of our most celebrated orthoepists make a decided distinction between the sounds of e and u in a syllable ending with r,* it might be deemed improper to represent the pronunciation of Berks by burks. On the other hand, were we to follow the method of Walker in similar cases, and pronounce it bêrks, we should be still wider from the mark. The mode adopted by us will, it is hoped, be found sufficiently definite, at the same time that it obviates both of the difficulties just mentioned.

IX. With a view to simplify as much as possible, we have rejected Walker's second and third sounds of o (the former being equivalent to oo, the latter to au), and his third sound of u (corresponding with oo in good, a sound which we have represented by ŏŏf). From the same motive we have dispensed with the figured vowels, whenever their use has not appeared to be necessary in order to avoid ambiguity. Thus we write simply TIL'-SIT, and not-TIISIT, til'-sit,-as Walker would have done. All marks or figures which are not needed, in order to indicate the exact pronunciation, must tend rather to embarrass than to aid the learner.

X. In the pronunciation of names belonging to England, or to those countries where the English language is spoken, a ending a syllable with the accent, should always have its first sound; in an accented syllable before a single r it usually takes the second, and before rr, or any other consonant, the fourth sound. Accordingly, in giving such names as PENNSYLVANIA, SA'RUM, FAR'MINGTON, CAR'ROLL, MAN'CHESTER, it has been deemed unnecessary to mark the A in the accented syllables, as scarcely any one in the least acquainted with the principles of English pronunciation, could mistake its true sound. In all cases which depart from these simple and general rules, it has been thought best to mark the sound, as Dål'ton, Mål'wah, Pâr'is, &c.

XI. The Latin names of foreign countries are nearly always to be pronounced with the English sounds of the letters. Thus, in BAVARIA, BULGARIA, LUSA'TIA, and TRANSYLVANIA, the accented a should have its first sound.

*Walker says that "Derby is pronounced NEARLY as if written Durby"and that "fir, a tree, is perfectly similar [in sound] to the first syllable in ferment, though often CORRUPTLY pronounced like fur, a skin." (See Principles of Pronunciation, 100 and 109.)

+ We have, however, retained his third sound of a, (though precisely similar to au), in order to indicate the broad sound of the a in such names as DALTON, CALDER, &c., without writing the pronunciation separately.

OBS. In PRUSSIA and RUSSIA, however, the u instead of being pronounced short, as it would unquestionably have been done in ancient Latin names of this kind, is usually sounded like oo, assuming in these instances the character of the German or Russian u. In like manner the first syllable of BULGARIA is to be pronounced bōōl and not bŭl.

XII. In the anglicized forms of foreign names, and in most well known names of foreign countries, the same rules of pronunciation, generally speaking, obtain, as in genuine English words, e. g., SPAIN, ITALY, NAPLES, &c.

OBS. 1. At the same time, we may observe a general tendency to adopt those sounds of the English vowels, which approach most nearly to the foreign sounds: thus the a in the first syllable of ADRIATIC, and Paris, has its fourth sound, which is much nearer to the French and Italian a, than its first sound, though this would probably be given to these names by the mere English scholar, who should be guided by analogy solely. In like manner the i in MILAN is made short, so as nearly to correspond to the Italian i, which is like our e.

OBS. 2. In the pronunciation of foreign names that have become thoroughly anglicized, it is interesting to observe the tendency of our language, to throw the accent as far as possible from the terinination. Thus PARIS is pronounced with the accent on the first syllable, though the French appear to place it on the last, and the Germans, who have not the same aversion to the ultimate accent, that we have, lay the stress of voice distinctly on the final syllable, thus PAR-IS'. HANOVER, which might be pronounced with the native accentuation (HANOVER,) without the slightest offence to the genius of our tongue,* has become irrecoverably HANOVER. Thus, also, we pronounce ANDALUSIA, (in Spanish ANDALUCIA,) AR'AGON (in Spanish ARAGON), &c. So in our own country, the old NIAGARA has become unalterably fixed as NIAG'ARA; and HURON', though still sometimes heard, is fast giving place to Hu'RON.

