in 1849, and included the inscription on the Black Obelisk found in 1846. So far only a few words had been made out with more or less of accuracy; but nothing had yet been done in the way of a connected translation. The few lines of the Khorsabad inscription which Hincks attempted in the Addenda to his paper (Feb. 26, 1850) had not as yet appeared.[821] Rawlinson, however, observed that many of the common expressions used at Behistun were adopted almost verbatim from the Assyrian annals; and it was the discovery of these known passages in the Assyrian inscriptions that first encouraged him to undertake their translation. He disclaimed all pretensions to be ‘a complete master of the Assyrian language’; and he still speaks of it as to a great extent unintelligible. ‘The first outwork,’ he says, ‘has been carried in a hitherto impregnable position, and that is all’ Indeed he is so discouraged by the difficulty of the task that he is sometimes disposed ‘to abandon the study altogether in utter despair of arriving at any satisfactory result.’ In consequence of the profusion of ideograms in proper names, he finds that their ‘pronunciation is a matter of exceeding difficulty, nay, as I think, of absolute impossibility’; and it was in this department that he achieved the least success. He, however, passes in review many of the principal inscriptions that were then known, and analyses the contents of each.[822] He begins with the earliest in date, the one taken from the North West Palace at Nineveh, which he ascribes to King Assur-adan-pal—really Assur-natsir-pal. He passes on to the inscription of his successor, whom he calls Temenbar II. (really Salmaneser II.), which covers the Black Obelisk, and it is to it that he devotes the largest share of attention.[823] His analysis, partly a verbal translation and partly a summary, fills no less than seventeen pages, and the achievement cannot fail to elicit unqualified admiration. The unfortunate failure to identify a large proportion of the proper names gives to it an unreal appearance that no doubt strikes the modern student unfavourably and may at first lead him to exaggerate its deficiencies. If, however, he is careful to remember that it is the first attempt of the kind ever made, his feelings will soon turn to astonishment that so much should have been correctly made out of what had hitherto been absolutely unintelligible. He may profitably compare a few passages with a modern version. For example, Temenbar begins: ‘At the commencement of my reign after that I was established on the throne I assembled the chiefs of my people and came down into the plains of Esmes, where I took the city of Haridu, the chief city belonging to Nakharmi.’ A recent translation of the same passage runs: ‘At the beginning of my reign when on the throne of the kingdom I had seated myself in state, my chariots and [my] armies I assembled. Into the depths of the land of Simesi I penetrated; Aridu the strong city of Ninni I captured.’ Again Rawlinson translates: ‘I went out from the city of Nineveh and crossing the Euphrates I attacked and defeated Ahuni, the son of Hateni, in the city of Sitrat, which was situated upon the Euphrates, and which Ahuni had made one of his capitals. Ahuni, the son of Hateni, with his gods and his chief priests, his horses, his sons and his daughters and all his men of war, I brought away to my country of Assyria.’ The modern version says: ‘I departed from Nineveh; the Euphrates I crossed at its flood; I marched against Akhuni, the son of Adini. The country of Shitamrat, a mountain peak on the banks of the Euphrates, he made his stronghold. The peak of the mountain I captured; Akhuni, with his gods, his chariots, his horses, his sons, his daughters and his army, I carried away and to my city of Assur I brought’;[824] and so on through the events of thirty-one years of the reign of the great king. On the other hand, it would be too much to say that even the sense is always preserved. There are, in fact, many and serious divergencies from the correct translation as it now stands, after more than forty years’ continuous study. It is impossible that it could have been otherwise with the means then at hand. The wonder is that so much could have been accomplished with one hundred and fifty imperfectly understood characters; and with only five hundred words arrived at conjecturally out of some six thousand. Rawlinson himself warned the reader that here and there ‘little dependence can be placed on the translation’; and he confesses that sometimes he could not ‘conjecture even the meaning of several passages.’ Notwithstanding his difficulty with the proper names, a host of new ones were now for the first time identified: Amanus; the Hittites; Chaldæans; the rivers Tigris, Euphrates, Belikh; the cities of Borsippa, Tyre, Sidon, Gabal, Caleh. On the other hand, the Akkadians now make their first appearance in modern history as ‘Hekdi,’ ‘which may be connected with the Armenian “Haik.”’ Hazael of Damascus is still concealed as ‘Khazakan of Atesh’ and ‘Jehu the son of Omri’ appears as ‘Yahua the son of Hubiri,’ ‘a prince,’ says the translator, ‘of whose native country I am ignorant.’ He called attention to the name of Yehuda in a Khorsabad inscription in connection with that of Hamath; but he hesitated to identify it with Judah. Indeed at this period he could not bring himself to believe that the son of the Khorsabad king was Sennacherib and his grandson Esarhaddon, as Hincks ventured to assert.[825] The signs for Sargon he transliterated ‘Arko-tsin,’ and those for Sennacherib, ‘Bel — — Adonim-sha’; but Esarhaddon came out almost correct as ‘Assar-Adan.’ He concludes his paper by an analysis of the inscriptions found at Khorsabad, containing the annals of ‘Arko-tsin.’

