seem that these letters were knowingly used indifferently; on other occasions ignorance of or insensibility to the true force of the Semitic may have prevailed; though in some instances, again, discrimination in their contrasted employment is evident, especially in words in which a complication already exists, arising out of the community of the sounds of R and w inherent in their common sign 2.1 If, in addition to these constructive difficulties, we add the imperfect phonetic aptitude or the want of system in the use of the symbols for J-D and -T, -G and S-K; and more important than all, the authorised dialectic interchange of — B, — P (F), and ,w, we have offered a goodly list of reasons why European interpreters have made such scant progress in Pehlvi readings.

One of the most curious questions in the whole range of this enquiry is presented in the history of that strangely influential vowel in the Persian tongue, the letter i; we have already seen the important part played by the normal form of that character in the supplementary definition of the concurrent signs of the Chaldæo-Pehlvi, and attention has been drawn to a somewhat parallel fundamental influence exercised by the typical curve of the Sassanian i, among the other letters of its own alphabet; it is further clear that neither of the very differently-fashioned letters of the joint Pehlvi systems of writing can be referred to corresponding Semitic originals as the latter are ordinarily determined; all of which adhere with more or less fidelity to a vague reminiscence of the archaic . A singular evidence of the community of Aryanism in alphabets suggests itself in these facts, though I am not prepared to claim any Noachian antiquity for the coincidence, but merely desire to show that the various branches of the Aryan pastoral races, as they are known to the modern world, only began to understand and appreciate the value of

1

It is . تیلدت and تیردت - فلمات and فرمات- ملكا and مركا :

a curious fact that all the early Numismatic legends use 2 both for R and w. does not appear till later, and then only irregularly. See J.R.A.S. xiii. 178. 2 Report of the Meeting of the Royal Asiatic Society, 9th April, 1866; Athenæum, April, 1866; Numismatic Chronicle (1866) vol. vi. p. 172; Journal Asiatic Society of Bengal, July, 1866, p. 138.

the art of writing when they came into contact with urban populations in their own migatory advance and domestication among more civilized peoples, or when they achieved, in force, the conquest of earlier-settled nationalities. In this present case, at least, it is strange that the self-same leading idea should have prevailed throughout, in the adoption of the crude form of the vowel i, within a range that can be traced upwards from our own capital or italic I, through the Roman and Etruscan outline of the letter, and the independent Greek design,1 whose but slightly modified shape is found typical in Armenia2 some centuries B.C., and which re-appears almost identically in its normal tracing with our own matured result, in the Bactrian reconstruction, under Aryan treatment,3 of the simple elements of the once current writing of Babylon.

The Sassanian alphabet manifestly incorporated the old Phoeniciani (the Persian Cuneiform 4 into its own system, and as it was already in possession of an ordinary short ; the Semitic letter was devoted to the representation of the long or duplicated sound of that vowel.5 A curious course 1 The following forms of the Greek iota approach very closely to the ChaldæoPehlvi outlines See also Gesenius, pl. ii.; Mionnet, volume "Planches,” etc., 1808, pl. xxxi. Nos. 1, 2; "Inscriptiones Græcæ Vetustissimæ," H. G. Rose (Cambridge, 1825), table i. Nos. 11, 15, 18. etc.; "Corpus Inscriptionum Græcarum," A. Boeckh (Berlin, 1828), p. 6. "Sed imprimis insignis est litteræ Iota forma, quæ etiam in ære Petiliensi reperitur, et tum in nummis aliquot urbium Magnæ Græciæ, tum in nummo Gortyniorum, derivata ex Oriente."-Swinton, Insc. Cit. Oxford, 1750.

...

2 Coins of Artaxias, Numismatic Chronicle. October, 1867, No. 3 [1],

3 The Bactrian medial i is composed of a single line thus. In composition it crosses the body of the leading consonant. The initial is formed by the addition of the sloping line to the short a, thus

iii. pl. vi.; Prinsep's Essays, ii. p. 161.

