\ «• . vV fc ^v^ L "‘ y ^ TGIT PIUXTC V. TI-IH0Ii06IGi.V^ i £ ? v ^rr r r.TNV9r{ * **•>> v-i- 1 I : "iV r , -,...P.K6!2-2 - c ;,.-,..aW.5 2, Y Digitized by the Internet Archive in 2016 https://archive.org/details/ondecipheringofsOOwest 9 , p T> T ~T p nriTT * -'v- i. At UiJ 7H30L0G ESSA1 SUR EES M0NNAIES DES B0UIDE8. 271 - ' canton du Tabaristan auquel Ibn Haukal , ed. Ouseley p. 179, donne le nom de 20. Sermenrai, La monnaie du n°30 de 1’an 341 a etc frappee dans cette ville qui est aussi nominee Samerra, elle est situee dans l’lrak de 1’Arabie ou elle fut fondee par le calife Motassem billah. Elle acquit bientot de la prosperity , mais elle se perdit aussi vite quandles califes cesserentd’yetablir leur residence. Ibn Haukal, ed. Ouseley p. 08. Cfr. Fraehn, Num.cuf. p. 61. 21. Siraf, La monnaie du n° 2 de Tan 327 , celle du n° 7 de l’an 334, et celle du n° 42 de Fan 341 out ete frappees dans cette ville. Selon Ibn Haukal (ed. Ouseley p. 105) les villes deChiraz et d’Arghan etaient avec celle-ci les plus considerables du Persis. ON THE DECIPHERING OF THE SECOND ACHiE- MENIAN OR IHEDIAN SPECIES OF ARROWHEADED WRITING; by N. L. Wkstergaard. The countries situated between the Euphrates and the desert of Persia, being the home of the four nations that were among the first which made their appearance on the stage of history, were also the birthplace of that peculiar system of writing usually denominated the arrow- headed or cuneiform, from the circumstance of the figure of an arrow-head, or cunens together with an angle being the principal elements by which numberless combinations of letters and syllables were formed. Of this arrow-headed writing we are now able to distinguish five species, differing from one another in the shape of nearly every letter or group , if I may be allowed so to designate the several « 1 O-C # ct -to _- >--< 272 MEDIAN SPECIES OF ARROWHEADED WRITING. combinations of arrow-heads and angles forming single signs and characters for letters or syllables. The inscrip- tions on bricks and cylinders from Babylon exhibit one species of arrow-headed writing which has been most appropriately termed the Babylonian , as it seems to be peculiar to that country alone , and this species is the most complex and intricate of them all. Another species is employed on the monuments on the Median frontiers, on the right bank of the upper Euphrates and the shore of the Mediterranean, and chiefly on those discovered at Wan, and now again appearing from beneath the mounds of ISinive. This species, being found both at the Assyrian capital and throughout the whole empire in its widest extent, most probably was peculiar to that country, and therefore may be called by the name of Assyrian. The three other species have as yet only r been found on the monuments of the Achaemenian kings of Persia , having been simultaneously used by them in their inscriptions. One of the languages employed , and which is written in the most simple species , is a sister tongue to the Zand, and both this species together with the language have therefore been called the old Persian and attributed to Persis proper, the native land of the Persian princes. As to the locality of the two other Achaemenian species, however, much doubt and uncertainty prevails, but as one of them is apparently connected with that species, which I have called the Assyrian, and the other with the Baby- lonian, l think they may most probably be referred to the neighbouring countries of ancient Media and Susiana; and as we must look in Bactria for the Zand language, there are no other countries but those to which these two species may with more probability be attributed ; although in so doing, and especially in ascribing the second MEDIAN SPECIES OF ARROWHEADED WRITING. “273 Achaemenian species to Media, we are obliged to overlook and disregard the testimony of Strabo, who plainly tells us that the Medes and Persians spoke nearly one and the same language. When and where this system of arrow-headed writing took its origin is also a matter of uncertainty. To me it seems most probable that Babylon was its cradle, whence it spread in two branches, east- ward to Susiana and northward to the Assyrian empire, from whence it passed to Media and lastly to ancient Persis, where it was ultimately much improved and brought to its greatest perfection. The three Achaemenian species became extinct with the last kings of that dynasty, the two others perhaps earlier, when the countries to which they belonged fell a prey to their more powerful neighbours. The Persian species of arrowheaded writing has been deciphered chiefly by the genius and labours of Lassen and Rawlinson, men who stand prominent among oriental scholars, few having contributed more than these gentle- men towards advancing our knowledge of ancient India and Iran. From their opinions I have ventured to dissent in but a few points only, further than the mere arrange- ment of the letters into a system approximating nearest to the Armenian, but as I think more regular and perfect. The Persian species exhibits a pure alphabetical system, every sound, whether consonant or vowel, having its own peculiar letter or group ; the short a after a consonant being left unexpressed by any sign or mark, as is the case in most oriental alphabets. The vowels are three nil TT> a. u, either short or long; the a being always long when written after a consonant; the short a , which is not expressed, I therefore render by a German a, noting the sound and at the same time distinguishing it by a peculiar sign from the same vowel sound, when 1810 — 1844 . 18 *274 MEDIAN SPECIES OF AKROWHBAOED WRITING* expressed by its own letter. The simple consonants are thirty in number, viz. ^Y, q, qh, g, real gutturals like the Ara- bian ^5 these are wanting in Sanscrit, but two of them at least occur in Zand. «YMTK k, kh, g TT~, ^Tt- ■ c, ch, j, -YrY- <^Y, TY, t, th, d, , Y«, -Y 7 p, ph, b, HftY> K^> ni ’ ,Ti > D * <-<7 *~«> ft, h > *> v > the Greek palatals x, x> T- the Sanscrit palatals tx, ar, rr. Sanscrit and Greek dentals. Sanscrit and Greek labials. The second occurs only before the vowel i, and I am unable to determine the exact sound, which it may have expressed, hard consonants springing from the following semivowels. K-, 3, -YE, y, E, Y-Hf, z . <<, HK, s, z. Y, »• r, w. The hard and soft sibilants with their corresponding aspirates. These I conjecture to be the Icelan- dic p and 3, the hard and soft th of the English , forming the link between the sibilants and dentals. The group, ff- seems to have been a sign for a compound letter thr, at least it corresponds in an etymological point of view with the Zandic thr and Sanscrit tr. Another group, YK, may also be a compound letter rp , or rpli ; it occurs in one word only ► ( Y^ " r ith the sense of king, which, if read narpa, or nurphd, would be exactly the Sanscrit nripa , ruler of men. Besides these two, we meet, in an inscription of Artaxerxes the third, with two other MEDI AN SPECIE* OP A RK OWHK A DK D WRITING. 275 groups, which would seem to be abbreviations of a later date, the one for bund and the other or <»< in NR in two places The latter of these is to be read dasta or tasta, and the former might therefore be read tusta , giving to the value of tu, or du, which seems to be corroborated by another form under which the same verb occurs , likewise in NR , viz. T=m -2T Z as it is also indifferently written corresponds to the Persian t in the word Kiitpaduku , and to d in squdra , and therefore , there MEDIAN SPECIES OF ARROW HEADED WRITING. ‘283 can be no doubt that it represented a dental t or d; and from its use here before the syllable tn 1 conclude that this was the way in which the hard sounds were ex- pressed, a method that is not extraordinary, as, to quote only one instance, it has been adopted in the Tamil ortho- graphy. The other group n, may have been used here to denote that the following nasal m was a clear and distinct m, or, to show that the preceding vowel had a nasal sound, a case analogous, in some respects, to the long a before m in the Zand language. Another word that occurs in both species, is that which in Persian is written tiicdram, and here is rendered II BO. The known groups give: tat, ram. It is from this apparent that either the language did not possess the sound of 6 (c/0, or that the alphabet had no peculiar sign to express it , but that it rvas ne- cessary to use two signs to represent this sound. The Persian word Zanunam being here written yy ►Jf I do not hesitate to attribute the value of za to the group yy, and we then have in Zananam exactly the same word as in the Persian, and in tatzaram, or rather tadzardm a tolerably exact rendering of the Persian tdcdrdm, and a word that even comes somewhat nearer to the modern than the ancient Persian w r ord does- The w'ord zdnandm occurs in the Persian as the last part of a compound of which the first is written pdruw (parti, the to being used to mark that the vowel is either long or accented , thus the genitive case is written paruwnam i. e. parunam). This first part of the compound is rendered in the Median by FIG, or y D 8, E7, 0 15, K 12. The first group i as we have shown, represents the syllable pa ; so 284 MEDIAN SPECIES OF ARROWHEADED WRITING. it would seem , does also the ►f , both in this word pfirun as also in •> Persian Pdrsu , and in =^Y or lawa, is, according to Lassen , a nominative plural of Pcirpu , Parthus , so that the name of the people in the plural number is used as the name of the country. In the Median this name is written with three groups, of which the centre one is not a little defaced. What still remains distinctly legible is y, the other traces appeared to me to have been 3 5 which would have made the entire group or rather in analogy with MEDIAE SPECIES OF ARROWHEADKD WRITING. 