It was not till the following year, and till after Rawlinson’s return to Persia, that the publication of the third column of the Behistun inscription was completed. It fills the fourteenth volume of the ‘Journal of the Royal Asiatic Society,’ and the greater part of it was laid on the table before May 1851.[826] A portion of the expense was defrayed by a Government grant, made at the suggestion of Lord John Russell; and the volume consists of seventeen large plates containing the cuneiform text with transliteration and a Latin translation. Then comes an ‘Indiscriminate List of Babylonian and Assyrian Characters,’ with their phonetic powers, and also such ideographic values as had been ascertained. The list includes two hundred and forty-six principal signs, many of which are followed by others varying in form; and generally representing the different methods of writing found at Persepolis, Babylon and Nineveh. An analysis of the text, extending over a hundred pages, follows, but it has not been carried farther than to the end of the first column. The ‘Memoir on the Babylonian and Assyrian Inscriptions’ is even more incomplete. It covers only sixteen pages, and breaks off in the middle of a sentence, before the analysis of the second sign was concluded. No explanation is given of this abrupt termination.

The discovery of a separate sign for each combination of vowel and consonant, explained by Dr. Hincks in his Appendix of January 1850, no doubt exercised considerable influence on Rawlinson,[827] and we are now in a position to recognise the full effect it produced in the progress of the study. How far Rawlinson independently divined the existence of some such principle is by no means clear, but we have no doubt that when he read his first paper to the Society he was still of opinion that the alphabetical system entered largely into the Assyrian language. ‘Many of the characters,’ he then said, ‘can only be explained as single consonants.’[828] In accordance with this view, when he had occasion to refer to the Assyrian characters, he uniformly gives them purely alphabetical values, although in many cases their correct syllabic values had been already definitely fixed in Hincks’s Essay.[829] Hincks himself pointed this out in one notable case. In 1850, Rawlinson said that the suffix of the third person plural is a simple n; but soon afterwards Hincks showed that the consonantal termination is followed by u. This opinion Rawlinson adopted in 1851, and remarked of the word ‘Yatipsu,’ ‘the termination in u marks, of course, the plural number like the Hebrew.’[830] ‘When,’ says Hincks, ‘the commentary was published [in 1850] no u could be discovered. The sign he now reads su was a simple s.’[831] These facts are not sufficiently accounted for by the explanation Rawlinson gives in his present Memoir. ‘In the articulation kat, for example, which is composed of two characters, ka and at, either one or other of these signs must represent a simple letter rather than a syllable; and as this peculiarity of expression pervades the whole Assyrian alphabet, I think I am justified in still adhering to the statement which I announced last year, that the phonetic signs were in some cases syllabic and in others literal.’[832] However this may be, he now finally abandons the description of the signs as letters, and no simple alphabetical values are to be found in his ‘Indiscriminate List.’ There can be little doubt that the alteration in the method of writing is to be best explained by a corresponding change of opinion.[833] However great may have been his obligation to Hincks, he soon made the discovery his own. He corrects the errors and supplies the deficiencies of his predecessor. He suppresses the twofold signs for a, and limits the regular syllabic combinations to the three vowels a, i and u, which thus yielded six instead of seven values for each of the consonants. Following Hincks, he accepts only fifteen distinct consonantal values, but he prefers to use z and kh in place of the j and g of Hincks, now written z and ḫ. With these he has given one hundred and seven simple syllabic combinations, a much larger number than really exists; but he has no less than seventy-eight correct, out of a possible number that slightly exceeds eighty.[834] He, however, took no account of the signs that indicated the exceptional combinations with the vowel e, which amount to about a dozen. This defect was soon afterwards noticed by Hincks, who, however, signally failed in his attempt to identify them. On the other hand, we are now introduced for the first time to the compound syllables that form so large a portion of the Babylonian signs. These, unlike the simple syllables so successfully treated by Hincks, consist of two consonants separated by a vowel. It is true that a few made their appearance in Hincks’s short translation from the Khorsabad inscription (February 1850), such as ‘sib,’ ‘kun,’ ‘bul’ and ‘gur,’ along with three others that are incorrect.[835] We also learn from a later publication that before the appearance of Rawlinson’s Syllabarium Hincks knew the values of upwards of twenty other compound syllabic signs.[836] But they do not seem to have been made known to the world, and they all, with the exception of five, are now met in the Syllabarium for the first time. Rawlinson gives sixty-eight of these syllables, and no less than fifty are correct. He has been unable to give any phonetic value to sixty out of his two hundred and forty-six principal signs, but in several cases he has determined their ideographic meaning. The distribution of the signs, as explained by Hincks, considerably diminished the number of supposed homophones, and the compound syllables now enumerated tended in the same direction. But Rawlinson dwells with increasing bitterness on the extreme confusion introduced into the language by polyphones. He complains that ‘after years of laborious research he has overcome the difficulty to but a limited extent.’ ‘The meaning of a word,’ he says, ‘may be ascertained from the trilingual inscriptions, or from its occurring in a variety of passages with only one possible signification; but unless its correspondent can be recognised in some Semitic tongue it is often impossible, owing to the employment in it of a polyphone character to fix its orthography; and this uncertainty presses on the student with almost crushing severity.’[837] In addition to this, he had to contend with the difficulty that besets all early decipherers—the inability to distinguish between his own correct and incorrect values, where the latter often cause more confusion than if the sound were still regarded as doubtful. Notwithstanding all these obstacles, his transliteration was sufficient to afford a considerable knowledge of the nature of the language; and to enable the student to recognise the connection of the words that resulted with their Semitic relatives. In the forty years that elapsed between the version we are now considering, and that given by Dr. Bezold, a whole army of scholars has been ceaselessly at work upon the Babylonian and Assyrian inscriptions, and the point they have reached is naturally far in advance of that in which it was placed by the first Essay of Rawlinson. The comparison of a few passages, taken almost at random, will enable the reader to appreciate the position the study had reached in 1851.[838] The first lines of the inscription are rendered thus by Rawlinson, De Saulcy and Bezold.

