opened a course of lectures on the Malayan language in the school of the living languages. Not to mention that this language has its own literature, it is of great importance for ethnography, as the restless and trading race of the Malays is spread over an immense range of coasts and islands, and the history of this idiom for the greatest part is also that of the maritime population of the Eastern and Southern seas. A great scholar, the late Mr. W. De Humboldt, had seized on the solution of the problem which the origin of these people offers, and most thoroughly investigated it in his masterly work on the Kawi language;30 the last two volumes of which have appeared last year under the auspices of the Academy of Berlin by the care of Mr. Buschmann. He founds his researches on the Kawi, the antient language of Java, reconstructing its grammar by analysing the text of Brata Yuddha. Then proceeding to a similar analysis of the other Malayan dialects from the Philippines to Madagascar, he supplies the insufficiency of his resources by the strictness of his method, and by the astounding penetration of his mind. The grammatical investigation is enriched in all parts of the work by memoirs concerning the influence of India on the Malays, on the antiquities of Java, on the influence of writing on language, etc. memoirs which render this work a mine of new and important ideas, and where the penetration and the mental power of the author are equally displayed.

Mr. Buschmann advertises, that he intends to publish the text and translation of Brata Yuddha, which will form the complement to Mr. De H.'s work. It is an epic poem, an imitation of the Mahabharat, of which Raffles had already given a part in Latin characters. Written in Kawi, it dates as the Indian Poem to a period when the influence of Indian ideas in Java had not yet submitted to the Musulmans.

After having presented to you this sketch, unavoidably incomplete, of the progress that Oriental literature has made since our last meeting, I would desire to add a few words concerning a subject which has occupied, and is now occupying a great many learned men, and which deserves the whole attention of a Society, destined for the interests of Oriental literature. I allude to the variety of systems, at present

30. Uber die Kawisprache auf der Insel Java, von W. Humboldt. Berlin, 1836, 39, 3. vols in 4to.

a

adopted, to express the Oriental by Roman characters. At the first intercourse of Europe with the East in the middle ages, oriental words were rendered in a most barbarous manner; and thence arose the origin of a certain number of monstrous names, some of which have been retained in all languages of Europe, as Mahomet, Mosk, Tamerlan, Gengiskan. Since the last half of the seventeenth century, the Latin translations of some Arabian works by Pococke, Golius, and others, and little afterwards the popular works of Herbelot and Galland introduced a more exact orthography, by rendering the Arabian words as faithfully as the comparative deficiency of this alphabet permitted. A long time people were satisfied with this method of writing, but at last, and especially since the discovery of the Sanscrit had enlarged the circle of oriental studies, the want of a stricter method became apparent. A degree of exactness was aimed at to render again in the original characters, what had been previously expressed by the Roman alphabet; the systems, however, previously adopted, were unfit for this purpose, and whosoever attempted to reconstrue in Arabian characters verses, quoted by Herbelot, must have been convinced of this.

Since that period, systems rapidly succeeded each other. Founded on the most different principles which were calculated to avoid difficulties of several kinds, they have produced the most opposite results. Sir W. Jones so early as the year 1788, complained of almost every author having a system of orthography of his own. What would he have said of the number of systems, and the still greater number of orthographies without any system in the present day. Historians, geographers, travellers who never study the languages of nations, take at random the different orthographies and confound them, so that it is impossible to trace them to their sources, and hence ensues a mass of confusion. Of this I shall give some examples by taking the easiest familiar names I at present recollect. For instance, the name of Ali in works of our time is found thus: "Ali, Aly, Ali, Alee, Ulee, Ullee, Alli, Aliyy, Ahli, Alee." I find nine ways of expressing the word Koran : "Kuran, Ckooran, Alcoran, Alcorawn, Qoran, Coran, Koran, Ckoran;" six to write the name of Aboulfeda: "Aboulfada, Aboulfeda, Abulfeda, Abowlfida, Abowlfeda, and Aboulfidâi," and seven for the name of the legislator of the Arabs: "Mahomet, Mehemet, Muhammed, Mohammed, Muhammad, Mohhammad, and Muhummud."

In names so well known as those just quoted, there can hardly arise errors from these discrepancies in orthography; but in names of obscure men and places, the confusion arising from it, may be easily imagined. I shall give an example. Mr. J. Prinsep quotes an official and modern map of the Duab, where the road from Akbarpore to Cawnpore, a road very much frequented, is doubly entered, because the topographic bureau of Calcutta had found two routes with names, written in such different ways, that their identity not being recognised, they were believed to refer to parallel routes.31

It would perhaps have been better never to have deviated from the old system, however imperfect it was, as the thing of real importance is uniformity. But now it is too late to retrace our steps; the want of exactness having once been perceived, we must endeavour to supply it, hoping the introduction of a system, infinitely superior to the others, will re-establish that unity from which we are so far at present.