291 and But as it is very difficult , if not imposs- ible, to be quite certain as to what such defaced, or nearly defaced traces may have been , the space between the separate wedges may perhaps (with as much probability) have been filled up by a which, when taken along with what is still legible , would make w and would thus give us the same group that we find as the second group in one of the forms under which the name of Persian is given in the inscription ]NR , viz. and which also expresses the name of the father of Hystaspes. This last name being written in the Persian inscription Arsti , we might conclude that this Median group m represented that very sound ot Arsti' or Ars. But now we find that in the Median wav of rendering the word Spdrda, the r is dropped before d, it being written HI >■% spliada, and here a question naturally arises : was the omission peculiar to this word only ? or may we not rather suppose that the Median language in general did not tolerate r before another consonant. This last supposition seems to receive corro- boration from another form of the above mentioned name of Persian, viz. T where r seems to have changed its place from before s to after it ; and therefore the group -“TTY must be supposed to have represented the syllable as both in the name Arsd and in that of Pursd. But although this may be admitted, there still remains one difficulty which I cannot solve, viz. why this name should be written Pha-as, or Pha-asdr and not simply P/i-as, or Ph-asdr. I may here as well mention a third form of this name occurring also in the inscrip- tion 1SK, viz. *- T «<; the last group is not found any- where else, but was most probably of a similar significa- nt 292 MEDIAN SPECIES OF ARROWHEADED WRITING. tion with the other os , perhaps ai. But to return to Par^awa ; the sound of p, which the Persians marked by the character was either wanting in the Median language, or the alphabet had no peculiar mark for it; for, we always find the Persian p rendered in a quite different way, s being generally substituted. Where- fore it seems likely that the centre group was that one to which I have ventured to give the value of as, and that the whole name, consequently, w r as written or Phusawa, a word that presents us the same diffi- culty as did the name Phasci , just above mentioned rendering the Persian Parsa. But if, on the other hand, the group were the real one, we would suppose it to have represented the syllable tha, a sound as nearly approximating to the Persian p, as that of s does, and being, moreover, in analogy with the Median rendering of the Persian syllable <5 u in the name of Katpabuka. The fourth name in the Persian inscription, Hariwa, Aria according to Lassen, is written here m The two last groups give rima coinciding with the Persian. The first group corresponds in this name and in that of Haruwatis to the Persian ha , but in Ariya to Persian a (or a), and in Arbaya, Arinina, to or. The same group occurs in the word rendering the Persian arddstand, which is written £ w hich also is used with the same value in the third species of Achremenian writing. The name corresponding to the Persian Hariwa w ould then lie Ariica , the h being dropped as is also the case with the Greek and Latin appellations of that country. The name of the fifth country is in Persian Bakhtris, in Zand Bakhdlii , and in Greek Baktria. This name is in the Median written with five groups of which the last is a little defaced , but traces enough remain to prove beyond a doubt that it must have been The whole name is therefore written *~YYYK ►YT* As the known groups give Bakh-rii, it is evident that the middlemost must have been a sign of a dental, which could only be the aspirate th, as the character ►Y * s the sign of the unaspirated dental sound t, or cl. We accord- ingly read Bakhthris. The aspiration of th must be attributed to the following r , w hich actually appears to have (occasionally at least) possessed in the Median language that aspirating pow r er which it always possessed in the Zand, of which we in the sequel shall show some other instances. The sixth name is Susuda, or Susda, in the Persian as also in the Zand language; Sogdia or Sugdia in Greek. It is written with three groups- in the Median. Of the middlemost of these only the last part was distinctly legible , viz. , before which I thought I could distin- guish traces of but I doubt not that the group was tY^ which we found in the beginning of the name of Cyrus with the value of qu. Here, of course, it represents the corresponding soft syllable gu; and as the first group 294 MEDIAN SPECIES OF ARROWHEADED WRITING. (being used here, where it corresponds to the Persian su) , appears to have represented the same syllable su, we have in yY,= , as I think the name was written, the Median SuQuda, corresponding, letter for letter, with the Persian. The seventh name is UwARAzmis, or, as it is written in the Persian inscription I, Uwarazmiya, which, according to Lassen, is the Zand Qdiri-zao , the land of nourish- ment, the modern Khavvarizm or Kharizm, and the Greek XopajpAa. In the Median the name is written I or Wariisvis , conse- quently with the omission of U in the beginning, and w ith the change of mi into vi. In this name , too , the 7V, or v would seem to have had in both places the hard or guttural sound we before noticed, which, if admitted, would at once account both for the omission of u and for the transition of w to the guttural Kh , and also for the change of the Persian z to s , as the soft letter z would not be tolerated before a hard letter. The eighth name is in the Persian Zarak, in which and in the Greek Zapayyai we have, according to Lassen, the proper East- Iranian appellation of the country of Drangiana , which derived its name from the lake still called Zareh. the sea, and as the Persian changes a Zand z into d, (as in Zand azam I, in Persian adam,) the other appellation of Apayyoc'. has probably originated from some West- Iranian language, perhaps even been the common appellation in Persian itself. Meanwhile, as this name in the Persian begins w ith a sibilant , so it also does the same in the Median, where the first group, although a little defaced YYf ) i can scarcely have been any other than the before mentioned YYT with the value of The second and third groups *~»=YT"~ MEDIAN SPECIES OF ARROW HUADE D WRITING. 295 ►►f j or r a, express here the same syllable ra that other- wise is represented hy »-yyy — a mode of writing ana- logous to what we have before observed was the case with the syllable yu. The last letter is ►f, or kha, and the whole name is thus Asraklia, which differs slightly from the Persian Zariik. The ninth name Haruwatis, the ancient Arachosia, Zand Haraqaiti, is in the Median rendered ►yy f or Aruicatis, which differs only in the omission of the li from the Persian name. The following name (the tenth) ^atacsus, the Sarra- •pSai of Herodotus, is in the Median written yf»H The three last groups give taQus, and the first ra. Of the signification of, the second group, though it occurs in five other words, it is a little difficult to form any certain opinion. The Persian pita, father, is translated which the last two groups alone would give us in dacla , or tata, a well known word with the sense required. The word that renders the Persian thuriya begins in the inscriptions D,F,K with f »-y£: phsati ; the inscriptions O and NR have :Hf ta instead of ti; in C a ►f, t is inserted before ti; and in E that syllable is preceded by tbe sign As this character is, in all these three words, found be- fore dental letters, and would even seem to be employed indifferently instead of t, we may conclude that it represents a dental sound , a value that also in J^y ►fy^ NR 16 would give us the same word as the Persian, datum. From the two other words (H 7 and 22) in which this character occurs, no inference can be drawn. Now, though it evidently represented a dental 296 media* species of arrowheaded writing. letter, still as 1 cannot determine its precise value, I pro- pose to render it by the letter t with a dot under it (t) to distinguish it from the other or t. The name then corresponding with the Persian Jatagus would, in the Median , be RattaQus. The same Persian [letter p is found, besides in the before mentioned name Parfiawa, (where it was, in the Median, apparently changed into $) also in two other words which I shall here examine. The one is AJiagina , a pillared hall , in the Median written f ~ The third group, as I shall endeavour to show in the sequel , seems to represent the vowel i , and then we have , Asina , the Persian word in a somewhat contracted form , and with s likewise substituted for p. The other word is the name of Assyria, AJmra, written here of which the two last groups give sur or sura. In the first group, one horizontal wedge at least has disappeared, which would make the group or g) but I regret that, at the time I copied the inscription , I did not observe whether any traces of writing were discernible before the distinct vertical wedge , which might make it probable that the group w r as rather or a; but at all events, whether this word be written Ssurii or Asura , it is evi- dent that in the middle of words, the Persian p is changed into $!, and that in the beginning, sometimes at least, ra is substituted. The eleventh name Gadara is that of the Gandhara- people in East Kabulistan. This name is written in the Median with three groups of which the two last are darn or dar. The first group appeared to me to be either »-yyy, or »-yyyy. In. the first case we should in »-yyy have a name somewhat MEDIAN SPECIES OP ARROWHEADED WRITING. ‘297 different from the Persian and Indian, viz. Radarci, instead of Gadara or Gandhara, for which alteration I am at a loss to account. We might, it is true, attribute to the other group the value of ga or ka , but as this group is nowhere else to he found , it would he very strange that such a syllable as ga or ka should not in all the inscriptions be met with more than once, and that, too, in a foreign name; besides, the external form of this group does not appear to have any analogy with the rest, it being the only one in which four equal wedges are placed together. The twelfth name is Hithus, the Sanscrit Sindhus, or rather the plural Sindhawah , the country of Sindh. Of the four groups, that compose the corresponding Median name, the last is a plain or i; the third is somewhat defaced, only remaining visible, but it is scarcely to be doubted that the two small wedges = have disap- peared from the original thu, or dhu; the second is the before mentioned or i; and the first sa ' As the group ►f kha joined toj ^ u forms the syllable kliu, so the group in a quite analogous way, joined to i might form the syllable si, though it is not easy to explain why one and the same syllable si should be ex- pressed in two different ways both by ►► as in this name , and by ^ as in the word dsina ; especially, as the latter is analogous to the common way of expressing similar syllables terminating in i, — thus the syllables pi, thi etc. are written p-i, th-i etc. The four groups ,then form •> °r Sitlius , a name , that has retained the Indian s, which in the Persian is changed into h. But I must observe that though the first group 298 MEDIAN SPECIES OF ARROWHEADED WRITING. in this name appeared to me to lie a quite distinct ga, still I noted down two other forms, it possibly might have been, viz. either or ►► , and if either of these should be the genuine , it might represent the syllable ha, as there is apparently no other group for that syl- lable ; and we should then have in HithuS exactly the same name in the Median as in the Persian. The Persian Sara is written Y ►f 5 sakkha, or sakha , with an aspiration of both consonants. By this name , according to Herodotus , the Persians called all the Scythian or nomadic tribes. The inscription on the tomb of Darius mentions, besides the Scythian, three Saka tribes, but their names are, unfortunately, just as defaced and obscure in the Median as in the Persian. The list of tributary countries in this| inscription begins with those of Media and Susiana, Persia not being men- tioned, as it was exempted from all taxes on account of its being both the chief country of all and the native land of the great kings. The enumeration then proceeds from Parthia in an easterly and north-easterly direction,, and then mentions, beginning with Babylon, the countries situate west of Media and Susiana , the list concluding with three or four mountain tribes on the borders of Media proper. The two first Saka tribes mentioned here after Giidara and Hithus, and before Babylon, must, therefore, be sought in the nomadic tribes to the north east of the present country of Kahulistan. The name of the first tribe begins in the Persian with Hcmawa, the rest being totally obliterated. In the Median, too, the name is in- volved in the same obscurity. The beginning is f ^7 hut the stone in this place presenting some unevenness, it may be dttubtful whether it was the group , now and then occurring, M, that was written, or, what would MEDIAN SPECIES OF ARROWHEADED WRITING. 