The passage is thus translated by the three scholars:

In the August of this year (1851) Rawlinson was able to announce that he had met with an inscription that satisfactorily fixed the date of the Lower Assyrian dynasty. We have said that this was in reality done long before by Longpérier when he identified the Khorsabad king with the Sargon of Isaiah (1847); and by Hincks, who was satisfied that the builders of the later palaces were Sennacherib and Esarhaddon (1849). But we have seen that Rawlinson long refused to recognise these identifications as satisfactorily established. At length, however, he found in a tablet from Khorsabad (Pl. 70 Botta) an account which he acknowledged referred to the capture of Samaria by Sargon in the first year of his reign; and he also succeeded in identifying ‘Omri.’[840] He thought that Sargon, whose identification he acknowledged in his Analysis of the Behistun inscription,[841] was the same personage as Shalmaneser, who figures in the same inscription as the conqueror of Ashdod; and he made out the names of other cities that fell before the arms of that king: ‘Hamath, Beroea, Damascus, Bambyce, Carchemish.’

But the inscription to which he now more particularly drew attention was recently found by Mr. Layard on a colossal bull at the great entrance of the Kouyunjik Palace; and Rawlinson speedily recognised that it contained the Annals of Sennacherib, its founder, and son of Sargon. The annals extend only to the seventh year of the king, but they recount the subjugation of Babylon in the beginning of his reign, and the defeat of Hezekiah and the capture of Jerusalem in his third year. The narrative agrees with what was already known from the Hebrew writings and from Polyhistor. The discovery, in a cuneiform inscription, of the three names Hezekiah, Jerusalem and Judah, and an account of events related in the Book of Kings, naturally stimulated the interest of a wider public than is generally occupied with archæology.[842] From this period dates the great popularity these studies enjoyed for a time, a popularity that culminated more than twenty-five years later by the despatch of George Smith on a mission to the East by the ‘Daily Telegraph.’