XIII. It may be observed that with respect to foreign names, not only in the French, Italian, and other languages that are written in the Roman letters, but also in Germant and Greek, (the characters of which may be readily converted into corresponding Roman letters), it is generally customary in English to retain the literal spelling, e. g., ANSPACH, (German Anspach), KÖNIGSBERG, (German Königsberg), CHIOS, (Greek Xios,) &c., excepting a very few well known names, as LYONS, (French Lyon), NAPLES, (Italian Napoli,) MUNICH, (German München), DANTZIC, (German Danzia,) LEIPSIC, (German Leipgig);—we often find, however, the last two names spelled literall Danzig and Leipzig.

• We have a multitude of words similar in accent, as promoter, devotion + It is, perhaps, scarcely necessary to remark, that Roman letters are n quently employed in writing and printing German.

XIV. On the other hand, names in languages of which the charac ters cannot be readily converted into Roman letters, or which are but little known as written languages, are usually spelled according to their sound in some well known European tongue. Thus the name of one of the cities of Persia is written in English, Shooster or Shuster, in German Schuster, and in French Chouster, precisely the same sound being expressed by these different spellings.

We find in English works of the highest character, these various modes of writing oriental and other names, employed indiscriminately. Thus in McCulloch's Geographical Dictionary, under the article SHUSTER, we find within the space of eight lines Khuzistan, (Khoozistan), Karoon and Dezphoul, (DEZFOOL); the first name being, as regards the sound of the vowels, German or Italian, the second English, and the third French. On the map of Persia, published by the Society for the Diffusion of Useful Knowledge, names written according to these three different modes, occur promiscuously in almost every part. Even the same name is frequently spelled differently in different parts of the same work. Thus on the map just mentioned OORFA is written "Orfa or Ourfa," while on another of the same set it is spelled Urfa. In the Penny Cyclopædia, we find Boossa and Boussa, Soodan and Sudan, Toorkistan and Turkistan, &c. McCulloch gives SHUMLA (Shoomla) under its proper head, but mentioning the town in another place writes it Schumla: in the same article, AFRICA, he has both Soodan and Soudan, each occurring several times: under MOGADORE he gives Shwera (more properly Sweera) as the Moorish name of this town; afterwards, in enumerating the principal fortified and garrison towns of Morocco, he gives Suira (pronounced Sweera), without so much as mentioning the name of MOGADORE. As neither Schumla nor Suira are to be found in his gazetteer, under their respective heads, it is scarcely possible that any one unacquainted with the different modes of writing these names, should know what places are meant by them, or in what part of his work to look for information respecting them.

XV. In order to avoid the perplexity and confusion resulting from the diversity which prevails in the mode of writing oriental names and others of the same class, we have made it a point always to spell them after the English manner, except in a few instances where a different spelling appears to have become thoroughly established by usage. Accordingly we have given OORFA, SOODAN, TOORKISTAN, &c., as

, the preferable mode of writing such names; at the same time, under the heads of OURFA and URFA, SOUDAN and SUDAN, and TURKISTAN, the reader will find a reference to the names as spelled in the English manner, to which he must look for a description of those places. By adopting this plan, it is believed that the correct pronunciation of oriental names will be taught in the simplest and easiest manner; the perplexity and error into which the prevailing inconsistent mode of writing such names, has sometimes led even well-informed geographers, will be avoided; and the apparent contradictions which are so often met with in our most popular geographical works, will be accounted for and reconciled.