299 seem to be more probable, whether the traces distinguish- able are two different signs, the first, f, the mark of distinction and the other, the letter u. Before the names of the other two tribes, it is true, the mark of distinction is left out, but then the relative pronoun ►►f ft is there inserted. The letter u, then, corresponds well enough to the Persian hu , especially as we have already seen that the h in the similar syllable ha is dropped in the Median names. The second group in this name, which also was indistinct, appeared to me to be either ►► sa, or, more likely which, as we shall afterwards show, probably represented the letter q, and might in this name correspond to the Persian syllable met, this being in other cases changed into rv, or rather into a hard or guttural sound of tv. The third group, which was distinctly visible, is tff»-, that seems to be of about the same value as tf»- bi, or pi, and might then have represented be, or pe, corresponding a little to the third syllable wei in the Persian name. The fourth group is tf t, or d. The fifth was somewhat indistinct, and I could not be satisfied whether it was a m or Rtt ; however, as the latter represents a dental letter only, and the former a syllable i/o, and as, in the case of the group standing here, the name would have terminated in two dental letters without any vowel following, it would seem more probable that the other is the proper one, and the whole name would then be ^ tff*- tf ^fff or Uqbetyo; but as to its geographical position I am as much in doubt as to that of the following tribe, which, doubtless, must also be sought in the same part of the world. This tribe is called in the Persian inscription Tigrakhuda, which seems indeed not so much to have been the proper native name 300 MEDIAN SPECIES OF ARROWHEADED WRITING. of the tribe, as rather some Persian appellation or nick- name, meaning the lords of arrows, as tigra might have the sense of an arrow and kliudd undoubtedly is the modern a lord. In the corresponding Median name , the second group is somewhat injured, only the latter part of it being discernible , hut there can scarcely be any doubt as to its having been a , kh, thus affording another instance of the aspirating power of r which we before noticed. The name then is written Hfc -ns or, with the exception of the last group, Tikhrakhuda. This last group, represent- ing, as I shall subsequently endeavour to make at least probable , the letter p , or b , might perhaps be nothing more than a plural affix , akin to the common plural termination bi in the Georgian language. With this name then, terminates the enumeration of countries lying to the East of Persia, and the list then proceeds from Babylon in a westerly direction. In the Median name corresponding to the Persian Babihjs or Babylon, the first group has lost the upper and lower horizontal wedges, and must be restored pa, or ba. The second group corresponding to the Persian bi, has undoubtedly the value of the same syllable bi or pi. The third, I have already stated, appeared to me to be traceable either to jEf, or and one of these, if true, might have represented the syllable lu’, but I am more inclined to believe ►y yy to have been the genuine group, which would give us in ^111 Babiru, the same name as the Persian, the final s only of the nominative being omitted, as was the case in Quro, the Median name of Cyrus. MEDIAN SPECIES OF ARROWHEADED WRITING. 301 The next name , A1>ura , in the Median ►^=YY"~ Ssura , or perhaps Asnra, we have already noticed. After that follows Arb(ya, written here »~^YH Abaya, or Arabia, which name perhaps was the Persian and Median appellation of the countries situate west of the desert, rather than that of Arabia proper, as this country, though twice subdued, never submitted to the Persian yoke, or paid any tribute. (Herod. 3, 88. 91.). The next name is in the Persian Ohudraya, accor- ding to Lassen, the Gordysei and Karduchi of the ancients, the modern Kurds. The first group in the Median name is which, on account of its occurring in this name, and of its being used in other words before yD qu, I conjecture to have represented the guttural letter q. The second group, (the last in the twenty-first line) is con- siderably defaced, only ►►► Y being distinctly visible, with some traces of other perpendicular wedges before the last one, making it probable that the group was m, knowrn as as from the proper name of Persian. The two last groups are rdya, and we have then in *V~*~YYY ►^YY*~ Qosraya, a name somewhat different from both the Persian and the Hebrew appellation , though apparently approximating to both. Then follows in the Persian inscription Ariuina, Armenia, written in the Median ^Y^ Aviniya, with the omission of r, with the syllable mi, as in other cases, changed into vi, and w ith the somewhat different termination of iya instead of the Persian a. In the next Median name, corresponding to the Persian KatpaDuka, Cappadocia, the third group was probably V a » ra ther than »~Y ph° > and the fourth apparently »%=Y i which the two middlemost small 302 MEDIAN SPECIES OP ARROWHEADED WRITING. wedges have disappeared , making the name ►f, Khatpathukha, with the palatal letter aspirated as usual. Then follows in the Persian Sparda, which Lassen has most ingeniously identified with the Greek appellation of Sardis, the capital of Lydia, in the country itself perhaps called Svarda, from which both the Greek and the Persian appellation may be derived. The Median name sphada (sfada), stands apparently half way between the supposed form Svarda and the Persian Sparda, which was undoubtedly the Eastern appella- tion not of the capital only, hut of the whole country of Asia Minor between Cappadocia and Ionia ; the latter being easily recognised in the following name Yuna in Persian, and ►tyy ^ Yun, or Yuna in the Median. Having now arrived at the most western part of Asia , we must look in Europe for some at least of the following nations. The first is a Saka tribe, called in the Persian inscription radarya, of which name, however, at the beginning , one or two letters w ould seem to have been obliterated. The corresponding name is , in the Median , composed of eight groups , of which the sixth may be (since only remains) either na, or tliUy the latter being perhaps the more probable, as ►f t precedes. The name, accordingly, is written ►T If the second group found in this place only , did actually re- present the syllable re, or la, we should have in Ares- vitthuven, or Alasvitthuvcn , a name somewhat approxi- mating to the name of the Scythian tribe mentioned by Herodotus (IV, 17), the Halazdnes, which is the only Scythian name that I have been able to find with any resemblance to the sounds of the known groups in the MEDIAN SPECIES OF ARROWHEADED WRITING. 303 name in question. In the following Squdra, we have the Persian appellation of the Scythians , who called them- selves Skoloto. The Median appellation corresponds ex- actly with the Persian , traces enough remaining of the two first groups to show their having been :■ The two last groups give yap, of which the p or b, ac- cording to our before mentioned opinion, is theMedian affix of the plural. The following and last name, Karra, or Karaka, Lassen has identified with KaXavwoj, hut this being only a small district of Assyria, which has already been enu- merated , we may perhaps with more propriety seek the people implied by the name of Karka in the tract lying between the river Araxes and the passes of the Caucasus, which likewise belonged to the Persian empire, though the inhabitants, the Colchians and their neighbours, were not tributaries to the great king, but brought only volun- tary gifts. The Persian name Karka may, I imagine, be identified either with that of Colchia , or perhaps better with that of the present Georgia, which the Russians call Grusia, and the modern Persians Gurjistan, or the country of Gurj , a name that might easily have originated from the ancient appellation of Karka. In the Median inscription this name is written The last group is the Median affix of the plural , the word itself being ►f* The latter group, , klia , corresponds of course to the Persian kd. The first group, ^5. is found in this place only , and cannot have represented either the syllable kar, or kra, since we find similar syllables, such as phetr, phra, dra expressed in a quite different way. The r, it is true, seems occasionally at least to have been omitted before other consonants; still, however, the value of ka for this group would be objectionable for the reason already assigned — that it would be strange, such a syllable should not be found more than once, and that, 1840—1844. 20 (500 MEDIAN SPECIE'S OP ARROWHEADED WRITING. too, in a foreign name. I am, therefore, inclined to think that the vowel has been a little changed , and that we have in the group a representation of the syllable Ku, or rather Kim, partly because the language appears to favour the use of aspirated letters as initials , and partly because the twofold way of writing this syllable Kim seems really to intimate the rarity of its occurrence. If this he the case , we have in Khuklia , instead of Khurklia, a name that stands halfway between the ancient Kcirka and . the modern Gurj. This name closes the enumeration of the nations subject to the Persian rule. It seems strange that neither Syria nor Egypt should he mentioned; the former might, however, as we have already stated , be comprised in the name of Arhaya , and as Egypt is also omitted in the Persian inscription marked I, where a similar list is given of those tributary nations, that worshipped fire , we may in some measure account for its omission, it being well known how much the Per- sians despised and detested the Egyptians because of their base idolatry. I shall now proceed to the examination of an intro- ductory passage common to the seven inscriptions C, D, E, F, K, ]NR, and O, in which I propose to explain each Median word as it occurs, and to collect under one head all the forms of each word , that are to be met with in the inscriptions at my disposal. The Persian passage I insert here , transcribed with Roman characters and with the literal Latin translation of Mr. Lassen. Baga wazarka Auramazda by a imam bufiiim ada hya awam asmanam ada hya martiyam ada hya siyatim ada martiyahya hya Daryawum (or Khsyarsam) khsayii- fiiyam aqunus aiwam paruwnam khsayajiiyain aiwam paruwnam phriimataram. MEDIAN SPECIES OF ARROWHEADEO WRITING. 