A year after the publication of the third column of the Behistun inscription, Hincks read a paper ‘On the Assyrio-Babylonian Phonetic Characters’ (1852), which may be regarded as having closed the early stage of inquiry into the subject.[843] In this essay he contributes no less than a hundred and eighteen new values, of which sixty-eight certainly, and possibly more, are correct. When these are added to the Syllabarium of Rawlinson, upwards of two hundred correct signs, in addition to those for the vowels and diphthongs, were now at the disposal of the decipherer. It had also been proved conclusively that ‘the characters all represent syllables and were originally intended to represent a non-Semitic language.’ In opposition to the system that still found an advocate in De Saulcy, it was shown that ‘instead of the vowels being unrepresented, or only represented by points, as in all Semitic writing that was first applied to a Semitic language, we have in the cuneatic inscriptions every vowel definitely expressed.’ This new Syllabarium demonstrates for the first time how extensively polyphony prevailed. Indeed one of its chief merits consists in the enumeration of the different values expressed by the same sign. This had been done to a slight extent by Rawlinson, who puts the polyphones in an apologetic manner in a separate column, under the heading ‘Phonetic Powers arising from Ideographic Values.’ This excited the contemptuous criticism of De Saulcy, who was still so far from appreciating the true nature of the language that he declared: ‘Either this language was for the Assyrian an inextricable gâchis, or one or other of these values must be chosen.’[844] The present Memoir of Hincks, which must soon after have fallen into his hands, ought to have convinced him that the former alternative is the only one available. Indeed the number of polyphones is so great that the two hundred and fifty-four characters which Hincks now deals with express no less than three hundred and forty-four different values. In the Appendix to the Khorsabad inscription (January 19, 1850) it will be recollected that he gave seventy-one simple syllabic values, of which we found fifty-seven correct. In a lithographed paper, presented to a meeting of the British Association in the course of the same year, he added to their number, so that, with the vowels, his contribution amounted to a hundred. These apparently include the twenty-five (correct) compound syllables already mentioned. In the present Memoir he added a hundred and eighteen new values (sixty-eight correct), so that he claims to have discovered by his own unassisted ingenuity no less than two hundred and eighteen values. He acknowledges that he is indebted to Rawlinson for seventy-seven in addition to these, and he states that they were substantially agreed as to the signification of one hundred and seventy-seven signs.[845] They disagree as to forty-nine; but the disagreement, generally speaking, does not extend to the consonantal value; it arises from the doubt as to whether the sign conveys the value of e or o, as Hincks thought, or of i or u, as Rawlinson maintained. Hincks frankly confesses he received seventy-seven values from Rawlinson after 1850; but Rawlinson has not told us how many values he borrowed from Hincks during that year. The probability is they were extremely few, if any; the transliteration and translation of the inscriptions taken at Behistun and from the Black Obelisk were made before the Appendix was sent to press, and neither could have been accomplished unless Rawlinson had previously drawn up a very comprehensive list for his own use. No doubt he took full advantage of Hincks’s paper to introduce occasional corrections and emendations, and it is to be regretted that he has not gratified our curiosity as to the extent of his obligations.

In the present Memoir Hincks modified in some respects his original mode of writing. Following Rawlinson’s example, he has discarded the use of c in favour of k for the Koph series; and he adopts s to express the three Hebrew sounds of s, s̱ and š̱ (s, ts, and sh). He also follows Rawlinson in substituting z for j, and kh for g. All these modifications have been accepted except the last, which is now written ḫ. On one other point, however, he was less conciliatory. In deference to Rawlinson, he drops his two sounds for a: his long ā becomes now simply a; but he insists on the distinct recognition of the union of the consonant with e or o. ‘We must,’ he says, ‘consider the seven forms which might belong to each.’ These forms, therefore, are now a, e or o, i and u; and one of the chief points of disagreement with Rawlinson is that the latter ignores the sounds of e or o and substitutes either i or u. Hincks was quite right in maintaining that Rawlinson unduly neglected the vowel e; for his ‘Indiscriminate List’ only contains one syllable formed with e, viz. ep. Hincks was, however, wrong in supposing that there is any regular syllabic combinations so framed. The regular syllabic combinations are six, not seven, and they are formed with a, i and u only—as Rawlinson rightly saw. The combinations with e are exceptions to the rule, and have been ascribed to local or dialectic changes.[846] They amount to about twelve and, strange to say, only one was correctly identified by Hincks (te). The o sound seems to be practically unknown.