.-Numismatic Chronicle, N.S.

There is some similarity of ideas in the form of the Pali 7 of Asoka's Inscriptions. Ex. gr. ũ ghi, ủ giữ

5 M. François Lenormant has devoted a lengthy article in the Journal Asiatique of Août-Septembre, 1865 (pp. 180-226), to "Etudes Palæographiques sur l'Alphabet Pehlevi, ses diverses variétés et son origine," in which he has done me the honour to quote largely from my first paper on Pehlvi writing which appeared in the twelfth volume of this Journal, 1849, as well as from a parallel notice on Arsacidan coins, etc., inserted in the Numismatic Chronicle of proximate date, without seemingly having been aware of the publication of my second contribution on the same subject, which was printed in our Journal for 1852 (vol. xiii. p. 373). M. Lenormant has not been altogether fortunate in the passages

attended the maturation of this literal sign in the parallel alphabet, which, though in the retention of its primitive forms, claiming so much more of a Semitic aspect, provided itself, from other sources, with a short ž, and lost all trace of the proper Semitic of Sargon's time, and hence had to invent anew the long i required for the due expression of the language it was eventually called upon to embody. The process by which this was effected is instructive, and may be said, in its

of my Essay which he has selected for adverse criticism,-a licence, however, I must confess he has been wisely chary of indulging in.

M. Le Normant is mistaken in supposing that Sir H. Rawlinson ever designed to insert a long final in the word Baga, so that his over-officious attempt at correction, in this instance, proves altogether superfluous (J.R.A.S. x. pp. 93, 94, 187), but the implication, in the general run of the text, is, that I myself had attributed this error to Sir Henry, which I certainly never contemplated doing, nor, as far as I can gather from anything I have printed, did I give any colour for a supposition that I desired so to do (J.R.A.S. xii. 264; Numismatic Chronicle, xii. 74). Sir Henry undoubtedly suggested that the group of letters ordinarily following the king's titles in the Sassanian coin legends and inscriptions should be resolved into the letters B. G., and hence he inferred, most correctly, that the term in question was Baga, divine (Sanskrit H), supposing that, in the ordinary course of Aryan tongues, the several consonants optionally carried the inherent short vowel a. My correction merely extended to the separation of the character composing the second portion of the group into the since universally accepted g. i. M. Lenormant has gone out of his way to assert that "Le savant anglais a prétendu, en effet, que le pehlevi ne possédait pas de D." This is not quite an accurate statement of the case. If I had not recognised the existence and frequent use of an, which letter duly appears in my alphabets (J.R.A.S. xii. pl. i.), I could have made but very little progress in Pehlvi decipherments. The question I did raise with regard to the origin of the earliest form of the Sassanian (xii. 266), as found in the Hájíábád sculptures, was not only perfectly legitimate and fairly and frankly stated, but there is even now no resisting the associate facts that the Chaldæo-Pehlvi version of Inscription No. vi. infrà, makes use of the in the penultimate of ¡, and that the corresponding of the Sassanian text

is susceptible of being resolved into the typical elements of . Moreover, it must be borne in mind that the Chaldæo-Pehlvi D was still unidentified, though I even then suggested the attribution which has since thrown new light upon the entire question (N.C. xii. 78). In short, the point of interest at that time was to determine the course and progress of the discrimination and graphic expression of the approximate sounds of z and s in the alphabets under discussion.

Boman, which

As regards my proposed rectification of M. De Sacy's D in 1 M. Lenormant confidently designates as “inutilement contesté par M. Edward Thomas" (J.A. p. 193), I am sanguine that the ample data adduced below will satisfy more severe critics that the mistaken interpretation M. Lenormant insists upon sharing, in common with so many of Anquetil's ancient errors, may be safely left to find its own correction.