307 Deus magnus (est) Auramazdes, qui hanc terrain creavit , qui istud caelum creavit , qui mortales creavit , qui fortunam creavit mortalium , qui Darium (or Xerxern) regem fecit , unicum multorum regem , unicum multorum imperatorem. The inscription C has in both places the word narpham instead of Jrhsayaftiyam , and the inscriptions F and K insert the sentence: hya majiista baganam, qui (est) maxi- mus deorum, after the name of the supreme God. Bdyd , God , is rendered in the Median by S^Y in all the inscriptions excepting E, which has »~*~Y J^Y -Y^' On this last form we may build an hypothesis about the value of the group S^Y which usually ends the word. As the word in the inscription E terminates in a hi or pi, it is most probable that the sound p, here expressed, must likewise occur in the other and more usual form , and as it cannot he sought for elsewhere than in the group ►►Y i I am, therefore, of opinion that this group has had the x 7 alue of p, in which case (as the doubling of the letter merely indicates a hard sound) the whole ditference between the two forms anap and ancipi would be that the word which generally termi- nates in a consonant does in the latter form terminate in the vow T el i ; being precisely what happens in the case of the word rendering the Persian nibci , which in the inscription D terminates in n, whereas in the inscription K it terminates in ni. It is true that in the inscription NR 15 ►►Y ^=Y- HTOK, pthri occurs, that is to say, three consonants before a vowel, but as the same word — or one very similar — is in the same inscription (1. 30) written ►►Y *~Y^ *~YYYK? ptiri , it is evident that the Median language itself has sometimes found the other form too hard. The group i^Y occurs, moreover, before •20 30S MEDIAN SPECIES Of ARUOWHEADED WRITING. NR 16; but vve have already adverted to the double way of expressing syllables terminating in the vowel i. Finally, that this group did really possess the value of p seems to he corroborated from its occurring in the word £^Y 5 which renders the Persian gupum, just as the corresponding instrumental case gapwa, is translated ^Y -Y -Y-Y Y^Y 5 tkhaphiwa , the value of the third group being phi, as I shall afterwards endeavour to show. We have already remarked that this group ►►Y seems to have been a plural affix, hut neither from this circum- stance nor from its occurring in four other words, Y^YY? S£f, St* It. ttl 1= ,V Is— f, and ttl ^Y 5 can I with certainty draw any inference as to the value it possesed. The genitive plural corresponding to the Persian baganam is written *~*~Y *~^Y ►►Y ►Y" - -^Y i anappituna , or rather anapituna , F 2 and K 2. In the inscription K the last group is somewhat defaced , but traces enough remain to prove its having been *~^Y? as ‘ n ^ ie °tber inscription. This case ana pi- tuna is formed by adding to the inflective root anap the termination ituna, whereof the last syllable na or n. (as we often find it in the genitive case both singular and plural) undoubtedly is an affix exclusively denoting this case; the first part, or itu, is probably the before men- tioned plural affix tu, joined to the inflective root by the auxiliary vowel i, which we likewise find together with the same affix tu in another plural case in the Median version of the two words liuda bag ibis, which I translate cum diis, as it seems clearly to appear , from the way in which these two words are expressed in the Median, that the first, hdda, is the preposition with — the Sanscrit saha — according to the opinion of Rawlinson, and not median species of arrowheaded writing. 309 the adverb here, as Lassen has conjectured it to be. This word ha da takes its position in the Persian before the word governed, hut after it in the Median, where it is always written ►y i or edaklia , for in the sequel I shall show the probability of the first group having represented a sound similar to that of i. This Median preposition, or particle, edaklia, is employed in a double way, being usually combined with the pure inflective ’form of the plural *~*~Y anapitu, which occurs in the inscription C 21 and 25, and apparently in E also , as the defaced space between and cannot have contained more than two groups , and the inscription K has undoubtedly had the same reading. But in the inscription H this word is joined to — I -: NR 13; the first, as it has reference to the preceding Parsa, must be of the masculine gender; the other, therefore, together with the word Daltyus, to which ') The group yf^; in F can only be a fault of the tran- scriber instead of the regular yY*~. MEDIAN SPECIES OF ARROWHEADED WRITING. 315 it belongs, may be supposed to be of the feminine gender. As the group seems sometimes to have represented a hard and guttural sound, we may perhaps in the affix- wa, or to, which in the Median language denotes the nominative case plural, seek the origin of the affix kh, the Armenian plural termination. The same form occurs, moreover, in H9, being evidently the loca- tive singular, rather than the nominative plural, and lastly the form ►► yf^f, H 8, seems to be either a dative or instrumental case of the prononn set; but of these two forms I shall afterwards speak, when I come to the interpretation of the inscription H. The Persian burnt, the earth, is always rendered by ►Iff which occurs twice in each of the seven inscriptions C, D, E, F, K, NR and O, in the first place in the introductory passage as accusa- tive , corresponding to the Persian bumim , and besides in NR 21 ; secondly, in each inscription and likewise in H 4 as genitive, corresponding to the Persian bumhjd, but without any additional affix or termination. Of the three groups of which this word is composed, the second is written tyj, instead of at the first place in the copy of the inscription C that is found on the high pil- aster , and is perhaps nothing but a slight mistake of the stonecutter in forgetting the last vertical wedge. In the inscription F 17 the second group is written *-yyy instead o f £yyy ; now as the difference here consists chiefly in the absence of the small vertical wedge above the central wedge of the three lower ones , the group »~yyy may very possibly he the result of a similar mistake either of the stonecutter or the transcriber; at all events as the one represents the syllable ra and the other ru, the difference is not very material. The value of the last 316 MEDIAN SPECIES OF ARROWHEADED WRITING. group I cannot presume to fix, its occurrence is confined to this word and to the word y|^§ or . which renders the Persian k/isathram; its use as a sort of interjection before the imperative, has been already noticed. The strange form of this word occurring in the inscription O 16 can only be a fault of the copier , some indistinctness of the inscription at this place may perhaps have misled and perplexed the unskilful and careless transcriber. The Persian add, he created, which occurs four times in the introductory passage we are explaining, is always rendered tusta , iu C, D, E and O; the inscription K has the same, only the groups are a little defaced and indistinct at the two first places. The inscription F likewise exhibits the same word in the first, second and fourth places, using at the third place the word that generally renders the Persian aqunus; lastly, the inscription KR has at the two first places tasta, and at the last -< Fffl -St ^tt S-TT, tuthtusta. The grounds on which my reading of tusta, tasta, and tuththusta (tuthusta) is founded, and the relation in which these three forms stand to each other, 1 have already (p. 281) stated. The demonstrative pronoun , that , is expressed in Persian by awd, and in the Median by -TK -<, yutu, the inscription D 8 only inserting another dental letter before the last syllable , writing K ? or yvthtu. This Median pronoun retains the same form yutu, or yuthtu, in all cases, and takes its place after the noun to which it belongs. It occurs in this introductory passage iu the accusative singular, and in ]NR 13 in the nomina- tive plural, corresponding to the Persian arvdm and awd ; MEDIAN SPECIES OF ARROWHEADED WRITINO. 317 it occurs further in the nominative singular both masculine and neuter, corresponding to the Persian aim, ille, NR 39, and aitd, Mud, NR 16; likewise in the accusative neuter NR 45, and H 23. The Persian preposition ami, wherever it occurs, is rendered by the same two groups, there being no other difference between the pronoun and the preposition than that the latter is placed before the word with which it is combined, as in the Persian; accordingly the Persian words amdqunmcim , or amiiaqunwam , (NR 30), area visdin (D 15, NR 39), awdsiiciyd (C24, D 20), begin in the Median with these two groups -W< ~< > m yutu. It is rather an uncommon coincidence that two so different words should have the same external form in two languages that in all other respects differ so much from each other. The Persian asmdncirn, heaven , is rendered m HfE in the five inscriptions C, D, E, F and NR, and apparently, too, in K; the inscription O adds to the last group , the difference being that the word commonly terminating in kh , terminates here in the syllable kha. The second group m representing a syllable beginning with kh, perhaps kho, as I shall subsequently show, the whole word may be read accordingly akhokh, or akhokha. To the Persian accusative martiydm , man, corre- spond in the six inscriptions the five groups Y2 ^rfy »-3YY*- i anf J m nr the two , Y*H - As we find the last form NR 37 corresponding to the Persian nominative singular martiyd , the first form is clearly to be taken as the accusative plural, the more so as the Persian plural mdrtiyd, the men, is rendered by the four first groups Y3 *~3YY*~: NR 46. We have then Y^ the inflective root, ►fyf *%sYY *~7 318 MEDIAN SPECIES OF ARHOWHEADED WRITING. rar, the affix denoting the plural, and y ff , ra, the affix of the accusative case; the first of these two affixes is clearly found in the Turkish affix Ic'ir , and the other in the modern Persian ra, both being used in these langua- ges for the same purpose as in the Median. We find in the same introductory passage *~^eYY*~ 1 as the genitive plural, (though corresponding to the geni- tive singular mdrtiyahyd , of the man, in the Persian), in which we easily recognise the same plural affix rdr and the genitive termination na, or n, which we have already noticed. As the real genitive singular we find S oJj of tlie pronoun *-£15 sa, and of the noun not hi, we may suppose to have been an affix of the same import as w or na, and denoting the genitive case. The different forms, then, that are met with in the inscrip- tions, of this word ku, king, are: Sing. Nom. ku. Acc. YTT”- 1 1 ku ; »~yyy > kura ; ^YYY *~^Y kuyoni. Gen. ^*“1 ku. Plur. Gen. YYY^" Y^YYY 1 kuthin {or -no'), and yyy- y-y?y -yy? -at-, The Persian aqunus, he made, is commonly rendered hy -yy< ^y yy^y ^»^yy , which is found in B 6, 0(5,23, 1)4,14, 20; E4;F6, 10; NR 5; and also in 09, where the transcriber has merely forgotten the little horizontal wedge before the two vertical ones in the group The same form occurs in the inscription K 7, the only differ- ence being that for the first group we find em- ployed instead of the more common HfT<^ that is , yo instead of yu l . In this form, whether written -yy< ^y tY^y 3*~yy ° r -yyy ^y yY^t *^~yy? an the groups are known, except the penultimate, the value of which I am not able to fix: (yut.