In the passages we selected to illustrate Rawlinson’s transliteration of the Behistun inscription, we placed that given by De Saulcy two years later side by side with it, not on account of its intrinsic merit, but because of the claims put forward on his behalf by some of his countrymen. It seems to be generally admitted that the honour attaching to the first decipherment of the Babylonian inscriptions cannot be justly claimed by more than three scholars—Hincks, Rawlinson and De Saulcy. We have endeavoured to lay before the reader the contributions made by the first two. There can be no doubt that Hincks displayed remarkable insight into the formation of the language, and that his ingenuity in detecting the value of the signs, and in recognising their relation to one another was very great. Whether he would have been able to go farther and acquire equal distinction as a translator is another question. His genius seems to have been more adapted to elucidate matters of grammar and philology. Rawlinson had a rare ability of assimilating the suggestions of other scholars so quickly as to be almost oblivious that they were not original, and of carrying them rapidly to a perfection that was all his own. Thus Hincks’s elementary Syllabarium of 1850 appears in Rawlinson’s Memoir of 1851, so vastly improved as to be practically an independent work. But the translation of the inscriptions was entirely his own, and in this department Hincks never entered into competition with him. Here, according to M. Menant, his rival was De Saulcy. It is unfortunate that De Saulcy’s early contributions should have become almost inaccessible, and we have not found any detailed account of their contents.[847] They date from 1847, but his earliest efforts, even according to M. Menant, only deserve mention because the author himself was disposed to treat them too severely.[848] On June 20, 1847, in a letter to Burnouf, he attempted to identify some of the kings mentioned in a genealogical fragment found at Van. Ten days later he endeavoured to solve the riddle of the second name in the Khorsabad inscription. In the following July he suggested that the name on the Michaux stone should be read ‘Saosdoukin’; and in December he imagined that he had found the sense of the Van inscriptions. He, however, honestly confesses that in the light of subsequent knowledge all these efforts were vain. ‘He has,’ he says, ‘passed the sponge over all he has hitherto done, and has recommenced the study of Assyrian ab ovo.’[849]

In 1849, however, he contributed two pamphlets which it is admitted were of greater importance. In the first, which appeared on September 14, he undertook a transcription of the Babylonian column of the Elvend inscription with translation and analysis. He succeeded, we are told, in separating the Babylonian signs correctly, so that each group could be compared word for word with the Persian. In his analysis he is said to have justified the values he attributed to each sign, and the meaning he attached to each word. He regarded the signs as purely consonantal, and sought to bring them all into relation with the twenty-two Hebrew letters. He seems to have thought that some of the signs are capable of division: one portion fixed, representing the consonant, and the other a variable appendix indicating the vowel. In the second pamphlet, dated November 27, he treated the Persepolitan inscriptions in the same manner, but with a growing suspicion that the Assyrian letters might after all turn out to be syllabic.[850] By this means he arrived at the consonantal sound of a hundred and twenty signs which M. Menant says were generally correct. When Rawlinson published his Syllabarium, in 1851 (two years later) De Saulcy observed that sixty-eight of these one hundred and twenty signs received the same values. Not satisfied with this success, he seems actually to have thought that Rawlinson had borrowed them without acknowledgment from him: ‘J’avais donc lu et publié avant M. Rawlinson soixante-huit des valeurs exactes publiées par lui; il eût été de bon goût, peut-être, de prendre, ne fût-ce qu’une seule fois, la peine de citer mon nom.’[851] It is, of course, quite impossible to admit the claims made on his behalf. We have not been able to see the signs to which reference is made, but it is quite certain that, in September 1849, there were not sixty-eight signs in the Persepolitan inscription still remaining unknown either to Hincks or Rawlinson. It must be recollected that De Saulcy’s pamphlets appeared more than two years after Hincks had already accomplished a somewhat similar classification with a considerable measure of success; and therefore after the consonantal values of a large number of signs had long been correctly ascertained. On the other hand, the earlier papers of Hincks were quite accessible to De Saulcy, and we learn from Mohl that they were well known in France a year before the appearance of his two pamphlets.[852] ‘What,’ asks M. Menant, ‘did Rawlinson owe to De Saulcy’s labours on the Assyrian text? It is impossible to say,’ he answers, ‘for Rawlinson has not given an account of his preliminary studies.’[853] The answer is, however, much simpler than this. Rawlinson owes nothing to De Saulcy, for the reason that his transliteration of the Behistun inscription was accomplished before he left Bagdad in October 1849,[854] and therefore before it was possible for him to receive even the earliest of De Saulcy’s pamphlets. Menant afterwards concedes that Rawlinson’s work on the Obelisk proves preliminary labours which he graciously admits may justly claim to be independent.[855] Nor is it true, as Menant says, that ‘it was by following De Saulcy’s steps that all later progress has been accomplished.’[856] The precise opposite approaches more nearly to the truth. It was by abandoning the alphabetical system, to which De Saulcy clung with strange pertinacity to the last, that all later progress was in reality accomplished; and two months after these pamphlets were written this was precisely what was done by Hincks with unmistakable perspicuity in the Appendix to his Khorsabad Essay.

