i Orthography of foreign Place-Mamas, Jas. Burgess 1892. •TOLIE«¥M¥IEiaSIIW- iuiiBiKirar «tK..!KCT^)^Wi^VnW4&^^uv-^k^ Reprinted from The Scottish Geographical Magazine for January 1892. THE ORTHOGRAPHY OF FOREIGN PLACE-NAMES. By Jas. Burgess, LL.D., C.I.E., Etc. The Report of the United States Hydrographic Office Board on the Orthography of Foreign Geographic Names, recently issued, is a docu ment deserving the attention of all geographers. In the introduction, the Hydrographer remarks that "the lack of a uniform and compre hensive system of spelling geographic names has long been felt, and many attempts have been made to overcome the difficulties operating to prevent the adoption of some one method that might in a measure settle the vexed question. To an office engaged in the publication of charts and sailing directions covering the waters of the globe, the want of systematic method becomes especially formidable, inasmuch as the publications emanating from it are regarded as having the sanction of official authority ; and, without some uniform standard system to proceed upon, doubt and confusion must necessarily be the result of inconsistencies 2 THE ORTHOGRAPHY OF FOREIGN PLACE-NAMES. and discrepancies, sometimes so great as to cause the same place to become unrecognisable under its various disguises in the way of spelling. Foreign Governments have long recognised this difficulty, and the Hydro- graphic Offices of Great Britain, France, and Germany have carefully considered the subject, each taking the initiative in their respective countries." In 1890 a Board of Foreign Geographic Names was officially created by the United States Government, to deal with the question, and its decisions were ordered to be accepted by the Departments of War, Navy, and Post Office, and by the Coast and Geodetic, and the Geo logical, Surveys. This Board communicated with the Hydrographic Offices of other countries ; and it appeared that in France a commission appointed by the Societe de G6ographie, on which the French Hydro- grapher, M. B. de la Grye, acted as secretary, had reported in 1886 in favour of a system largely formulated upon the British, and on proposals made by the French Hydrographic service. The German Hydrographic Office also adopted rules in 1888, closely following the earlier British system. In examining these systems, the Board considers that they divide all geographical names into two classes, on the following propositions : — " 1. Geographic names originating in languages that are written in Roman character should be spelled as in the country having jurisdiction. The French and German systems allow exceptions to this rule in the case of names which, through long custom, have become generally known in another form:" as (French) La Mecque, Naples, Calcutta: (German) Kopenhagen, Neapel, Genua, Zanzibar, Zulu. " 2. Geographic names originating in languages without literature, and in languages that require transliteration into Roman characters, should have their spelling based on the true sound of the word as locally pronounced." It is not apparent why here and in its " rules " given below, the Board has so entirely separated, in this second class, all languages with alphabets, however perfect for the representation of their sounds, from those using the purely Latin alphabet, and has, moreover, grouped them along with those devoid of any form of writing. Surely it would prevent some confusion of thought to make at least three classes, of which the second should consist of those geographic names originating in lan guages using alphabets like the Arabic, Russian, and the various deriva tives of Sanskrit, which are susceptible of scientific transliteration. Languages like Chinese — which is not alphabetical, and those without any form of writing, might either be classed together or separately. It is a recognised fact that the alphabets referred to are essentially phonetic, and represent the sounds in their own languages with far more precision than the Roman alphabet does those of the tongues of "Western Europe and America. Further, under conditions as little inconvenient as those implied for the first class, a well-advised system of transliteration would present in Roman letters the sounds of names in these foreign lan guages much more consistently than any phonetic system — which must ever be dependent on the ear and intelligence of the individual alone. THE ORTHOGRAPHY OF FOREIGN PLACE-NAMES. In regard to the values of the letters to be used in transliterating names under this second class, the three European systems are in such close agreement as to be practically almost identical. This has resulted from the practical adoption by France and Germany of the British system, which accepts the Continental sounds of the vowels. The following com parative table illustrates this, and gives also the American Board scheme