-ta or yot-ta). Another form of the same third person -yy< ^y yy^Yi which differs from the preceding in the omission of the last group ^*~yy, is met with in € 19 and NR28, as also in K 20. for in this last place the half effaced traces scarcely admit of any The third group in K 7 is somewhat effaced, the two great vertical wedges only being still visible, but no reasonable doubt 324 MKDIAX SPECIES OF AttROWHRADED WRITING. other form. This last form renders likeAvise the Persian aqunwii, which is the third person plural, NR 1G; for the group which immediately follows at that place, belongs , in my opinion , to the next word in which I recognise the Persian datum ; hut granting even that the ^=»~YY belongs to the preceding verb , it is evident that the same form is used in the third person both of the singular and plural. Another and a longer form seems peculiar to the third person plural, viz. -TT< T 2HN t < m NR 41, which translates the Persian aqunwd, as Lassen has most justly restored the defaced Per- sian groups. In the Median word the second and fourth groups are somewhat mutilated, the former however is easily restorable to or t, and the latter can scarcely be any other than the group Y*~K* which occurs now and then, but of which I am unable to determine the real value. The restored form then of the third person plural is ►Y 3*~YY Y»^Y ) whereof the last two groups Y^eY seem to be the verbal termination. As the first person singular , corresponding to the Persian aqunwdm, I made, we find both *~YYK ►Y *H*"Y Y •> yutta, D 12, E 11, NR 40, and ^YY< -Y ^HfY ^YY~> yuttar, C 22 l , D 13, 19, E 12, and lastly ^YY< ^Y fY-Y* NR 30, which is the same form Ave found employed as the third person both of the singular and plural. Of this perfect tense we find, lastly, the first person plural, corresponding to the Persian aquma tvritten *~YYK ^ Y ZSi ►Y »E*~YY’ *) We may as well remark here that the Persian phrase: vtd tyamiya karlcim utd tydmiyd pithra Ddryaicahus ’ ndrphiihyd kiirtdm is rendered in the Median : et quod feci et quod pater Darius rex fecit. In the same way are rendered the similar phrases in the inscription D and E. MEDIAN SI’ECIKS OK AKROWHKADED WRITING. 325 DIG, where the middlemost group, as the two other in the inflected forms of this verb, is unknown to me. If we consider the six forms under which this perfect tense occurs, it will he evident that *~YYK yut, is the theme or root from which the several persons of this tense are formed in different ways, hut chiefly hy the affixes and employed separately, or conjointly. We have thus : 3 sing, and yut - yf^Y - ^(Y (or ^fff ^ rYMf jHf)^ I sing, yut - yfirf , and yut 0I '~3^YY 3 plur. yut-tf^, and yut Y^K Y^Y* 1 plur. 2 /m/~^ 5Y ►Y »£*^YY 5 and in this person the last ►Y > s undoubtedly 7 nothing but an orthographical sign, to mark the hard sound of the following consonant. We may' here add that we meet, no doubt, with the same root *~YYK ►Y’ V ut » ' n ^ ,e nou n *"YYK -Y lY^-? yutro, or ~ff< sf ,T= -I yutrot, corresponding to the Persian kart a , which seems to have been formed by the affix yY^j ro, or yY^ ►Y) roif > from this root. The Persian airvcim, one, the only one, which occurs twice in each of the seven inscriptions , is always ren- dered hy d 5 ,nr 6 ,k 8 , 9 .» -m ^E or Fm yy *-3YY*~ *~^Y MY^ j phrumattaranm. We shall , therefore , here only try to explain the new word mm <- my- >yy m ^yy~, which the inscription NR exhibits corresponding to the Persian phrfimutdrdm. The three first groups MM MY^- correspond likewise to the Persian phramana NR 47, and it is therefore evident that they must compose a word conveying the fundamental notion of both Persian words, that of disposing, commanding. Setting out from this. the word mm <- my- :hy m m- is not difficult to explain. It is a compound word , of which the last part ^5 MY M M^- *~^YY*~* dattir, or dat.ir, has undoubtedly the sense of, he who holds, who possesses ; the w hole word being of a formation analogous to the modern Persian jL\JUw> firmnndhr etc. , and the sense of the whole word would then be , who holds a command, a commander, a ruler, like the sense of the ancient Persian plirumatnram. In the first part of the compound word MM MY^- I hod 0 word, 328 median species of arkowheaded writing. that stands in etymological connection both with the ancient Persian phrdmdnd and the modern firman. It is true that the Persian syllables phrama are written phr&vaa , but as this is a mere tran- script in Median characters of a Persian word, we may safely suppose that the same word might have taken a somewhat different form in the language itself. We might then consider the two last groups nim, as a sort of Persian accusative of a word that terminates in n, or ni, or perhaps indifferently in either, like the two words we have already mentioned ( , or -Hf <1 EEf a„d which E8 and apparently too F 18 translate the Persian apit/a, and which being read ophi, would present us, in a slightly altered shape , exactly the same word , as the Persian apiyii , that might also have been pronounced 8 Anap rasarar Aera- ZDA, KKHA SA Ore. TE- STA, KKHA AKHOKH YE- TU TESTA, KKHA WoTHIRA RRA TESTA, KKHA SI- YATIM TESTA WoTHI- RARAN, KKHA DaRIYA- wes (or Khsarasa) Ke- RA YET. DA KHORA RA- SAKHOTHIN KE, KHO- RA RASAKHOTHIN PH- RAWATARAM. Various readings. *) ppi, E. — 2 ) K and F. add if # v -m -sir - 8 -:r "T -t- -< -:r, rasarar anappiten. — 3 ) NR omits it; D, E, F place it after the next word. — 4 ) NR ht n nr , TASTA, and likewise in the next line. — 5 ) *-^3 Hfj khkha, O. — 6 ) Hff thte, D. — 7 ) NR omits the three groups -yyy ^TT?i barra. — 8 ) NR Y^TTT i5Tr 1 teththesta, and likewise in the next line. — 9 ) D, E, O. — 10 ) om. D. — n ) V, ssa, D, K. — l2 ) -yyy, yo, K. — 13 ) C and O insert at these four places y. the mark of distinction. — 14 ) ppu To, na > or n > joined to the inflective base by the group y, the value of which I am unable to determine. We have thus only two forms of this MEDIAN SPECIES OF ARROWHEADED WRITING. 333 Median pronoun of the first person, viz. used both as nominative, accusative and genitive (and perhaps also dative) and — ~YYY employed only in the geni- tive case. The following words we have already explained, and have likewise deciphered (p. ‘280) the word that renders the Persian dahyu, occurring in this passage in the geni- tive plural ddhyunam ; therefore, after observing that the Median word is always separated or distinguished by the mark f, (except in the inscription E, which uses the other mark of distinction, *-) we will here only notice the diffe- rent forms, that occur of the Median word. Sing. Acc. ^»~YY yy K ^5YY i dahyus, NR 34. yy *~YYK ^5YY 7 dahyus, D 12, as the last part of a compound adjective. Plur.Nom. ^E^YY yt Vt K »5YY i dahyus, NR 13/38. Gen. ^=*~YY 11 ^YY ^5 dahyunam, D7. ^-YY 11 HflK ^5YY dahyustun, C 11. ^-YY 11 ^YYY ^5YY *~^Y» dahyostun, B 3, C 11, E 7, F 15, H 3, K 12, O 15. ^-YY 11 "YYY i5YY dahyostu, NR 10. The adjective belonging to this word, written in the Persian inscriptions pdruzunanam , or pcifuivzdndndm , is in the Median merely transcribed by ^YYY 11 *%-! -^Y ►YY^i paruzananam , F 16 (where by a mistake of the transcriber the one *~^Y * s omitted), or, more com- monly, — Y -YU tt *~»-Y *~Jf — YY— : • D 8, E 7, 0 15, with the aspirated pha instead of the unaspirated pa. The 334 MEDIAN SPECIES OF ARROWHEADED WRITING. inscription K 12 has £yyy j{ *~^l 7 which is the inflective base £yyy |( pnr«- zana, joined to the Median termination -< -3f, stun, or stuna. In analogy with this form the inscription B 3 has TT ®w2»a- zanastun, of which compound vispa (all) forms the first part, and to zann is joined the termination stun, or stuna, differing from the other genitive form only in the s not being aspirated; and such a vacillation in the writing may have originated from a similar vacillation in the pronun- ciation. Instead of pdru-zdndndm the inscription NR has, as mentioned above, vispa- zdndndm, which in the Median corresponds to V »-^y , vissadanastun. We here meet the same termina- tion i stun, or stuna, hut joined to a somewhat different word which I cannot fully explain. It may be, however, that the first part Y , vissd, (wha tis not at all unlikely) is either a fault of the stone- cutter, instead of vispa, or a Median assimilated form of the Persian word, and the second part s-n ~zi dana, may be the proper Median form of Persian zanii, with a transition of z to d, of which we shall show one probable instance more in the sequel, and according to this supposition we should in the Median vissd-dana, or perhaps vispa-dana , have a word of the same import as the Persian vispd-zdnd. But the word dana is also used in the inscription C 12 , where the Persian pdruwzdnandm is rendered bv | *~^yy which renders the Persian nibdm, farther in the termination of the imperative in the verbs *~Tf and of which we shall speak in the sequel; lastly in NR 28 at the end of a word, and H 23 in the middle of another. The Persian epithet dpiyd is left out in all the Median inscriptions except in the three NR, E and F. In the two first it is rendered which we have read above ophi and compared with the Persian dpiyd. The transcript of F 18 has tf, but it is not at all improbable, considering the carelessness with which this transcript has been taken, and of which we have already shown instances enough, that the tran- scriber has overlooked the wedges ►f, which of a t, would make a phi, and then we should have the same word in all three inscriptions. The Persian putlird, son, is rendered v -s- -m<> sakri, or sagri in B 4, D 9, G3, K 15, O 19, NR 10, 1 1. The one copy of C 15 has the same, but the other leaves out the final group having Y only; the 1810 - 1844 . 