The two tracts of De Saulcy, written in 1849, gave the transliteration according to his peculiar system of the whole of the Achaemenian inscriptions accessible to him. The text of the Behistun inscription had not, of course, escaped as yet from the jealous hands of the English Major. The translation of these inscriptions was comparatively easy, for it was only necessary to follow the Persian version, which was already known. De Saulcy may therefore claim to be the first who accomplished this task, which neither Hincks nor Rawlinson thought necessary to attempt. But De Saulcy did not rest satisfied with this achievement. On February 3, 1850, he published a transliteration of ninety-six lines of the inscription engraved upon the bulls at the entrance to Khorsabad; and he accompanied it by a translation, which two years later he still considered was sufficiently accurate. On February 12, he also contributed a Memoir on the Royal Names at Nimrud.[857] These two publications appeared between Major Rawlinson’s lectures of January 19 and February 16. In the first lecture, Rawlinson gave the earliest translation of a purely Assyrian inscription that had ever appeared, with the exception of the few lines rendered by Longpérier and Hincks. It was taken from the Black Obelisk, and he promised to read at the next sitting a précis of the Khorsabad inscription.[858] It was clear, therefore, that he had already prepared it. Meanwhile, after this announcement was made, and thirteen days before it was carried into execution, De Saulcy’s translation appeared. This forms the second long Assyrian inscription to be translated, and it can scarcely be denied that De Saulcy and Rawlinson had worked upon it independently of each other. The report of Rawlinson’s second lecture was given in the ‘Athenæum’ on March 2. It is, as we have said, impossible to estimate the comparative merits of the two works, because we have not seen De Saulcy’s pamphlet. It can only be judged by what we know of that writer’s later acquirements.

The transliteration from which we have quoted made its appearance in February 1854, having apparently been sent to press in June 1853, or some three years after Rawlinson’s version of the same text.[859] It cannot, therefore, claim the indulgence so willingly accorded to a first effort; on the contrary, it is distinctly put forward as a rival Essay, intended to prove ‘the essential error of Rawlinson’s method of reading.’ It was designed at the same time to show that not one of the general results reached in the two Memoirs of 1849 had now to be abandoned, and also to establish his legitimate right to a large share of priority of discovery to which De Saulcy says he attached great value. We have already disposed of his claims to priority so far as regards the consonantal values of the Babylonian signs found in the trilingual inscriptions, and we are at a loss to imagine in what other direction he fancied that he had achieved priority. So far indeed from his studies having ever been in advance of his contemporaries, they uniformly lagged far behind, and he evinces a singular inability even to follow the results obtained by their genius. A remarkable instance of this is seen in his rejection of Longpérier’s reading of ‘Sargon’ in the Khorsabad inscription. ‘The kh, k or g,’ he writes in 1852, ‘is in reality a d’; and in the Table he published in 1854 it is actually found under that heading. He regrets that he is obliged to renounce all hope of finding the ‘Sargon of the Bible’ at Khorsabad; and he reads ‘Sardon’ instead.

De Saulcy still adheres to his alphabetical interpretation, and it is no doubt the syllabarium of Rawlinson that is ‘the essential error’ he sets himself to remove. His transliteration accordingly consists, as the reader will have observed, of an immense agglomeration of consonants which the student is left to bring within the possibilities of human utterance as best he may. It is clearly a comparatively easy task to arrange a number of signs according to the simple consonantal sounds they contain. Hincks reached this point in 1847, and De Saulcy’s latest effort seems to carry us back to that rudimentary stage of the inquiry. Here we find the signs distributed among the various classes of gutturals, dentals and so on, exactly as in Hincks’s Table seven years before. It is possible that the later writer is more complete and accurate; it could scarcely have been otherwise, considering the large amount of data now available and the impossibility of not being guided, to some extent at least, by the Syllabarium of Rawlinson and Hincks. Yet in this elementary work of simple classification there are numerous errors he might have escaped if he had condescended to place more dependence upon their authority. Thus, for example, among his gutturals we find (of course bare and stripped of their vocalic garments) the signs for ip, up, al and zi. His dentals include two gutturals, the signs for ga and gi. Among his labials he gives us the signs for as and ku. Among his linguals are those for ki and su and they include the signs for the syllables tar, kur and rit; while the sign for ul is found among the sibilants. Notwithstanding all his efforts to escape syllabic values, he was forced to enumerate a few—kam, ak, akh, at, bar, far, in or an, ar, as, is. Some of these are correct; but not even here would he submit to authority, and he has accordingly blundered. His ak should be uk, his akh al, his far par, and of the two signs he gives for as, one should be si and the other sur, while his is should be us. With few exceptions all these might have been found correctly given by Rawlinson three years before. Hincks had long ago pointed out in his Khorsabad Essay (June 1849) that a clear distinction is maintained between the vowels, and between the surd and sonant consonants. Yet here we find the signs for m, w, ou and b, and those for l and r all classed indifferently together. His treatment of the vowels is not less behind the knowledge of the time. The single vowel a is represented by no less than seven signs that really express an, a, ap, i, ruh, man, it. Two of these signs, according to De Saulcy, also express ha, and one either e or i. Ha has four signs, none correct. They are really the signs for a, it, i and il. He was correct in supposing the language contained two diphthongs, ai and ia; but neither of his signs for ai is correct: one has, in fact, the value of tir.

