verb, with which it agrees. On the whole, I guess,

that the diftich fhould thus be written;

Yecádu hhina tunájícum d'emáïrunà ̧
Yahdi álainà 'lásay lau là taássina.

When our bofoms impart their fecrets to you, an "guifh would almoft fix our doom, if we were not "mutually to confole ourfelves."

The principal verbs may have a future sense, and the laft word may admit of a different interpretation. Dr. HUNT, I remember, had found in GIGGEIUS the word dhemayer, which he conceived to be in the original. After all, the rhyme feems imperfect, and the measure irregular. Now I afk whether fuch perplexities could have arifen, if D'Herbelot, or his editor, had formed a regular fyftem of expreffing Arabick in Roman characters, and had apprized his readers of it in his introductory differtation?

If a further proof be required, that fuch a fyftem will be useful to the learned, and effential to the ftudent, let me remark, that a learner of Persian, who fhould read in our best histories the life of Sultan Azıм, and wifh to write his name in Arabick letters, might, exprefs it thirty-nine different ways, and be wrong at laft. The word fhould be written Aazem, with three points on the first confonant.

There are two general modes of exhibiting Asiatick words in our own letters: they are founded on principles nearly oppofite; but each of them has its advantages,

and

and each has been recommended by refpe&table authorities. The firft profeffes to regard chiefly the pronunciation of the words intended to be expreffed; and this method, as far as it can be pursued, is unquestionably ufeful: but new founds are very inadequately presented to a fense not formed to receive them; and the reader muft, in the end, be left to pronounce many letters and fyllables precarioufly; befides, that by this mode of orthography all grammatical analogy is deftroyed, fimple founds are reprefented by double characters, vowels of one denomination ftand for those of another and poffibly, with all our labour, we perpetuate a provincial or inelegant pronunciation. All thefe objections may be made to the ufual way of writing Kummerbund, in which neither the letters nor the true found of them, are preferved; while Kemerbend, or Cemerbend, as an ancient Briton would write it, clearly exhibits both the original characters, and the Persian pronunciation of them. To fet this point in a strong light, we need only suppose that the French had adopted a fyftem of letters wholly different from ours, and of which we had no types in our printing-houfes: let us conceive an Englishman, acquainted with their language, to be pleased with MALHERBE's well-known imitation of Horace, and defirous of quoting it in fome piece of criticifm: he would read it thus:

La mort a des rigueurs à nulle autre pareilles:

• On a beau la prier:

La cruelle qu'elle est se bouche les oreilles,

Et nous laiffe crier.

Le pauvre en fa cabane, ou le chaume le couvre,

• Eft fujet à fes loix,

Et la garde, qui veille aux barrieres du Louvre, 'N'en défend pas nos rois!"

Would

Would he then express these eight verses, in Roman characters, exactly as the French themfelves in fact exprefs them; or would he decorate his compofition with a paffage more refembling the dialect of favages, than that of a polifhed nation? His pronunciation, good or bad, would, perhaps, be thus represented :

Law more aw day reegyewrs aw nool otruh parellyuh, • Onne aw bo law preeay:

Law crooellyuh kellay fuh boofhuh lays orellyuh, Ay noo layfuh creeay.

Luh povre ong faw cawbawn oo luh chomuh luh 'coovruh,

The fecond fyftem of Asiatick orthography confifts in fcrupulously rendering letter for letter, without any particular care to preferve the pronunciation; and, as long as this mode proceeds by unvaried rules, it seems clearly entitled to preference.

For the first method of writing Persian words, the warmeft advocate, among my acquaintance, was the late Major DAVY, a member of our fociety, and a man of parts, whom the world loft prematurely, at a time when he was meditating a literary retirement, and hoping to pass the remainder of his life in domeftick happiness, and in the cultivation of his very useful talents. valued himself particularly on his pronunciation of the Persian language, and of his new way of exhibiting it in our characters, which he inftructed the learned and amiable editor of his Institutes of Timour, at Oxford,

He

to

to retain with minute attention throughout his work. Where he had acquired his refined articulation of the Persian I never was informed; but it is evident that he spells most proper names in a manner which a native of Persia, who could read our letters, would be unable to comprehend. For inftance; that the capital of Azarbaijan is now called Tabriz, I know from the mouth of a perfon born in that city, as well as from other Iranians; and that it was fo called fixteen hundred years ago, we all know from the Geography of Ptolemy; yet Major DAVY always wrote it Tubburaze, and infifted that it fhould thus be pronounced. Whether the natives of Semerkand or Samarkand, who probably speak the dialect of Soghd with a Turanian pronunciation, call their birthplace, as DAVY fpelled it, Summurkund, I have yet to learn; but I cannot believe it; and am convinced, that the former mode of writing. the word, expreffes both the letters, and the found of them, better than any other combination of characters. His method, therefore, has every defect; fince it renders neither the original elements of words, nor the founds reprefented by them in Persia, where alone we must seek for genuine Persian, as for French in France, and for Italian in Italy.

The fecond method has found two able fupporters in Mr. HALHED and Mr. WILKINS; to the firft of whom the publick is indebted for a perfpicuous and ample grammar of the Bengal language; and to the fecond for more advantages in Indian literature, than Europe or India can ever fufficiently acknowledge.

Mr. HALHED having juftly remarked, that the two greatest defects in the orthography of any language are the application of the fame letter to feveral dif'ferent founds, and of different letters to the fame found,' truly pronounces them both to be fo com'mon in English, that he was exceedingly embarraffed in the choice of letters to exprefs the found of the • Bengal

Bengal vowels, and was at laft by no means fatisfied with his own felection.' If any thing diffatisfies me, in his clear and accurate fyftem, it is the ufe of double letters for the long vowels, (which might, however, be juftified,) and the frequent intermixture of Italick with Roman letters in the fame word; which, both in writing and printing must be very inconvenient. Perhaps it may be added, that his diphthongs are not expressed analogously to the founds of which they are compofed.

The fyftem of Mr. WILKINS has been equally well confidered; and Mr. HALHED himself has, indeed, adopted it in his preface to the Compilation of Hindu Laws. It principally confifts of double letters, to fignify our third and fifth vowels; and of the common profodial marks, to afcertain their brevity or their length: but thofe marks are fo generally appropriated to books of profody, that they never fail to convey an idea of metre. Nor, if either profodial fign were adopted, would both be neceffary; fince the omiffion of a long mark would evidently denote the fhortness of the unmarked vowel, or converfely. On the whole, I cannot but approve this notation for Sanscrit words, yet require fomething more univerfally expreffive of Asiatich letters. As it is perfect, however, in its kind, and will appear in the works of its learned inventor, I fhall annex, among the examples, four diftichs from the Bhagawat, expreffed both in his method and mine*. A tranflation of them will be produced on another occafion. But, in order to render this tract as complete as poffible, a fuller fpecimen of Sanscrit will be fubjoined with the original, printed in the characters of Bengal, into which the Brahmans of that province. tranfpofe all their books, few of them being able to read the Dévanagari letters; fo far has their indolence prevailed over their piety!

* Plate IV.

Let