22 338 MEDIAN SPECIES OF ARROWHEADED WRITING. same is (lone by both copies of E, although in the one (that found on the staircase) the ~ttt< 5 as before obser- ved (p. 285), seems to have strayed into the following word Okkhavenisiya. The inscription F 20 has -m< only ; hut the transcriber has there undoubtedly over- looked the wedges j , as also the mark of distinction f, that precedes this word in all the inscriptions with the sole exception of NR. We may as well observe that the word V *~fff^: sakri, or sagri, seems to lie etymologically connected with the Mongol or/hul, son. The whole passage which we have just explained , containing the titles of the great kings Darius and Xerxes, we here insert , as we did the preceding one , in the Median character , and transcribed with Roman letters. It runs thus in the inscriptions of Darius: i =m * i ht -m< E-A tst < zw s r m= * -w V -w -Iff- * T TTT - * T TTT- f-fff -m -Iff- • * f fff ; = *.f HT Y TT -TTT ^TT - ■< -I r * f -f H -If ' -IT -TT- " * f ffF * - ~ -rrr -YY ► ► iTS 3 flf »- -I T !! -ffl ►w-i * ^Y 5 YY -I -Y T HT -t? 6 -T * Ht -f-f ’M I <- pa, instead of — f, pha, F . - 4 ) C y -r? ~ -m RA , F (•)• — 6 ) Thus D, E, F ; K has ]]Z ]] iy, ozakha, and C -"ffy V *^T?T RAsarar. — 7 ) -Y -|t, tti, C; ^y -yr, tti, E. — 8 ) RA, F (?) — y ) C, D, K omit this word. — 10 ) Y sag, E and the one copy of C. — ll ) D, F, K omit the group -i- - ,2 ) i which iv e read nnnri , or perhaps ntiri. It is placed, as here, before the name of Xerxes in all the inscrip- tions of that king (C 15, D 10, 17, K 10), but after the word in thosegof Darius (H7,19; NR 12, 26, 39); MKDIAN SPKCIKS OK A RHOWMKA DKD WHITIXO. 343 the place of the corresponding word being in the Persian inscriptions always before the proper name of the king. The following three words are: Xerxes rex magnus. To the Persian masnii, bv the grace, or, according to the will, correspond the four groups yy ^ which are likewise found C17, Dll, 10, H 10, XII 12, 29 \ 40. It would seem that a slight error has been committed in the transcript of the inscription K in omit- ting the angle ^ forming the second group ; but even supposing it to be wanting in the inscription itself, the omission _may in my opinion be defended , as I find in the two first groups }} ^ the modern Persian preposition, J»f or j of; the other two groups, then, which we read vii, or rather viyi, would constitute the word that properly corresponds to the Persian wasnr'i, and per- haps even stands in etymological connection therewith. Then follow: Auramazdis hanc. The Persian hrichis, which occurs in this inscription and in C, Mr. Lassen has ren- dered, aula columnaris, a pillared hall, and that translation at least agrees exactly with the ruins of the building N, in which the inscription E is found, and which, therefore, must have been the lutchis that Xerxes erected. Here the bases of thirty six columns still remain in the great ball, and in the adjoining looms traces of a double row of pillars. On the other building K , which bears the inscription C, and which, therefore, was the Art this built by Darius, the father of Xerxes, the rubbish has accumu- lated to such a degree as to conceal every vestige of a column. But as the Median word (»- that renders the Persian hdchis, translated also the Persian At Hus place the first group, yy. is partially defaced, the last y half y only being left. 344 MEDIAN SPECIES OF ARROWHEADED WRITING. vip , court people , or courtiers , I am more inclined to attribute to the Median word the signification first of the court, or courtiers, and secondly of the place where the court was held, the Durbar of the Eastern princes, and farther to give the latter acceptation to the Persian hcichis. The corresponding Median word here E 19 (or 10), and C 17, is preceded by the mark of distinction and con- sists of the three groups °f which the two last give sathi, the first group being unknown, nor am 1 able to fix its real value, though it is also found NR 14, 28, 33. Then follows: Ego erexi. Me Auramazdes o ! The Persian imperative paSuma, tuere , is rendered by the five groups <► iSfi iiTT ^yy in the twentieth (or eleventh) line of this inscription, and like- wise C 20, 25, D 18,20, H 20 and NR 42. Of ^yy we have already observed that we are unahle to fix the value; the same is the case with the middlemost group -n-E found , besides in this word , likewise in two others NR 29, and f-.., NR 14, in both places somewhat defaced , as in the former the foremost horizontal wedge, and in the latter the middle- most has disappeared. The word rendering the Persian barjibis has sustained some damage in all the copies of the inscription E, the three groups i anap , only remaining , and in one copy a fragment of the fourth after which the fifth has totally disappeared in all; but in none of the copies does there seem to be space for any more groups, and hence we may conclude that the inscription E exhibited the same form anappitu (anapitu), that we find in the inscription C and MEDIAN SPECIES OF ARKOWHEADED WRITING- 345 have explained above (p. 308). Then follow in the Median the three groups if, composing the word which in our opinion corresponds to the Persian h add. As the first group ►£:, from its occurring NR 35 in a word which in the thirty -seventh line is written with an i, appears to be a vowel of nearly the same sound, I suppose it to have had the value of an e, and the word ►f is then to be read edaklta, of which the two first syllables may be compared with the Persian word huda. Its place is always (C 21, 25 and H 13, 20), as here, after the noun governed. The Persian utdmiyd khsdtkram is rendered here by rTS cT 3HN - ai sii -E r_, by (13 M - SIT cl? (13 iE <£c, D la The three first groups its cf 3-n, qutta, or quta, constitute the Median word that renders the Persian utd , et , and is found, besides in this inscription, in C21,22; D 14, 18, 19; H15,17; NR 23, and no doubt likewise in the half defaced wedges y|^= K20. The last of the groups in D, vi, seems to denote the demon- strative pronoun, as it is used, besides, NR43, in a place where there is no other group that can be thought to mark the pronoun which is found in the corresponding Persian passage. The same therefore must be likewise the case with the f*-, z, of E, though they were perhaps both and ra ther pronominal affixes than real pronouns. Corresponding then to the Persian khsathram we have in the inscription E the three groups sn ►!! - and in D the four ►{{ y|3 s0 that the word contains in the latter inscription one syllable (13 > qu , more than the other form. The first group , 340 MEDIAN SPECIES OF ARROWHEADED WRITING. which we meet here for the first time is also found in the word »eTY but as to its probable value I can form no opinion. The word ►Y following then next after ►Y quttn (quta), is the Median relative pronoun. Of the two groups, the first, ►►Yi bas the value of p; the second Y^ therefore, which is likewise found D 14, NR 34, and perhaps also in the indistinct and defaced traces Y^“ and Yi NR 44 and 36, may be supposed to represent a syllable beginning with p, which could only be po or pu, as we have already pointed out signs for the other syl- lables beginning with p. The relative pronoun, then, would be read ppo , or rather po, the first group being merely an orthographical sign to mark the hard sound of the following consonant. This pronoun always preserves the same form wherever it is used. It occurs as the nominative singular neuter, K 14, and farther NR 16, in the phrase: datum quod mihi (est), id possessum fuit; more frequently, NR 30 and D 13, 14, as the accusative singular, corresponding to the Persian tyam and tya, but in the Median the gender cannot with certainty he determined. As an undoubted neuter in the accusative case it occurs both E 23 (12) and C21,22, D 19 where, as we have already observed, the Median has a whole sentence: quod erexi, or quod pater erexit, instead of the Persian, meum palatium., or illud patris palatium, AVe find the same pronoun ►►Y Y^: employed as Hie nominative plural masculine NR 46 and 23, corresponding to the Persian hya and tyiyt'i, and likewise NR 15 and 20, in which places the relaHve pronoun is left out in the Persian. The pro- noun ►►Y Y- occur * farther as the nominative plural NR 32, and as the accusative plural NR 13, agreeing in JIKDIAX SPKCIRS OK AHROWIIKADKD WRITING. 347 Loth places with the preceding dalujus. The occurrence of this pronoun ►► is in all these places quite analogous to that of the other relative pronoun which we have above explained; and from this we draw the inference, that these two pronouns kha (kkhrt) and po (ppo) , were used to supply and complete each other, in such a] way, that ►f, was used in the nominative , and perhaps , too , in the accusative case of the singular masculine only, and in the other cases of the masculine and the other genders. This pronoun is farther found XR39, and likewise K19 and jNR 47, placed between a noun and its adjective, in all places corresponding to the Persian tyd and tyihn, hut apparently used more as the demonstrative than as the relative pronoun. Lastly these two groups are found H 23 , forming, as it would seem, a conjunction with the sense of the Latin quod, that. The last word in the Median inscription E -yy< ►Y yattar , we have explained above and translated, I made; and it is thus only in the last sentence that the Median inscription differs from the corre- sponding Persian, having the words „and which I made” instead of the Persian: l4 and this palace”. We conclude with the insertion of the Median lines which we have explained, transcribed, as usual, in the Latin characters. 17. XAARI KH 18. SARASA Ku RASARAR ZU VI- 19. yi Aurazdax SA .SA 20. THI Yo YUTTA Yo AoRA- 21. ZDA . MS.S. AXAP- 22. ' PITU KDAKHA QUTTA . . 23. fJZ qUTTA PPO YUTTA II 318 MEDIAN SPECIES OF ARROWHEADED WRITING. THE INSCRIPTION A. The Persian inscription A is found on the Northern side of the large terrace bearing the great colonnade, at the western flight of steps leading up to it, between the triple row of figures bringing gifts and presents to the king and the representation of the fight between the unicorn bull and lion. The Persian inscription contains thirty lines, of which the five first are broken off together with the parapet wall along the terrace. It is in substance nearly the same as the preceding inscription E, the whole difference being that instead of the phrase which is found in E after the word wazarka : 23. wasna Auramaz- 24. dahii ima hachis adam 25. aqunwam. fhe inscription A exhibits : 23. tya mana kartam 24. ida uta tjamiya 25. apataram kartam awa v- 26. isam wasna Auramazdaha 27. aqunwam. which Mr. Lassen has translated : 23 hoc mihi palatium 24 heic, turn hoc 25 alterum palatium ad commorandum 26 e voluntate Auramazdis 27 exstruxi. The corresponding Median inscription has totally dis- appeared together with that in the third species of Achae- menian cuneiform writing. THE INSCRIPTION C. This inscription is preserved in two copies differing from each other only in the number and length of the lines. They are found in the building K on the plan of Kerr Porter, the one on the southern staircase, and the other on the high pilaster at the top of the steps. It begins, like the other inscriptions, with the introductory passage MKMAN SPECIES OP ARROWHEADED WRITING. 