Each of these sections was then gone through, in company with Mr. Ross and, usually, with one of the texts at hand. The uncertain passages were then discussed and filled in, or marked off for further investigation; but the revision of the English, and the addition of footnotes, together with the solution of the puzzles in which the book abounds, were subjects left for me to take up at a later opportunity. In editing the English version, my object has been to render the language fairly clear and readable, without so changing the translation as to alter the author's meaning. But this has proved to be not always an easy task, for numerous passages occur where the translation will only bear slight amendments, without impairing its accuracy. As a rule, indeed, the latitude that could be allowed was small, so that it has been necessary to leave many sentences and paragraphs standing in rather awkward phraseology, which it would have been easy—and was even tempting—to improve, had the author's words permitted the license.
As regards the spelling of proper names, it was at first intended that all should be reproduced according to the system in use at the British Museum—i.e., an exact transliteration, where each Persian and Arabic letter which has no single equivalent in English, is distinguished by some accent or diacritical sign. This design, however, could not be followed out for various reasons. Indeed, the work of translation was scarcely finished, when Mr. Ross left England to pursue his Oriental studies on the Continent, and the task of attempting to harmonise the spelling fell to me. I found it impossible, with the time at my disposal, to carry out the original intention, seeing that many of the names would have had to be searched out, afresh, in the vernacular texts—an undertaking that would have involved frequent attendance at the British Museum. I decided therefore, to write all according to the simple method of Sir W. Jones, or that adopted by the Government of India in their official documents and publications. This system may be, to a certain extent, imperfect, for it does not, in every instance, show exactly what was the Persian or Arabic spelling of the word represented, and therefore is not a complete guide to re-conversion; but it gives, as nearly as possible, the correct pronunciation, while it secures a degree of accuracy sufficient for practical purposes.[1] It has moreover the advantage of
- ↑ It may be said to consist of accentuating the a, in order to give the sound of that letter in father; while other vowels are expressed (as Mr. H. G. Keene has tersely put it) "by the English sounds in 'ruminant' and 'obey'"