of vocalisation — where names are to be represented phonetically — and of transliteration, where that is to be accepted : — United States. a as in father e as in benefit i as in ravine o as in mote u as in flute o nearly as German 6 or French m ii like German and French u ai as i in ice au as ow in how ao as in Nanao ei nearly as in they b English b c always soft as in center ch as in church d English d f English/ g always hard always pronounced when inserted English j English k, also replaces hard c Oriental guttural, as in Khan Oriental guttural, as in Ghdzi English I English m English n as in finger and singer English p as in loop-hole, not to be used for/ q not employed kw replaces qu (as in quiver) English r English s as in shelf English t English v English w x English x y always consonantal, as in yard z English % British. French. a a e e 1 i 0 0 u u oe u ai au ao ei b b c c ch tch or ch d d f f g g h h j k d orj k kh kh gh 1 gh 1 m m n n ng n P P ph kw q1qu r r s s sh sh t t V v w w X X y y z z German. a, a eiou a or sa o i or ce ii or ue ai auao ei, as i in mine bc chd f Jk kh gh 1 mn k, also ku r sshtV w Xy, also j z Spanish. bc, also z ch hard before a, o, u, ue, and ui y, consonantal k, also qu .1 j 1 rnniiP qu rssh t v w xy z 1 For the Arabic q&f in Qoran, Talaqan, Bilqa, Reqem, Qe'lb, etc. same as the Hebrew p. This is the 4 THE ORTHOGRAPHY OF FOREIGN PLACE-NAMES. A study of this table, says the Report, suggests " the possibility of future international agreement as to the phonetic value to be attributed to each letter in spelling geographic names requiring treatment from a phonetic basis. Practically but few difficulties now exist, and these might be eradicated by an international congress." But in the way of the practical application of these systems, the difficulties, as the report adds, " are various and many, and not the least is the uncertainty existing in regard to the correct local pronunciation of many of the names in distant foreign lands that use another than the Roman character in their writing, and many years will pass before these difficulties will have dis appeared. Where the local pronunciation of the name is absolutely known, it may be readily spelled under the general system ; but where the written form only is available, detailed systems in harmony with the general system can be advantageously used for transliterating, and thus assist in bringing about the ultimate object." With reference to this last statement, the Report also furnishes detailed systems for the transliteration of Greek, Russian, Japanese, and Korean names, and supplies some data towards formulating guidance for Malay and Chinese. One is rather surprised, in presence of all this, that no reference is made to Turkish, Persian, and the other languages using the Arabic alphabet. The Board submits the following rules for the orthography of place- names : — • " {a) The more prominent geographic names, the orthography of which is already fixed by universal adherence to a certain form, or which are spelled in but one way in American and English literature, shall not be changed, although their spelling may not be in accordance with rules (b) and (c). Examples: Athens, Bombay, Ceylon, Germany, Spain, St. Petersburg. " (b) Geographic names in countries which use the Roman or Gothic characters in writing their language shall be rendered in the form found in the latest official publications, maps, and books of the country having jurisdiction. In cases, however, where the form of a place-name hitherto used is so different from the official form that the identity of the latter with the former might not be recognised, the old form shall be added in brackets ; as, for instance, Livorno (Leghorn), s'Gravenhage (The Hague), Wien (Vienna). " Where there is a conflict of jurisdiction, as it were, over a geographic name, as, for instance, in the name of a body of water separating two or more countries, or as in Polynesian or African names, rule (a) or (c) shall be applied. Thus, Swedish and Norwegian ' Skagerak,' Danish ' Skagerrak,' is an example where choice lies only in the spelling, and under rule (a) preference is given to double r, 'Skagerrak'; but in the Baltic (Baltiskoe More, Baltisches Meer, Ostsee, Ostersjon, etc.) the choice lies first in the proper name, and Baltic is retained under rule (a); while the Dutch, Spanish, or other rendering of Polynesian names must be treated uridei rule (c) if not in accord with the general system. "(c) Geographic names from languages that require transliteration into Roman character, and geographic names originating in languages THE ORTHOGRAPHY OF FOREIGN PLACE-NAMES. without literature shall have their spelling based on the true 'sound of the word as locally pronounced, and in accordance with the general system of sound equivalents given below : — Letters. Pronunciation and Remarks. Examples. a ah, a as in father. Java, Banana, Somali, Bari. e eh, e as in benefit. Tel-el-Kebfr, Oleleh, Yezo, Medina, Le- viika, Peru. i English i, as in ravine ; the sound of ee in beet. Thus not Feejee, but Fiji, Hindi. 