adopting on the one hand Arabic and Turkish pronunciations such as “Moḥammed” (“Mahommed,” and, still worse, “Mahomed,” “Mahomet,” “Mehmed,” and the like, are monstrosities of which no Orientalist should countenance the use), or, on the other hand, usages based upon the phonetics of French and German. There are in Persian only three vowel-sounds, each of which may be long or short; and it is essential, both to correct pronunciation and to correct comprehension, to distinguish the long vowels either by a long mark, or (which I prefer) by an acute accent. These vowels are:
| a (short) as in “man”; | á (long) as in “all”; |
| i (short) as in„ “sin”; | í (long) as in„ “machine”; |
| u (short) as in„ “pull”; | ú (long) as in„ “rule” or “pool”; |
There are also two so-called diphthongs, consisting of the short a followed by one or other of the weak consonants w and y, Of these aw is pronounced exactly as the same combination is pronounced in Welsh (“mawr”), or like the English ou (“house,” “out’’), or like the German au (“auf,” “aus”); while ay is pronounced like the English ay in “hay,” “may.” There is therefore no occasion to use e and o at all, nor, as a matter of fact, do those who use them do so consistently. Those who write “Yezd,” “Resht,” “Enzeli,” and the like (to indicate, presumably, that the vowel is short), should, to be consistent, also write “Tebriz,” “Hemedan,” and “Isfehan.” And if it be said that some of these inconsistencies are sanctioned by usage, and that they ought not to be altered, the answer is that it is both easier and more philosophical to transliterate on a fixed and definite principle than to decide in each case whether a given spelling has or has not been sanctioned by usage. Therefore even in the case of the most familiar place-names I have rigorously applied the system which I have adopted, writing always “Ṭihrán” (not “Teherán”), “Anzalí” (not “Enzeli”), “Najaf” and “Karbalá” (not “Nejef” and “Kerbelá”). Similarly, in speaking of the Bábís, I have abandoned the spellings “Ezelí” and “Behá’í,” which I formerly used, in favour of “Azalí” and “Bahá’í.” It must also be borne in mind that in the case of Arabic derivatives, which are of constant occurrence in Persian