Finally, I am bound to place on record a distinct protest against the general accuracy of M. Lenormant's illustrative facsimiles. I imagined, in the first instance, that the French artist had reproduced in a crude and clumsy way the conscientious originals of the English engraver; but I see that M. Lenormant claims whatever credit is due upon that score for himself, in the declaration, nous avons relevé nous-même les figures que nous donnons sur les plâtres offerts à la Société Asiatique de Londres par M. Rawlinson" (J.A. p. 188).

66

very mechanism, to add an independent proof of the true value attaching to the fellow character ~. The configuration of the clearly proceeded upon the duplication of the simple or short (); and in order to avoid the possible confusion of the new compound with the ordinary a concluding curve was carried upwards and backwards from the second i through its own down-stroke and into the leading letter.

In course of time both these double letters disappear from public documents, but the Sassanian letter is preserved in the Parsi alphabet, and is but little changed in its Zend form ?. While the short i was subjected to considerable modifications, till, on the Arabico-Pehlvi coins it appears as in its independent definition, or in the latest introductory stage towards the Naskhi "Kasrah-i-Izáfat.”

As regards the true force of the fellow letters, though we may, for simplicity sake, designate them as long or double i's, it is clear that the duty they had to perform in the less matured orthography of the third century A.D. will be represented by a very extended range of optional transcriptions when reduced into the elaborated characters of the present day, leaving the Chaldæo-Pehlvi letters to answer for their parallel power in the double ". The Sassanian counterpart must clearly be admitted to stand, according to the conSor, and their several medial corres

,ي text, for or ي ي دي دي

pondents.

An apt illustration of the difficulty the limited characters of the Chaldæo-Pehlvi had to contend with in the definition of the mixed Aryan and Semitic speech they had to respond to, has lately been contributed, on the occasion of the natives of Persia having been called upon to reconstruct an alphabet suitable for the expression of their modern tongue out of the self-same literal elements they had abandoned so many cen

1 Spiegel, Grammatik der Pârsisprache. Leipzig, 1851. I observe that Dr. Haug still adheres to the old lesson his Parsi instructors at Surat so erroneously taught Anquetil in 1760, and persists in interpreting the power of this letter as See preface to the "Farhang-i-oím yak," p. 21. Though he seems at one time (1862) to have been prepared to accept the reading of j, converting the old 'Boman' into 'Barj.' "Sacred language of the Parsees," Bombay, 1862. p. 45.

من

turies ago. The motive for this experiment arose out of the desire of our Bible Society to furnish the Jewish converts in Persia with a version of the New Testament in the Hebrew character, with which they were already familiar, but textually couched in the spoken language of the country. The subjoined table will show how this singular compromise was effected, and its details are of considerable value in the present inquiry, as giving us a clearer perception of how the modern ear was prepared to deal with the sounds of the actually current speech, and how, with a clear field and enlarged and matured powers of alphabetical development, those sounds were held to be critically defined and discriminated in the general reconstruction of the ancient alphabet.

HEBREW ALPHABET ADAPTED TO THE DEFINITION OF THE PERSIAN LANGUAGE.2

One of the most curious results of this adaptive revival of the ancient letters is to prove to us, what I have already perseveringly contended for, that is, the use of some form of a double i, and some acknowledged method of writing such a compound with a view to avoid the possible confusion of the independent repetition of the short vowel, amid a series of letters in their nature so imperfectly discriminated inter se. Examples of

The New Testament in question, designated "JUDÆO-PERSIC," was printed by Messrs. Harrison & Co. in 1847, under the editorship of Mr. E. Norris, from a text arranged by the natives of Persia according to their own perceptions of equivalent letters.

2 Michaeli's Arabische Grammatik (Gott. 1781) arranged the discriminative marks as follows: — ♬ = ☺, ñ = ☺, n = 2,5 = 7,5 = ¿,i = ¿,

رج

وت
ת
دلت
ח

.8 = 1 ق = P ,غ = b, i = , I = ۵ ,ض = لا

وص