0 o as in mote. Tokyo. u Long it as in flute ; the sound of oo in boot. Thus not Zooloo, but Zulu, Sumatra. 0 German o (oi). Sochhdn. ii The French sound of u (u) ; or the German ii (ue). All vowels are shortened in sound by doubling Bdrkiit. the following consonant. Yarra, Tanna, Mecca, Jidda, Bonny. Doubling of a vowel is only necessary where there is a distinct repetition of the single sound. Nuulua, Oosima. ai English i as in ice. Shanghai. au (ue as in how. Thus, not Foochow, but Fuchau. ao is slightly different from above. Macao. ei is the sound of the two Italian vowels, but is fre quently slurred over, when it is scarcely to be distinguished from ey in the English they. Beirut, Beilul. b English b. c is always soft, but is so nearly the sound of s that it should be seldom used. Celebes. If Celebes were not already recognised, it would be written Selebes. ch is always soft, as in church. Chingchin. d English d. f English/; ph should not be used for the sound of/. Thus, not Haiphong, but Haifong, Nafa. g is always hard (soft g is given by j). Galapagos. h is always pronounced when inserted. j English j. Dj should never be put for this sound. Japan, Jinchuen. k English h. It should always be put for the hard c. Thus, not Corea, but Korea. kh the Oriental guttural. Khan. gh is another guttural, as in the Turkish Dagh, Ghazi. as in English. n ; ng has two separate sounds, the one hard as in the English -word finger, the other as in singer. As these two sounds are rarely employed in the same locality, no attempt is made to distinguish between them. P ph as in English. as in loophole. Chemulpho, Mokpho. q should never be employed ; qu is given as ho. Kwangtung. .; ! as in English. THE ORTHOGRAPHY OF FOREIGN PLACE-NAMES. Letters. Pronunciation and Remarks. Examples. H as in English. Sawakin. x ; y is always a consonant, as in yard, and therefore should never be used as a terminal, i or e being substituted. Kikiiyu. Thus, not Mihinddny, but Mikindani. not Kwaly, but Kwale. z English si. Accents should not generally be used ; but where there is a very decided emphatic syllable or stress, which affects the sound of the word, it Zulu. should be marked by an acute accent. Tongatabu, Galapagos, Palawan, Sarawak." This alphabet, it will be seen by comparison with the previous table, is almost the same as that adopted by the British Hydrographic Office, with the addition of the vowels 6 and U, with diacritical marks, and is almost as satisfactory as could well be expected. Both provide kh and gh for two of the Arabic gutturals, but reject q for the third. The French, however, have wisely retained the q as a good representative of the Turkish, Arabic, and Persian guttural qdf (jj), familiar to us in the word Qoran. It has a distinctly different sound from the letter kdf (t-» <"), which corresponds to our k; and since q is not required for any other sound, it is well to adopt it for this letter, as it is already largely in use among scholars in transliterating quotations from the Arabic alphabet. It is required to preserve the sound in names like Quraish, Qairwan, Qulzum (Clysma), Qatif, Qazwin, Qapchaq, Aqlid, Qablah, Dauraq, Qansarin, Qus, Qarqisia, Qarqub, etc.1 The British system has also recently added ' the digraphs aw (as in law), hw, and zh? This British Admiralty system, adopted in 1885 by the Royal Geographical Society and by the War Office, has also been accepted by the Colonial Office, which has disseminated circulars to the Governments of the various colonies. There is thus no backwardness on the part of our Government in recommending a system by which some principle in writiug geographical names might be introduced in our maps and books. The difficulty is to get the general public, or even scientific societies and official authorities, to adopt and consistently adhere to this or any inter national system. The American system- does not essentially vary from that of our own Admiralty, except in some of the detailed applications to different alphabets, and in inclining to spell many names rather by the ear than according to their vernacular written forms. Both lay too much stress on sound, to the neglect of vernacular spellings. i We may not perhaps be quite prepared yet to introduce it into such familiar names as Asqalon, Qaisariah, Samarqand, Qilniaq (Calmuck), Demeshq (Damascus), or Ladiqiah ; but this may possibly yet come about ; less familiar names might well be used in their correct form at once. 2 As iu Hwang-lio and Ngan-hwi; and in Muzhdaha. THE ORTHOGRAPHY OF FOREIGN PLACI0-NA11ES. 7 In France a ministerial decision, dated 21st June 1887, decided to use the rules proposed by