It is, meanwhile, I hope, not without use to classify the difficulties which such a system offers, and the attempts which have been made to remove them. These difficulties, it appears to me, are the following:

1. Oriental alphabets have a much greater number of letters than

ours.

2. Orientals do not always pronounce according to the rules of orthography.

3. They disagree in the pronunciation of the same letter in every country.

4. Europeans disagree in the pronunciation of the same letters.

1. Oriental alphabets have a much greater number of letters than ours. This especially has application to the Arabian and Indian alphabets. The means to obviate these difficulties, may be reduced to three classes.

a. The attempt has been made to enrich the Latin alphabet with some new characters. Thus has Meninski introduced the Arabian Ain; Volney modified the form of some Roman characters; Mr. Gilchrist invented a short u, and other learned men at a still later period used some Persian and Greek characters in their systems of rendering.

31. See the Map in "The application of the Roman Alphabet to all the Oriental Languages." Serampore, 1834, in 8vo.

None of these systems, however, were universally adopted, and the European public is not willing to tolerate the introduction of new characters into its alphabet.

B. It has been proposed to represent Arabic and Indian sounds by groups of European characters, as dh, th, kh, tt, ss, etc. This system has produced a great number of essays, but it has real inconveniencies; for if partially applied only, as the greater part of the learned do, the object which was in view with regard to it, is not attained; and if carried to the extreme, it renders strange the form of Oriental words, affording combinations of characters, which must appear barbarous to a European reader, as "Ckasr or Qasr, Hhadrat, Hadjdjadj," etc. Moreover, the system of expressing by double characters the simple ones which we do not possess, has the great drawback of leaving the reader in the dark concerning the orthography of the original, because he cannot know, whether the double character represent two characters, or be only the conventional representative of a single one.

C. Lastly, others have tried to modify the Latin alphabet by marks, not very apparent, which without producing new characters, exhibit various forms, by which the letters of Oriental alphabets may be easily exposed. This system, I think, was first proposed by Sir W. Jones, and adopted by the Asiatic Society of Calcutta, which, however, did not always adhere to it. According to it, the vowels are multiplied by accents, when they are short or long, and the consonants by points above or below. This system has had many imitators, and almost all Indian scholars have made similar ones for their rendering. Gilchrist has partly preserved it; the Geographical Society of London has adopted it with a few modifications; Mr. Erchhoff in France has made use of it in his parallel of the European languages; and lately, has Mr. Brockhaus proposed a similar one in Germany; Mr. Weijers has published another, resting on the same basis, and Mr. Arni, of Turin, has formed characters, on which he marks the different t, d, s, etc. of the Arabs by the same points, by which they are distinguished in the Arabic language. This method has the inconvenience easily to occasion errors, and to require a much larger printing apparatus, but it atones for these material difficulties by evident advantages. The European is not inconvenienced in his reading, for if he do not know the signification of the points added to the characters, he may easily overlook them, and

without their introducing an error, the reading of the words is not crammed with a mass of supplementary d'h and other characters; lastly, it approaches much nearer to that which only attempts to render the simple sound, without pretending to imitate all its shades, so that it is easy to identify words, written by a scholar, with those which a traveller, according to the mere pronunciation, has written down. The great mischief at present is the variety of systems, founded on this method; for we cannot expect, that the public shall become accustomed to this modification of the alphabet, unless the signs be generally adopted.

2. The Orientals do not always pronounce according to the rules of orthography, and this difference between the manner of writing and of pronouncing especially arises from euphonic laws. They, for instance, write al-Raschid, while pronouncing ar-Raschid. Mr. Weijers pro

poses to distinguish a character, subject to such a change, by putting it in italics; but this expedient displeases the eye, without indicating to the reader the real pronunciation. This problem is evidently indissoluble, and we have the choice between the sound and the orthography. The custom of the European nations with regard to this has established the excellent principle of submitting ourselves to orthography; thus is written in all European languages "Shakespeare, Bordeaux" etc., though the sound to be derived from this combination of letters, be much different from the real pronunciation. To follow the orthography is the only means not to efface the etymology of a word, and to preserve a chance of unity in renderings; yet there always remains a great confusion in the representation of short vowels, so differently pronounced in different words of the same language, that it becomes difficult to express them in all instances by the same vowel of our alphabet.

3. One and the same letter is differently pronounced by every Orien tal nation. The Turks, for example, generally substitute for the short A of the Arabs and Persians a short E; the Musulmans of India in many instances pronounce an E long, when the Persians pronounce a long I; in Persia a long A is substituted for a long Ou.* The Ara

• My readers will remember to give the sound of these vowels as in French.