The reader may be referred to the case of Schirvan and Shirvan, already men. tioned in the Preface. There can be little doubt, that the practice which prevails among the English, of writing oriental names after the manner of other European nations, has sometimes led geographers of the highest character into error. Thus Mantchoo, the name of a tribe of Tartars, inhabiting the north part of the Chinese empire, is written by some of the most respectable authorities, Munchow. It is pro bable, that in the first place some English writer or writers, spelled it Mantchou after the French manner, and that others supposing it to be English, and wishing to adopt a mode of spelling less equivocal, wrote it Manchow. It may be, however, that the latter indicates the true pronunciation, and that the name was originally written correctly Mantchou, the ou having its genuine English sound; and that some English writer, naturally supposing it to be Freneh, (for many, if not most of the English, appear to prefer the French mode of spelling such names, to their own,) without investigating the subject, converted the name, as they thought, into English, by writing it Mantchoo. As Mantchooria (the country of the Mantchoos) is rarely visited by Europeans, this question may long remain undecided. At present, Mantchoo appears to have become almost universal. In one of the earlier numbers of the Penny Cyclopædia, (article CHINA.) we find it written Manchow, but in other parts of the work Mandshoo, which does not differ materially in sound from Mantchoo. Had the English uniformly adopted the practice of writing oriental names according to the sounds of their own language, it is probable that CABUL would never have lost its native sound (Kâ-b')), so far as to be generally pronounced by the English and French, Cabool or Caboul, (see CABOOL, in the body of this work.) It is true that it was formerly written correctly in English works Cabul or Caubul; but the practice of spelling oriental names according to the German or Italian mode, is so common among English writers, that analogy would naturally lead us to adopt or confirm that pronunciation of Cabul, which appears to be now so thoroughly established. Had it been the uniform practice of English travellers and geographers to write such names as Surmul, Dezful, (01 Dezphoul), according to the sounds of their own tongue, viz., Soormool and Dezfool, no English reader would have thought of pronouncing the u in Cabul like oo. The erroneous pronunciation of a single name may, perhaps, justly be regarded as of little moment, but the embarrassment and error into which the prevailing practice of English writers on geography, is continually liable to lead both the ignorant and the learned, appear to us to involve considerations of the highest importance to the interests of geographical science. From a sincere wish to add as much as possible to the utility of the present work, we have had recourse to the system of writing oriental names, which has already been explained. It may be remarked that though this system is frequently at variance with common usage, it is sanc tioned by the example of the most distinguished French geographers, as well as of several English writers of the highest character.

XVI. In giving the pronunciation of this class of geographical names, to represent the accent correctly, is the principal difficulty to be encountered. Those acquainted with French, are aware that this language has no accent in the sense in which we employ the term. The same may be said of the Hungarian, and perhaps also of the Arabic, Turkish, and Persian. At all events, those best acquainted with these languages, are not unfrequently at a loss, when they wish to represent accurately in English, the accentuation of Arabic, Turkish, and Persian words. Nevertheless it will be found that here, as in the French and Hungarian, there is something analogous to our accent, which will generally serve to guide us in marking in English, the accentuation of names in these languages. The different syllables of oriental names, however, like those of French words, are not unfrequently pronounced with a stress of voice so nearly equal, that it has been deemed proper in a number of instances, to use the secondary accents in order to indicate more precisely the true pronunciation. (See Remarks on the

French accent, Section XIX.)

OBS. 1. It may not be improper here to remind the reader, that the accents which we often see upon oriental names, are by no means to be understood as always indicating the manner in which an Englishman should accentuate these names in pronunciation, as they are often employed to denote some particular sound in the vowels over which they are placed. Thus some authors place an accent upon a, when they wish merely to signify that this letter has the clear full sound of a in far. In the same manner an accent is placed upon u, in order to show that it has the Italian or German sound, or in other words, is to be pronounced like oo.

OBS. 2. The sound of a in several of the oriental languages is often very broad, approaching nearly to that of au, in English. Hence we often see AFGHAUN instead of AFGHAN, CAUVERY instead of CAVERY, sultaun instead of sultan, &c. In writing NEPAUL and BHOPAUL, the improper diphthong au appears to be almost universally employed by the English. The French indicate the same sound, by using a with a circumflex, e. g., NEPAL, BHOPAL, &c. This method is, perhaps, preferable to ours, the sound of â being intermediate between that in the English word far and that in fall.

OBS. 3. The Arabic article al or el, is often changed in pronunciation, so as to correspond with the initial consonant of the word to which it is prefixed, thus, EL-RASHEED (Rashid), EL-SHAM, (the Arabic name of Syria), and EL-SIOOT, are pronounced and should be written in English, Er-Rasheed, Esh-Sham, Es-Sioot. The vowel sound of the article also varies considerably, sometimes approaching that of oo, thus the "Country of Dates," is usually pronounced Beled ́ool Jer-eed'. So Es-S100T (or ŠI0OT, without the article,) is sometimes written As syout and Osioot.