the Soci6te de Geographie of Paris in April 1886 for the spelling of native place-names.1 These rules are remarkably international, and discard distinctively French representations of sound like ou, u, eu, ch, dj. In countries using the Roman character it is ruled, in accordance with the British system, that geographic names shall be written in the orthography of the country to which they belong. In others the following alphabet is adopted for trans! iterative purposes, as well as for the phonetic spelling of names in countries having no proper alphabet : — a, e, i, o to be sounded as in Italian, French, Spanish, and German. e shall never be mute. ii with a trdma like German ii, to represent the French Sound of u. u as in Italian, etc., to represent the French sound of ou. ce as in oeil, for the French sound of eu. consonants mostly as in French, etc., but — g and s always hard as in gamelle, sirop. sh to represent the French sound of ch, Sherif kh and gh to represent the strong and soft Arabic gutturals. th as in English path (Greek 6). dh as th in English those (Greek 8). y to represent the semi-vowel, as in yole. w as in English William. h always aspirated, or modifying the preceding letter. S n with a tilde, to sound as gn in seigneur. q used to represent the Arabic qdf (as in Qordn). Arabic ain may be represented by a double dot. A The circovflexe indicates the lengthening of a vowel. The German system, adopted in 1888, a*lso follows closely the British system.2 It adds, however, the German symbols a, 6, ii, with their German sounds, and the Skandinavian a ; uses j for the English sound of j ; ch andyft respectively for tsch and sch ; and permits the doubling of a consonant to render the preceding vowel clear. This last is an expedient that is liable to misuse, however, and should be employed with care and not at random. In Spain, though a system has been under consideration, it does not appear to have quite come into operation. Its principal features are given above in the general table. It is thus clear that there is not only a fair consensus of opinion in favour of a uniform and international system of spelling geographical names, but a practical agreement to discard national peculiarities in favour of one phonetic Roman alphabet. The differences on this point are almost confined to the addition or rejection of diacritical marks and accents on a few letters. These mai'ks are certainly very helpful to the correct pronunciation of foreign names, and German cartographers, at least, seem to have no difficulty on the i Bulletin Soc. Geog. 1886, pp. 193-202. 2 Published in the Annalen der llyilrographie, 1S88, p. 543; and NucJirichten fur Seefahrer, 18S8, p. 824. 8 THE ORTHOGRAPHY OF FOREIGN PLACE-NAMES. score of using them on their maps, and appear to adhere most creditably to consistent transliteration of names from languages with other alphabets. The objection to the use of diacritical marks seems to be often pushed to an extreme. What is their use 1 Let any one open a good pronouncing English dictionary to see their necessity even for English words. They indicate varieties of sounds expressed in type by the same symbol. We learn the particular sound, and then dispense with the diacritical marks in our books, except in a few cases. When once names have become familiar, there is much less need for such phonetic helps, and they might be given only in the gazetteer, the "index geographicus," and school geography. Crecy, Chalons, Nimes, Beam, Bareges, Mezieres, and hundreds more, we should pronounce the same with or without the accent ; yet, for sufficient reasons, it is well to mark it. In other languages they play a larger part than in English. And the American Board has found it desirable to introduce two marked vowels that had not been employed by the British Hydrographic Office. There are, moreover, in the Slovenik, Sanskritic, and Arabic alpha bets, letters to express modifications of the sounds of s, e, h, n, t, and d, for which the Roman alphabet has no equivalents, and if we are to indicate these additional sounds, we can only do so by diacritical marks. This is inconvenient in printing ; and so it would be to print English books indicating by marks the seventeen different sounds or thereby which we represent by the five vowel letters, and the ten others we represent by the four symbols, c, g, ch, and th. But as we do with ordinary words, we can do with place-names. In the pronouncing vocabulary or index we may represent them with diacritical marks on letters having modified sounds : on maps and in ordinary literature, we may spell as in the index, but without these marks — unless, as may be desirable in some cases, we. introduce italic letters for those that are modified. Let only a rational system of transliteration be established, with diacritical marks for the index and the use of the learner, and