Of course our scheme applies only to Indian languages of the Aryan family and, no doubt, even if we restrict ourselves to this limited area, we shall find differences of detail which commend themselves to scholars of various nationalities.

I notice, however, that Professor Oldenberg, in his latest work on the Veda, uses almost every one of the same equivalents for the Nagari letters as those of our scheme. He agrees with us in rejecting the French ç, merely differing from us by his preference for instead of s to represent the palatal sibilant.

Many other German Orientalists, too, now follow Lepsius' later ideas, employing c for the English sound of ch in 'cheer'; and j for that of j in 'jeer,' and y for that of y in 'your,' all of which sounds are common in Indian languages.

Still we find among English, German, and French scholars a residuum of differences which might, I think, be discussed in a spirit of mutual harmony and concession, with a view to the eventual adoption of complete international uniformity not only in Aryan, but in Semitic languages.

Whether such an eventuality is possible in Turanian languages with ideographic systems-especially ChineseI leave our present President (Sir Thomas Wade) to inform us; merely observing that the existence of about 43,000 distinct ideographic symbols in Chinese (according to Prof. Legge and the Chinese Ambassador whom I once met at the Professor's house 1) makes the uniform transliteration of Chinese-if possible at all—a question which seems likely to glow with incandescent heat, should our missionaries ever take up the matter seriously in its bearing on the spread of Christian education. As to Japan, I understand that a Japanese New Testament transliterated into Latin or Roman characters is already in existence.

I need scarcely add that my present paper leaves wholly untouched the difficult question of deciding upon the best alphabet for unwritten languages such as the African.

The following are a few of the points which it seems to me

1 This gentleman, however, confessed that he himself only knew about 8000, and that few of the most highly educated knew more.

desirable to discuss and in respect of which some definite conclusions ought to be arrived at-assuming of course that the desirability of employing c, j and y as equivalents for the Indian palatal letters is already generally admitted.

1. Whether the saddle-back-shaped circumflex, or the horizontal mark, or the acute accentual mark should be employed for long vowels?

2. Whether the Sanskrit vowel sound ri should be expressed by r (as some German scholars hold) or by ri?

3. Whether c alone or c with some diacritical mark should stand for the Nagari and our sound of ch in church?

4. Whether the guttural nasal should be represented by ǹ ?

5. Whether s, or 8, or ș, or the French ç should stand for the Nagari palatal sibilant? (Note that the use of ç is common in Germany and America, and that Professor Cowell of Cambridge, Professor Weber of Berlin and Professor Whitney of America support it.)

6. Whether a simple symbol, such as rather than sh, should stand for the cerebal sibilant?

x

7. Whether a should be adopted (as it has been by Lassen, Dr. J. Muir and others) for the common compound ksh?

8. Whether some formal exception could not be taken to the Bengali system of transliterating the inherent short ǎ by the symbol o (e.g. Brahmo-Somaj for Brahma-Samāj), and to the Burmese system of transliterating s by th, etc.— practices which result from mixing up the question of transliteration with that of pronunciation.

9. Whether any agreement can be brought about in transliterating Semitic languages ? (Note that evidently much disagreement still prevails in regard to certain Arabic symbols, as may be seen by comparing Professor Nöldeke's method with that of our scholars.)

10. Whether, at least, some authoritative transliteration of common names, both Aryan and Semitic, could not be drawn up and recommended to be adopted by those who write books on Oriental subjects for general readers? In regard to this point we know that the Government of India has sanctioned

the generally scientific system of transliteration adopted by Sir W. W. Hunter in the Gazetteers edited by him; and it becomes a question whether scholars are to uphold the Government decision.

11. Whether in certain of the commonest names (especially of places), scientific accuracy should not be sacrificed to timehonoured usage (as in the name 'Benares,' etc.).

Probably Sir W. W. Hunter will testify to the importance of this last question. It concerns also the spelling of names of persons (as, for example, all the names in our Bible beginning with J, such as 'Job,' etc.).

As to the names of persons occurring in Indian languages, I merely instance two common names:-1. that of Buddha (the first syllable of which is constantly mispronounced, like our English word 'bud,' through the prevalence of the false idea that transliteration is always to be a guide to pronunciation, while this name is still spelt Bouddha by French writers); 2. that of the Arabian prophet (Muhammad or Mohammed or Mohammed or commonly Mahomed, Mahomet), and names connected with him (Ahmad or Ahmed? Muslim or Moslem? Kuran or Quran or commonly Koran? etc.).

To those who doubt whether chaos does not still reign. supreme in the transliteration of common names, I commend a study of the lists occurring in the Calcutta University Calendar for 1890, extending for nearly 300 closely printed pages.

I conclude with a proposal that our Royal Asiatic Society should communicate (through its Council) with other Oriental Societies in other parts of the world, as well as with the Editorial Committees of the religious societies mentioned before (pp. 625, 626), and recommend that two Delegates should be sent from each Society to the next Oriental Congress or Congresses, charged with the duty of conferring together on the possibility of formulating a uniform international scheme of transliteration.

[The above paper was read on April 21, 1890.]

ART. XII.-On a Proposed Method of Transliterating the Languages written in the Arabic Character. By H. T. LYON, M.R.A.S.

[Read March 17, 1890.]

IN bringing forward a scheme for the conversion of the Arabic character to the Latin-the second most widely spread alphabet of the world to the first-a word may be said as to the objections which the mere mention of such a proposal invariably evolves.

First comes that of the minor savant who, having acquired by years of laborious practice the facility of deciphering through his knowledge of word forms the sketchy hieroglyphics representing the Arabic letters, resists any attempt to provide the learner with an easier path to knowledge than that so painfully pursued by himself.

In reply to the arguments advanced by this class of scholar, I have no hesitation in declaring that the student who devotes himself to acquiring a certain vocabulary through the means of a good system of Latin transliteration will find himself months-aye, and years-in advance of his comrade who has fettered himself from the outset by attempting to cope with the mechanical difficulties of the character, which, by reason of its unfamiliarity, affords him little or no aid in recalling the visual form of words, so long as he is still ignorant of the language.

Another objector is the individual who bases his arguments on the truly Oriental ground that it is impossible to obtain the agreement of all scholars on the point, and that therefore it is needless to discuss the matter. But I do not think that he requires any more serious reply than does his more sympathetic though irrelevant colleague, who tells you that he would gladly see some universal scheme adopted for, say,