It may be said in conclusion that on all points of difference between Rawlinson and De Saulcy, both as regards the theory of the language and the details of its expression, Rawlinson was right and De Saulcy hopelessly wrong. De Saulcy was not only unable to teach Rawlinson anything, but, as we have already observed, he was incapable to a very remarkable degree of apprehending the truth from others. He lived for many years afterwards, but his Essay on the Behistun inscription seems to have been the last occasion on which he meddled with cuneiform studies. He probably recognised more clearly than some of his admirers how incompetent he was to make any useful contribution to the subject. It is impossible to refrain from sympathy with him. He tells us he spent a whole year in ‘comparing sign by sign and transcribing all the Achaemenian texts without exception.’[860] Nothing is more calculated to overwhelm the mind with despondency than to pass years of fruitless toil amid such arid wastes as these and to discover in the end that the natural ability to make useful application of the knowledge acquired is wholly wanting. For the true genius of a decipherer is a rare gift, and no amount of industry or learning can compensate for its absence. Hincks and Rawlinson possessed it with exceptional intensity. Many of the other scholars whose labours we have reviewed were endowed to a less degree—Grotefend, Jacquet and Lassen. Even Longpérier, in the few lines he contributed to the subject evinced no little aptitude in this direction; but De Saulcy was singularly deficient in the special qualifications it required. It would have been more worthy of the position he occupied in other departments of study if he could have restrained the irritation that the consciousness of the waste of so much effort could not fail to produce. It was lamentable that he should fretfully pretend to have anticipated the discoveries of Rawlinson, or that he should have presented his own crude performance as a possible rival to his. It would almost seem, from the extreme rarity of his pamphlets, that he endeavoured to suppress the evidence of his failure, and it would be well if his countrymen were to allow his work in this department to pass out of the reach of farther controversy.

The translation of the Babylonian Column of the Behistun inscription was apparently thought at the time to dispense with the necessity of any special publication of the Semitic columns of the other Achaemenian inscriptions at Persepolis and Naksh-i-Rustam. De Saulcy had indeed devoted himself to this portion of the subject in 1849, and Menant informs us that ‘all the trilingual inscriptions then known were already translated.’[861] But they do not appear to have attracted the attention of any more competent scholar till 1859, when M. Oppert published a portion of them in the second volume of his ‘Expédition en Mésopotamie.’ He gave the text, with transliteration and translation of the Window inscription L and B (Darius); D and E (Xerxes), and the unilingual H (Darius) from Persepolis; the long inscription and the three short ones at Naksh-i-Rustam, the K of Xerxes at Van, and the S of Artaxerxes Mnemon at Susa. He also gave a new translation of the Behistun inscription without text or transliteration. A peculiar feature of his book is that he has made a transcription of the inscription into Hebrew characters. A complete edition of the Babylonian columns of the Achaemenian inscriptions was published by Dr. Bezold in 1882, with text, transliteration, translation and commentary, and is now the standard edition.

We have now brought to a close this tedious history of the various steps that led up to the decipherment of the Achaemenian inscriptions; and we have described the share taken in its accomplishment by a long succession of scholars, from Tychsen to Oppert. The whole of these inscriptions were now interpreted and their contents made known to the world. The difficulties of the cuneiform character, which at first appeared insuperable, were at length surmounted.

The subject for a long time seemed to yield no results at all commensurate to the labour and ability lavished upon it. Its interest seemed to be limited to the arid domains of philology, or at best to throw a sidelight upon a few matters of no great importance in ancient history. Some scholars were gratified to find that their old and greatly maligned friend Herodotus was ascertained to be much more trustworthy than was long supposed; but these were matters that could only affect a small and comparatively worthless class of dilettanti. At length, however, there came the great Assyrian discoveries and the apparition in the cuneiform records of ‘Jehu, the son of Omri,’ and a host of other notabilities of sacred history. The study was raised at once, especially in England, to an entirely different plane of interest. Lectures began to be delivered upon it throughout the provinces; books were written by Vaux, Bonomi, Fergusson, and many others, to explain the subject to the public. The great work of Layard was quickly followed by a popular edition, and was translated into German. Curiosity was stimulated by the appearance, in 1853, of an account of the results of farther explorations. Whether the new learning would tend to confirm the ancient records or whether it would compel a revision of cherished beliefs began to be debated in many quarters, far beyond the circle of learned societies.