the way of progress in this matter will be much cleared. But it need not be altogether overlooked that for maps, whether lithographed or engraved, diacritical marks can be inserted without the sort of inconvenience they present to the com positor. Captain W. J. L. Wharton, the Admiralty Hydrographer, remarks on the practical outcome of the system, that " to revise the names at present existing is a herculean task, and one that must occupy many years ; for our system being necessarily on a phonetic foundation, the proper sound of a name is required before it can be reduced to writing. This is, I find, the most difficult item to deal with, and I am obliged mainly to content myself with reducing to system such names as are- manifestly spelt in contradiction to it. This is done whenever a book of sailing directions is revised, as it is desirable that charts and books should go together, and on these occasions many charts undergo a complete over haul in this respect." What Captain Wharton calls the phonetic system must, of course, rule in all languages without an alphabet ; in others the native spelling ought surely to be as authoritative as in German or Italian. It is here, we THE ORTHOGRAPHY OF FOREIGN PLACE-NAMES. 9 reiterate, that the real difficulty lies : if the geographer would call in the aid of the scholar and accept his decision as to transliteration, the greatest of all the difficulties would disappear ; persist in aiming at spelling names by the ear only, and all progress will be stopped by multiplied difficulties. It will, moreover, be found that, as already stated, in cases under the Board's second class, the written form, accurately transliterated, will yield a far better representation of the name than any attempt to write it down from the pronunciation alone: local pronunciation varies so greatly, and is so very difficult to fix with certainty, that personal differ ences of pronunciation and of hearing- will always combine to produce innumerable divergences in the representation of names ; whereas all other alphabets are more uniformly phonetic than the Roman (especially as used by English-speaking peoples), which renders transliteration the best guide to sound.1 Wisely, too, the Admiralty considers no geographical dictionary hitherto published as a standard, and examines different authorities which may be considered most worthy of credence for each separate district. " The question of those names which are already well known under a certain form is another great stumbling-block. It is impossible to sweep away forms known for years to all English-speaking people, and where to draw the line is the difficulty." Here the American Board rule (/;) would best solve the difficulty, and familiarise the public with the correct forms before those so long known are swept away. New forms can only be introduced gradually, and by carrying the knowledge of the public along with such advances and corrections. In the detailed system of the American Board for Greek names, it is proposed to transliterate 77, i, v, et, 01, and vi, all by the same letter i, /3 by v , . Surely ph, though a digraph, might have been retained for the former, so as to preserve the difference for philo logical purposes — which we cannot afford altogether to dispense with — rather, which scholars wisely seek to retain as far as practicable. Further, the semi-vowel i-s'-kratkoi, the last letter of the alphabet, is not trans literated after i. The system proposed is contained in the following table : — THE ORTHOGRAPHY OF FOREIGN PLACE-NAMES. 11 RUSSIAN. ENGURIT EQUIVA LENTS. HUSSIAN. ENGLISH EQUIVA LENTS. A.., . a (TO #. a T ...T J^ m t B..„„. 6 T..tf. cf ?> y.... y * y a B. B .# / V . 4> $>-.. 4> f r. ..... r ,r * 9 X X kh A... ••* .#...£. ¦¦•? d II.. u 2/j 24, to E..... ...e £,.... ....* e q.... ....n Y. y ch Jit- 3K . i^C j E... m VL^.ftu. sh 3, 3 >... lb 1. * I K i K. .7 / i k b £ 6 : ui X .*• .1\ A ./* .* I U... ...* J& ..b...7t ye, M.._. ..M J£ .*•* m 3.. a 3. 3 e H , rl X _./£- n KL . K) ya 10 yu 0 0 ^ <7 0 JI... >r JL JL ya n.,.. TT ,r .71 P e. . e d. e f p. -P j? ../I r Y. .. V 7/1 v 6 c C ,« C s ft H a -a V Here it appears to have been entirely forgotten that the Slovenik family of languages generally employs the Cyrillic alphabet, with some necessary omissions or additions to suit the necessities of each tongue; and that, alongside this, we have a co-ordinate alphabet in Roman char acters, which is used in Poland and elsewhere, and which in Serbian is em ployed by the Catholic Croats, while the Cyrillic is favoured by the Serbs of the Greek Church. The one alphabet is transl iterative of the other, and is established in the literature of the countries concerned ; and with 12 THE ORTHOGRAPHY OF FOREIGN PLACE-NAMES. this any proposed mode of spelling place-names in Austria, Servia, Bul garia, Russia, etc., must take account. In face