349 and that stating the royal titles , and then proceeds in the Persian : 16. Jiatiya Khsyarsa narphii wazar- 17. ka wasna Aurahya mSzdaha 18. imi hachis Daryawus narphii 19. aqunus, hyii miina pit- 20. a. mam Auriimazda pa5u- 21. wa hada biigibis uta t- 22. yamiya kartam uta tyami- 23. ya pithrii Daryawahus narphahy- 24. a kartam aivasaciya Auramiiz- 25. da pa&uwii hada bagibis. The Latin translation of this passage is : 16. Generosus (sum) Xerxes, rex magnus, 17- e voluntate Auramazdis 18. hanc aulam columnarem Darius rex 19. erexit, qui meus pater (fuit). 20. Me Auramazdes tuere 21. cum diis, turn hoc 22. palatium turn hoc 23. patris Darii regis 24. palatium. O propitiande Aura- 25. mazdes tuere cum diis. To this corresponds in the Median : n. !! < * * Hf < -3T- L HT 15 . * -e * - <-r= -e is * t nr -m< E=r'> rsr < nr * r m= * -tk -r 18 . 350 MEDIAN SPECIES OF ARROWHEADED WRITING. 19. yH ^ T •> -T * T * T =ST S-ff s-rr * r =m * ~y 20. < r_ nr m -rrr * =i! * <- ^rr -tt-e ^rr -n * 21. Hf O | H =T- -< * E= HT -T * iTS cT S-ff * ~T 22. r= * t| |-ff -Iff- * tTS =T s-tt * ~t r= * r -sr 23. TT * T S-TT -TTT< E-u TST < =~EE YSf * ~Y Y= * Y -YYT * -YY< -T S-YY 14. * tYS -Y S-YY * ~T Y- * Y -SY S-YY ht* -yj< A AA S- -YY * ~1 1- V Y : -Y * 15. -yy< = r -Y * EYYY E=u YSY - j- * 1 = 19. i- * t =m * -n< -i i-r f Hh * i ] 12 =r i-n * "t r= * i -i f hi > n * 20. -n< -i n-r i-n * -rr< -< ; HI * Hf < sn- r- i-n • <- ^n -ns gn -n * We shall, in the interpretation of these Median pas- sages, as we have done before, translate the words already known and explained by the corresponding Latin terms, dwelling only on such as we meet here for the first time. Line 10. Generosus (sum) Xerxes; L. 11. rex. Ex voluntate Auramazdis hanc. The Persian thurvarpim, a portal, is rendered by the word preceded by the mark of distinction Both of the groups that compose this word are unknown. The first Syy , which we have found as a dubious lection of in one of the copies of C , occurs likewise in two other words i-n tfti a„,i <=r= -e 12 -A m the genitive case of the noun > and ^ ie other T^TT which is found in this place only, is per- haps a mistake in my transcript instead of which we shall find of more frequent occurrence, though its value is still quite unknown to me. The word visd- diihyum following after thuredrpim, is rendered V TT •> vtisadaRyus , which is clearly the Persian word, only transcribed in the Median character and adopted in the Median language, the nature of which undoubtedly accounts for the retention of the Persian nominative termination, and for the change of s into the aspirate s (or written according to the Median ortho- 1840— 1844. * 23 354 MRDIAX SPRCIRS OF ARROW HR A DEI) WRITIXO. graphy ss). Tlie next words are: ego extruxi majora, the Persian jpdsiyd being, as before mentioned, here and K19, rendered by the word »-yyy ■> rasakho, which likewise translates the Persian parti. The following Persian auiydsdciyn and the corresponding Median yy £yy ^f^y , (fah.phi , are undoubtedly, like wasiya and rasakho, both adjectives, belonging to the next word, in the Persian nibdm, which Mr. Lassen has translated propykea, and which here 1. 12 and 15. is rendered in the Median by 5 sis.n , but K 20 by iya, hya maria pita, huwa wa- if). sna Auramazdaha wasiya tya 20. niham aqunus uta ima st- 21. anam huwa niyastaya. Mr. Schulz’s copy has kutundm ( ►Tlf iTf ^“YyY 3? which Mr. Lassen, adopting a conjecture of mine based on the corresponding Median word , has corrected to Y3; ►Iff Iff stantim, the whole difference between a Persian k and s being that the former has two and the latter 0E> *) ti> ree horizontal wedges after the first vertical one. The Latin translation of these lines is : 16. Generosus (sum) Xerxes 17. rex. Darius rex, 18. qui meus pater, ipse e vo- 19. luntate Auramazdis majora haec 20. propyliea exstruxit, turn hunc lo- 21. cum ipse inhabitavit. The corresponding Median passage I here insert from the copy of Mr. Schultz, only separating the words by an asterisk, and restoring a few groups that are a little 358 MEDIAN SPECIES OK ARROW HEADED WRITING. defaced and indistinct in the copy, but which have retained traces enough to show beyond a doubt their original form. The Median passage is: 16. -:y - y -yy< * i f -m ~ -m ^yy v * 17. 1 w * y ht J ryrc rsr < ^yy * y yyy- * ■> 18. ►T * f : TTT * y =3 ny ht * r -n< ^ -yyy< * tt Tf \-~ 19. y < -ar- - y. ny -:y * 1. 23, are apparently the fragments of the first personal pronoun ; lastly the nyrim , 1. 24 , is conjectured after grammatical rules in accordance with the niycistdya, 1. 21. The transcript of the corresponding Median passage is still more defective. The three first groups compose in my opinion the word that renders the Persian kiitd- niyii, and supposing this to he etymologically connected with the modern kandan , I attribute to it the sense of the cutter, a stonecutter, nr artist, and this acceptation would show the reason why the corresponding Median word is preceded hy the mark of distinction )f. The three Median groups, {{, give satza, or rather sadza, but I do not hesitate to correct the ►Y to > , the difference between these two groups being so trifling as easily to escape the notice of even the most skilful transcriber, especially when, as was actually the case, the inscription can only be copied at a great distance, and in a dazzling sunshine. We have then ►► jy , sakhaza , in which I find the two modern Persian words, sung , a stone , and ziidiin , to cut , so that the literal sense would be the stonecutter. The rest of the Median passage I leave unexplained , which I think is more ad- visable than running from conjecture to conjecture in the wide field of hypothesis. I only mention that we find in these lines two new characters and that are met with nowhere else, and therefore it is very dubious whether they are to he enumerated among the real charac- ters. We find farther twice the group ^Yi which as stated above (p. 300) is perhaps also to be found as the third group in the name of Babirus corresponding to the Persian ru. The four last lines of the inscription in the third species coinciding exactly with the final pas- MEDIAN SPECIES OF ARROWHEADED WRITING. 361 sage of the inscription E, makes it evident that the two other inscriptions were concluded by the same passage, containing an invocation to the supreme god Auramazda. In the Persian inscription, however, no traces are left, and in the Median all that is distinguishable is the initial y -yyy *~*~y k c° : *~^yy*~) •> yo Aum (zda), in the twenty fourth line, and farther the final groups of the two next lines, of which the former, na, must have been the second group of the word ►►f? Y Y anap, god, and the latter, y can scarcely be any other than a fragment of the £yy, which occupies the second place in the word a kingdom; the rest is wholly obliterated. The inscriptions of Xerxes, which we have now passed in review, are on the whole of a meager and poor tenor. We learn from them the pompous titles which he had inherited from his father, and something about the part which he had taken in the building of the royal palace of Persepolis, but nothing as to his own history, or the internal state of his empire. On the other hand , the inscriptions of his father Darius , which we shall now proceed to explain, are of a more interesting character both in a historical and philological point of view. THE INSCRIPTIONS OF DARIUS. The Persian inscriptions which we have relating to this king are equal in number to those we possess of Xerxes. Four of them (B, H, I and L) are found among the ruins of Persepolis; a fifth (NR) on the royal tomb oppo- site to the palace; the sixth (O) at mount Alwand close to Hamadan ; the seventh, the greatest of all, is found at Bisitun , but except the beginning that has been tran- scribed by Mr. Rawlinson , this inscription is still left MEDIAN SPECIES OF ARKO W H E A UK D WRITING. 362 uncopied. Of those only the four B, L, Ml and O, are found in a corresponding Median translation; the other two H and I, being accompanied by a Median inscription (marked H) of an original and quite different tenor. THE I INSCRIPTION B. This is found on the pilasters of the doors in the building K, and, like the inscription (!, over the figure representing the king. The inscription consists of six lines in each of the three species, of which the Persian is: Daryawus khsayajiiya waziirka, khsayapiya khsa- yafuyanam, khsayafiiya dahvuiiam, Vistaspiihy- a puthra, Hiikhamanisiya , h- ya imam tiicaram aqunus. Darius rex 2 niagnus, rex 3 regum, rex 4 regionum, Hystaspis 5 filius Achremenius, qui 6 hanc aedem aediticavit. TIip Median inscription is as follows: 1- r h? -m< Em\ m < =r=r ht * 1. Dariyawus Ku rasa- 4. stun, Vistaspa sag- 2. bar, Ku Kuthin Ku 5. ri Okkhavenisiya, kkha 3. D YHYOSTUN Vispazana- 6. sa Tadzaram yut.da We have in the foregoing pages explained all the words here occurring: The Median inscription differs from the Persian only in the addition to dally ostun , which translates the Persian diilojunnin , of the adjective visyu - MEDIAN SPECIES OK ARROWHEADBD WHITING. 