An account of the progress subsequently made in the knowledge of the Assyrian language lies beyond the scope of our present work. It was indeed a happy accident that the power of reading Assyrian should have been acquired just as a countless number of inscriptions in that language were brought to light. Excavations continued at Nineveh and elsewhere under the direction of Mr. Layard during the winters of 1849 and 1850; and in the autumn of 1851, Colonel Rawlinson returned to Bagdad. He was charged with the general supervision of the work, while Mr. Hormuzd Rassam assumed the practical direction in the field. Parliament sanctioned a grant of three thousand pounds, and many large sums were contributed by private individuals, including five hundred pounds from Lord John Russell.[862] Soon after, Rassam discovered the famous inscription of Tiglath Peleser, which afforded the earliest glimpse into a long-forgotten history. In it the genealogy of the Assyrian kings was traced back to the fourteenth century B.C., and the names of no less than twenty-five sovereigns were recorded.[863] The inscription itself was written at a time when Assur was still the capital of the kingdom and Nineveh was too unimportant to be mentioned. Rawlinson was surprised to find that the language was more polished then than at a later time, and he was obliged to admit that the discovery ‘annihilates all my theories about the modernicity of Assyrian civilisation.’[864] Shortly after, he was able to announce that ‘all the Assyrian kings mentioned in the Bible have now been identified,’[865] and many others who occur in profane history, so, that almost a perfect list has now been obtained. Two French expeditions were engaged at the same time on the work of exploration. M. Place, from 1851 to 1854, devoted himself chiefly to Khorsabad, though with scarcely the success his perseverance merited. But the chief effort was made by the Commission headed by M. Fresnal, 1852-4, which included M. Oppert among its members, and which concerned itself principally with the exploration of the ruins of Babylon. Meanwhile Southern Babylonia was explored by Mr. Loftus, whom we have already mentioned in connection with Susa; and by Mr. Taylor, the Vice-Consul at Bussorah. It is to their labours during the winters of 1853 and 1854 that we owe the recovery of the history of the Early Babylonian Empire that long preceded even the foundation of Assyria. The forgotten cities of Nippur, Erech, Larsa, Ur and Eridu were once more summoned to surrender the records of a civilisation reaching back many thousand years before the Christian era. In 1854, Rawlinson was able to send home a list of eighteen of the primitive kings of the ancient Babylonia and of twenty other personages of less exalted station; and he records his surprise at the discovery of ‘monarchs who must have reigned before the establishment of the Assyrian Empire.’[866] Till then it was generally held that Babylon owed its foundation to the late period of the great Nebuchadnezzar.[867] Early in 1854, a fruitful discovery was made in the Lion Room of the North Palace of Assurbanipal at Kouyunjik.[868] Here large numbers of tablets were found, which subsequent investigation showed to consist of lexicons and phrase-books to enable the student to acquire the primitive language of Babylonia, from which it afterwards became apparent the larger portion of the Assyrian literature had been derived. Rawlinson was the first to detect the existence of this language in a tablet sent to him from Larsa by Mr. Loftus.[869] He announced the discovery in a valuable paper, contributed to the ‘Journal of the Royal Asiatic Society’ in 1855, on the Early History of Babylonia. At that time, however, he had made but little progress in this new study, for he says: ‘I have no hesitation in pronouncing the language to be Semitic.’[870] In the following year he found out his mistake, and, having carefully studied the vocabularies from Kouyunjik, he speaks of it more guardedly as ‘the Chaldean or Hamitic language of Babylonia.’[871] Six years later, we still hear of the ‘Hamitian language, of which not much is yet understood.’[872] For a time it was known also as the Proto-Chaldean;[873] Hincks seems to have been the first to call it by its later name of Akkadian,[874] but Rawlinson was the earliest to make any considerable progress in its study. In 1866, he endeavoured to translate the tablets bearing on astronomy and other scientific subjects; but he found ‘the primitive Babylonian language’ was so extensively employed in these documents that it was ‘advisable to undertake a thorough examination of this ancient and most difficult language.’ The result was that he thought it was intermediate between the African languages and the Proto-Turanian or Finno-Ugrian, which he proposed to classify under the name of the ‘Erythean Group.’ He considered it came from the uplands of Central Africa, and was the speech of the Akkads or Highlanders. From that circumstance it gradually acquired the name of Akkadian, suggested by Hincks.[875]