of a recognised system no counter proposal of a society or congress can possibly prevail. If we accept the spellings on the maps of the Austrian Government, as agreed in the first rule, we therein adopt this system of transliteration for large areas of that empire. Shall we adopt a different system for the neigh bouring lands of Servia, Wallachia, and Bulgaria, or even Russia ; or shall we accept an established system of transcription, recognised and used in certain Slovenik countries, and apply it consistently to all 1 The Roman alphabet is adapted to the Cyrillic, by adding diacritical marks on the z, c, and s, and by a few digraphs, to represent the sounds we would express in English by zh, ch, fh, etc. If we find it is, for typographical reasons, desirable to avoid consonants with diacritical marks, then the analogy of the h in each case, in place of the mark, would be preventive of confusion in thus rendering the 7th, 24th, and 25th letters of the Russian alphabet. The 22nd, X, is represented in this system by the Roman h, which is sounded as a guttural, as in German "acht"; the 23d is transliterated by the Roman c with a sound like ts. The 28th letter (yiri) has not the sound of ui except accidentally, but of i and y in "pity," and is represented by y; the 35th (izhitza) has hardly the sound of German 6, but of Italian i somewhat long ; while the last letter (i-s'-kratkoi) is nearly mute, as y in " toy," and is in fact 1. The 27th letter (yer) is like our mute e, but makes the preceding consonant hard, as if doubled; yeri, the 29th, makes the preceding consonant soft, and often at the end of words has the short sound of l in " bit " ; it is represented in Bohemian by the apostrophe ('). Lastly, is distin guished from G (pldta or thita) by rendering the latter by ph, which can occasion no difficulty. The 31st is the only letter not found in the Cyrillic alphabet. The Serbs use six letters not in the Russian alphabet, and which they transliterate by — gj (or dj) with the sound of dy (Magyar gy) ; j (y) ; Ij — like the Italian gl (Magy. ly) ; nj (fi) like gn in French and Italian ; 6 with a sound between ch and ts ; and dz a guttural j. But it drops all letters after the 25th in Russian ; other Slovenik alphabets drop the last three, whilst Wallachian adds x (£) and ps (-uV). The transliterated alphabet in general use (together with English equivalents to avoid diacritical marks and the continental sound of j) is — a, b, v, g, d, e, z (zh), z, i, i, k, 1, m, n, o, p, r, s, t, u, f, h, c, c (ch), S (sh), sc (shch), (yer\ y, ' , je (ye), ju (yu), ja (ya), ph, y (t). The American proposed alphabet lor transliterating Russian, then, does not take sufficient account of recognised conditions ; but to do so only a few changes have to be made in it (i.e. in the transliteration of the 7th, 22nd,. 23rd, 28th, 34th, and 35th letters); and this would adapt it generally also to all the Slovenik countries not already using the Latin alphabet exclusively. Japan may be said to have already an official geographical nomen clature in Roman characters, and this has been adopted by the Board with slight changes. With regard to Chinese names, the prevailing diversity is well THE ORTHOGRAPHY OF FOREIGN PLACE-NAMES. 13 illustrated by the varying spellings of the names of six Chinese provinces by different authors, thus : — Keith Johnston Quei-chow, Sze-chuen, Hoo-pe, Fo-kien, Ngan-whi, Chi-li. Black . . Kuei-choo, Szu-chuan, Hou-pe, Fu-kian, Ngan-hoee, Che-lee. Stieler (Kwoi-tschou, Sz'-tschwan, Hu-pe, Fo-kien, Ngan-hwei, ' \ Tschi-li. Stanford . . Kwei cliau, Sz chuen, Hu peh, Fu kien, Ngan hwui, Chi li. Dr. W. Williams Kweichau, Sz'chuen, Hupeb, Fuhkien, Nganhwui, Chilili. The Board gives the preference to the system of Dr. Wells Williams, which it fairly deserves. In regard to Indian names, the authority of Hunter's Gazetteer has been, perhaps rashly, adopted by our Admiralty. In a former paper (S. G. M. vol. vii. p. 357) reasons have been given why the spellings in this work should not be accepted en bloc. It is to be understood, however, that it is not the system — which was intended to be translitera- tive — that is at fault ; a large number of names have been corrected in that work, and others have been brought nearer to their proper forms than was previously the case ; the dissent applies only to those names that are still presented, against all system, under wrong forms of spelling. Were the work thoroughly revised by competent scholars, it might then be regarded as authoritative, and it will be a pity if the next edition is not very carefully revised in this respect. Burmese place-names have hitherto been badly represented on all our maps. The Burmese, Siamese, and Kambojan alphabets are purely Sanskritic, and even the forms of the letters may be distinctly traced in the older types of the Indian alphabets from which they have branched off. The letters also retain the Indian sounds of the vowels and con sonants, arranged in precisely the same order. In Burmese, however, the letters for s and sh are omitted, the vowel ai is sounded as a lengthened e (Italian), and the palatals are pronounced more like ts, ch, z, and zh, than the sounds of ch, chh, j, and jh, given to them in India, while the sibilant s has often in Burmese the sound of th in " thin," but also that of fh in " fhin " ; it is in fact a lisped cerebral s'. The Jonesian alphabet, as modified by Wilson, would answer equally for the translitera tion of all Burmese,1 Siamese, Kambojan, Singalese and Tibetan names, as for those of purely Indian ones. For Turkish, Arabic, and Persian names, as already mentioned, the American Board has supplied no detailed system. Here it would be well to adopt a mode of transcription already accepted by scholars, with only such modifications as are necessary in ordinary printing — reserving the details of diacritical marks for scientific indexes and guides to the 1 The transcription of the Burmese palatal and guttural nasals by gny and cjn is undesir able ; they have simply the sounds naturally given to the nasal n, when it accompanies a palatal and guttural respectively, and do not require digraphs and trigraphs to indicate this. In the standard alphabet they are indicated by ft and n. The combinations ht and hp, some times used to represent the aspirated forms of t (palatal and lingual) and^i, are also deceptive and may well be dispensed with, for the corresponding Indian transliteratives of th, th, and 2)h. 14 THE ORTHOGRAPHY OF FOREIGN PLACE-NAMES. pronunciation of geographical names. The difficulty connected with this alphabet, used over the extensive area from Marocco to the Indus, and from the coasts of the Indian Ocean to the Danube and the confines of Siberia, lies partly in the representation of the numerous sibilants. Some of these, too, are sounded in one way by Arabs and otherwise by Persians and Turks. The transliteration according to a recognised system, how ever, presents little difficulty in the consonants; it is in the vowel- accents and tashdid, so generally omitted in Arabic, that the difficulties of accurate transcription lie, and even this can be overcome by taking trouble. The following table presents the elements of transcription for Turkish, Persian, and Arabic : — 4, a a Letters. Names. 2 3 Sounds. Ph ¦5 1 Alif is only a support for a vowel. <__> Be b b b as in babble, rub. <_-> Pe P P as in puppy, sip. C15 Te t t t soft dental as in Italian, nearly as in tube. o Se s i th Ar. nearly as in thin ; Persian as s in sister. z Jim j j j as in jet, jam. z Che ch ch soft ch, as in church, Italian c in dolce, etc. Z He h h h a strong aspirate pronounced forcibly in the throat. like ch in loch, or the Gaelic ch. t Khe kh kh kh j Dal d d d soft dental as in Italian, nearly as in dew. o Zal z z dh as in zeal ; but rather as dh by the Arabs. Eg r r r as in ran, bur. J Ze z z z as in English, like <^ in Persian. t, J Zhe zh zh like s in pleasure, or % in azure. u» Sin s s s sharp s. LT Shin ft ft ft as s in sure, session, shine in sfe. U° Sad s s s as in sire. U° Zad z z d t Ta t t t a hard t, sounded with the tongue full against the front part of the palate, as in true. 1? Za z z z % hard. L 'Ain ) ) j a guttural formed in the lower part of the throat. t Ghain gh gh gh a peculiar hard guttural like the Northum brian r, or the French in grasseye. L— J Fe f f f soft letter as in fill. THE ORTHOGRAPHY OF FOREIGN PLACE-NAMES. 15 Letters. Names. ¦3%H 1 § q Sounds. J V Qaf q q a very hard guttural sound stronger than ch in thich. soft h, as in Icing, hill. c) Kaf k k k ^J Gaf g g always hard, as in good, give, dog. ; y as in year, yet, loyal ; used both as a semi vowel and. vowel. Maddah i a a aa aa short a as in America, rural ; answering to Arabic fat'ha. long a as in ball, tar. ai ai ai i in nigh, fine, ai in aisle, in grammar, — ya- i-sakin ma qabl uiaftuh. j>< au au au as in our, in grammar, — wav-i-sakin ma, qabl maftuh. ,_£ J. j! e e e as ei in neighbour and in prey, ya-i-maj'hul. \ t ' Kasr i i i i in fin, sit. s i i i same ; in grammar, — hamza-i-mulaiyana. lS £ i) i ? i i i in marine, police, ya-i-fna'ruf. ;J 0 0 0 as o in bone. 1 Zamm u u u u in full, put. u u ii u in rule, wav-i-marfif. 0) Tashdid doubles the letter it is placed over. o Sukfm The letter it is placed over has no vowel. 16 THE ORTHOGRAPHY OF FOREIGN PLACE-NAMES. To carry out this much-needed spelling reform of geographical place-names, not merely in charts but in all maps, our first and most pressing need for each foreign country is authoritative indexes of all names (like those for Austria), printed in the alphabet of the country, whether Greek, Russian, Arabic, Japanese, Hindi, or Burmese, and, where necessary, with a transcription (liable to revision) into Roman characters. It would add to the value of such lists if pronuncia tion could also be represented in cases where the spelling does not quite convey it. These lists could only be drawn up in the countries to which each relates, and would require that the transcriptions should be revised by scholars in accordance with the transliterative systems adopted by international arrangement for each language.1 Except to- indicate the quantities of vowels in certain cases, diacritical marks could be dispensed with on maps and in ordinary books ; in the standard indexes, they would be usefully retained. For such works there appear to be considerable collections already in existence for some countries, but which may not be sufficiently known to cartographers. "The following are mentioned in the American Report : — For Austria there are official lists of the names of every town and village in each province of the empire. They are — Special Ortsrepertorium von Steiermark ; von Karnten ; von Krain ; von Bohmen ; von Tirol und Vorarlberg ; von Kiistenlande ; von Mahren ; von Schlesien; von der Bukowina; von Galizien ; von Nieder Oesterreich ; von Ober Oesterreich; von Salzburg. Geographisch-statistisches Re- pertorium der bewonten Orte im Kbnigreich Dalmatien ; Slovnik geografizny des Kbnigreiehs Polen und anderer slavischer Lander ; Ortslexikon des Kbnigreiehs Polen ; Ortschafts und Bevblkerungs Statistik Bosniens und der Hercegovina ; Ortslexikon von Ungurn by Jekelfalussy ; Postlexikon der Lander der ungarischen Krone ; and Orstlexikon von Russland. For Denmark, etc. : — Statistisk-topografisk Beskrivelse af Kongeriget Danmark, ved I. P. Trap, Kjbbenhavn, 1879. Uppdrattr Islands a fjdrum blbdum. Den Grbnlaudske Ordbog af Sam. Kleinschmidt, Kjbbenhavn, 1871. For Dutch countries : — Woordenlijst van de Aardrijkskundige namen in Nederland van bet Neder- landsch aardrijkskundig Genootschap, 2d druk : Amsterdam, C. L. Brinkmann. Veth's Java. Havenga's etappe kaart van Java, de residentie kaarten van Java, etc. For Spanish-speaking countries : — Don Pascual Madoz's Geographical Dictionary of Spain, 1845-50. Serior Mariana's Dictionary of the Corporations (cities and towns) of Spain. Riera's General Geographical Dictionary, 1880-85. Spanish- American Encyclopaedia (being published). Spanish Universal Geography, 1830. 1 Leaving the Armenian and .Sthiopic out of account, there are practically only three alphabets extensively used — the Arabic, the Sanskritic in its various equivalent forms, and the Cyrillic (including its basis— the Greek)— for which transliterative equivalents have to be accepted ; and these have been satisfactorily fixed for all geographical purposes by scholars and the Latinised alphabet of part of the Slovenik race. THE ORTHOGRAPHY OF FOREIGN PLACE-NAMES. 17 Royal Historical Academy's Geographical Dictionary of Navarra, etc., 1802. Don Jacobo de la Pezuela's Geographical Dictionary of Cuba, 1863. Prof. Don Juan Vilanova y Piera's Vocabulary of Geographic names. 101 Diccionai'io jeogrtttiuo de Chile, por F. S. Astaburuaga, New York, 1867. Lin uniforme pasado a la Universidad, por Don B. Vicuna Mackenna, in the Annals of the University, 1863. Derrotero del Archipelago Filipino, 1879. Diccionai'io Geogratico de las Islas Filipinas, por Buceta y Brava, 1 85(1. For China, Japan, and Korea : — Names of places on the China Coast and the Yangtse River, published by order of the Inspector-General of Customs, 1882. G. M. H. Playfair's Cities and Towns of China, London, 1879. A Concise Dictionary of the Principal Roads, Chief Towns and Villages of Japan, etc., by W. N. Whitney, M.D., Toyko, 1889. A List of Korean geographical names, by E. M. Satow, 1884. French-Korean Dictionary, with an appendix of geographical names, 1880. We want similar materials for the place-names of Russia, Greece, Turkey, Egypt, Algeria, Arabia, Persia, India, Ceylon, Burma, Siam, Tongking, and Central Asia, etc.1 1 A bibliography of the whole subject would be important. The following works arc valuable:— Bender, Die deutschen Ortsnamen; Forstemann, Die deutschen Ortsnamen; Buttmann, Die deutschen Ortsnamen ; Vilmar, Ortsnamen in Kurhessen ; Piderit, Ortsnamen in Niederhessen ; Meyer, Ortsnamen des Kantons /Mrich ; Finnson, Islands Landnaniaboh ; Edniuuds, Names of Places ; Skene, Celtic Topography of Scotland in Trans. Ii. Soc. JSdin., vol. xxiv. pp. 207-220 ; Fergusson, River Names ; J. Macdonald, Place-names of Strathboyie ; Robertson, Gaelic Topography of Scotland ; P. W. Joyce, Irish Local Names (two series) ; General Alphabetical Index to the Townlands and Towns, the Parishes and Baronies of Ireland— Census 1861 ; Professor J. J. Egli's articles, Veierdie Fortschritte in der Geographi- schen Namenkunde in Wagner's Geographisches Jahrlucli ; Nomina Geographica Neerlandica, uitgegeven door het Nederl. Aardijkskundig Genootschap.