303 zamsfun, to which nothing corresponding is found in the Persian, hut which renders the expression in the Persian INR, vispuzdnduum , omnigenis populis hajdtatarum. THE INSCRIPTION L. This is met with in the same building K, as the inscription B. It is found on all the windows and niches in the great hall, and consists in each species ol one long line placed so that the Persian occupies the space above, the Median running up on the left side, and the inscription in the third species running down on the right side ol the Persian, which last, transcribed in Roman characters is as follows: Arda«tana ajiagina Daryawiihus narphahya vijiiya karta. Alta (h.x*c) arx (est) Darii regis geutis palatium. Which in the Median is rendered thus: -m h» Ht -:r * - -m n * y ht -m< UK^n-'i w * - <=f= -E Tl TIT * -TK -T rT" =T * Thp two first rvorrls ~£f£ jiTT HTT ~Z1 ► • ddastana Asina, have appa- rently been adopted from the Persian , and altered con- formably to the nature of the Median language. To the Persian genitive viphjn correspond the five groups pre- ceded by the mark of distinction , *~£e Ul and as the Persian accusative of the same word, vipdm, is rendered by the three first groups preceded by the same mark of distinction, »- ^^^E ►► it is evident that the two last groups of the former Syy must denote an affix of the genitive case, and the three first ^i^y^ ►► y^ compose the word that is equivalent to the two Persian words vif) and /ific/ti# , and which, therefore, must have the signification, first, of the court, 364 MEDIAN SPECIES OF ARROWHEADED WRITING. and next, of the place where the court isheld, the durbar as we have above translated the Persian huchis. The last word , i yutrot , we have already explained. The whole Median inscription then is: Adastana Asina Dariyawus Ku .sathi yutrot THE INSCRIPTION 0. The transcript of this inscription which is met with at mount Alwand , w'as , like that of the inscription F, found amongst the papers of the late Mr. Schultz, and has been published by the Asiatic Society of Paris. The inscription consists of twenty lines in each of the three species, and contains only the introductory passage, with that of the titles, which we have explained above, at the same time correcting the faults with which this transcript ot the Median inscription abounds. THE INSCRIPTION NR. This inscription is the largest of those to which I have had access. It is found on the outer face of the tomb of Darius (which is excavated in the mountain on the opposite shore of the Pulwar river, over against the ruins of Persepolis), above the Sassanian sculptures now called by the natives Xakshi Rustam , or the sculptures of Rustam, the great hero of the Persian fabulous history. Of the inscription marked 3NR the Persian version, con- taining sixty lines, is placed to the left behind the figure of the priest in the act of praying, and the Median, con- taining forty eight lines, to the left of the Persian; the version in the third species , which contains only thirty six lines, occupying the space on the left lateral wall at right angle with the front wall of the tomb. The Median inscription, which is lithographed Tab. XII, is of exactly the same tenor as the Persian, they being properly speaking mere translations the one of the other. They both begin with the two oft repeated passages, the MRDI A X SPRCIES OF ARROWHEADKD WRITING. 3G5 introductory one and that of the titles , which we shall pass over here , referring the reader to our explana- tions above, where we have also noticed the various lec- tions exhibited by this inscription. After the word Hakhd- mcinidiya the Persian proceeds: Parsii Arsahya puthra ariya ariya dathra 1 ; of which • 1 consider Pdrsu- Arsahya and farther ariya-ariya-duthrci as compound words of the species called karmadharayd in the Indian grammar, and translate, a son of Arsa, the Persian, of an Arian and an Arian ancestor, that is, of the one Arian ancestor after the other. Arsa 1 am inclined to consider as the same with the Arsaniii of the inscription P and that of Bisitun , and to regard the ariya as having been an honorary title of the Persians also, although Herodotus restricts it to the Medes, giving to the Persians that of arta, the sense of which does not, in fact, differ much from that of ariya, the former being a participle of the preterite, bonoratus, and the latter a participle of the future, honorandus. To those six words correspond in the Median tiie two last groups in the tenth and all but the last group in the eleventh line , viz : t a * t -&-m * v -2- ~m< * T ~EtE HM< E=,\ * 1 ~ETE -W< E=,\ * Em v which I read : Phaasa Asa sagrt Ariya Ariya .ssa of which we have in Bm^n Y the Median word that corresponds to the Persian genitive dathrd , of an ancestor. The first group, as has been already mentioned, I am unable to decipher ; the second, or is appar- ently inserted here for the same purpose as in the words ') The Italics indicate the groups conjectured and restored. 360 11 K D I A \ SPKCIKS OK ». RD OWHK A DU I) WRlTIVfi. Khsarassa, vissa , to mark the hard sound of Ihe follow- V V sibilant y , .VO. Then follows in the Persian inscription : 15. jiatiyii Daryawus khsaiyo- 16. jiiya ivasna Aurainazdaha into 17. dahyawa tya a dam agrabayam 18. apataram hiica Parsa adani sa 16. pmiyakhsiya mana bajini abiiro 20. hamyasam hacam a{)a lidm a- 21. (junvva: datiim tya mana a it a . • 22. adari. I differ from Lassen only in restoring apaliam a at the end of the twentieth line , instead of his apdf/indm a , and likewise a it a in the twenty first line , for which 1 shall presently assign my reasons. To this Persian passage corresponds the Median one that begins with the last group in the eleventh line and terminates with the first in the seventeenth. The last group in the eleventh line , is , as shown in the plate, somewhat defaced, only the lower part yy remaining tole- rably distinct, and the upper part being traceable to which would make the group yy, h , or hd. Thi s joined to the first group in the next line , forms a word yy /ink, which is found distinctly written I. 25 and likewise H 6 and 18 in the same manner as here before the name of Darius. Therefore, though the word is left out in the corresponding Persian phrase, 1 suppose it to have been a demonstrative pronoun akin to the other, sn, which we have explained above. This pronoun y*y occurs farther H 21 and 15, apparently as the accusative, and H 12 probably as the adverb ideo , of which I shall speak more fully when explaining the in- scription II. The following ten words: Dariyawus Ku V K MAX SPKCIKS OF A ttllOW H R A OKI) WRITIXO. 367 naari zu viyi Aurazdan sa dahyus ppo Yo, Darius rex j„ generosus; e voluntate Auramazdis ilke regiones quas ego, having l)een already explained , we will only observe that the Median r — r -m< 5 naari, rendering the Persian pntiyu, is here and likewise NR 26, 39 and H 7, 19, placed after the word , ku , rex , hut occupies in the inscriptions of Xerxes the same place as the Per- sian word before the proper name of the king. The last six groups in the thirteenth and the two first in the fourteenth line Sfl- HW< V V T- * compose the two words that render the Persian ayrdbaydm ay afar am , hut which I am unable to separate with any certainty, for it is scarcely more than a loose guess, when I divide the groups into =YY- HNf< Y»- V a,,( l U ►f and suppose the three last to compose a word , ziirdkha , etymologically connected with the modern Persian digdr, having the same sense ! that is ascribed to the ancient apdtdrdm. The word then that renders the Persian ayrdbaydm would be formed by the five groups *~^YY*~ Y*“ V i °f which the found before \\ is perhaps inserted to show that the following consonant is the soft aspirate, z, that corres- ponds to the hard s. True it is we should expect, that as the alphabet has distinguished between the hard s and the soft 2 by different signs , it would have done the same with the aspirate s and z , hut as the contrary is analogous to the common way of expressing the hard and soft consonants hv the same sign, and as farther we occasionally find the group inserted before Y Y and y , which would lie quite superfluous did those groups represent only the hard syllables sa and si, we may, I think, venture to suppose that these also represented .'568 MRDIAN SPKCIES OF ARHOWHRADED WRITING. syllables beginning with z , and that the , z , was inserted for the purpose of marking clearly and distinctly the soft , and 5 the hard sound of these syllables. The word would then he read berirzza, or berirza, which would seem to he from the same root as the beris/a, NR 33, and beris, NR 16 (that correspond to the Persian addrdyd, and addri), and perhaps, too, as the berida, which we found above rendering the Persian visum. The groups f ^ ^ following after must be restored to r Hf «< according to 1. 38, and give us the word that corresponds , as above stated , to the Persian Pdrsdi , which is placed after lined , whilst the Median occupies the place before the corresponding word klikhabe, and this word rendering likewise the hdcam in the twentieth line of the Persian inscription , proves that the two w ords lined and hdcam must he of the same import , which from the context in both places, seems rather to have been that of help, assi- stance , than that of adoration which Mr. Lassen has ascribed to the accusative hdcam’, for the sense implied in the first phrase seems to he, as has been conjectured by Lassen, that Darius together with his Persians con- quered the countries, or, according to our supposition as to the signification of the words luied and klikhabe , literally: These (are the) countries, which I conquered , and the Persian assistance; that is, which I conquered with the assistance of the Persians. The Persian Pdrsd must therefore he taken not as the nominative plural Persic, but as the adjective feminine Persica, belonging to hacd; which likewise must he the case with the Median , Phas'd. The two following Persian sentences: adiim sapmiyii- khsiya , mana bajim abarii, are rendered by the groups MEDIAN SPECIES OF ARROWHEADED WRITING. 369 from y -m in the fourteenth to z?fl in the fifteenth line. We have f Yo, corresponding to the Per- sian ndam , and y -yyy ^TT to miind as before mentioned ; the three groups ^ nr- ^yy following after this word corres- pond apparently to the Persian abdrd, as we find a similar form yy^ ^y ya-., qutrea.., deduced from the same root ^ s=y, qut, rendering the Persian bdrdtiya TNR 34. We have , therefore , to look for the words that render the Persian i apmiydkhsiyd and brijim in the groups between the two personal pronouns in the fourteenth and fifteenth lines. Of these groups the first is undoubtedly to be restored to , the middlemost wedge having disappeared; and the last group being somewhat indistinct might have been either £^y*~, be, or perhaps rather m. As the words are found nowhere else in the inscriptions, I am unable to separate the words expressed by these ten Median groups, u DHy y-yyy - *cy -yy-, and have only to observe that the Median word rendering the Per- sian brijim has evidently changed its place from after mana to before the corresponding Y The following Persian sentence luimyasrim hticrim apdhrim a quitted seems not to be quite literally rendered in the Median. The corresponding sixteen groups, after its -T= qutis, in the fifteenth and sixteenth lines, contain first the relative pronoun ppo, and then the first personal y-yyy 1 Yo , both of which are left out in the Persian; while on the other hand we find in the Median nothing corresponding to the Persian hdmyd- srim , unless we should suppose that I have made the 1810— 1S44. 24 370 MKDIAN SPECIKS OF A K KOWHK A OKI) WUITIXU. mistake in my transcript of separating the wedges y instead of joining them into one group FYYY , tit, and then that the three groups £^Y Y~- Y^ YYY i ppoth , com- posed the word rendering the Persian hiimyiiiam, which supposition would certainly restore the want of harmony, and considering the great distance at which I was obliged to take my transcript through a telescope, it cannot he very sur- prising that r -yyy might have been erroneously copied for Y*~YYY” The three next groups ►YY*"' i khabe, render, as we have just mentioned, the Persian hticam, and in this place, too, the sense of help, assi- stance, seems to accord prefectly well. The next four groups ►►Y *~YYYK -Y ' pthriklia, compose the word that renders the half effaced Persian yyy KT * r Ht -m< m < iitT * t m- * 11 T IT -It Hf -m< * !! < * TS S-TT Y ■ -ay- ^yy 20. -ST ^TT * T V -S- -T * T < yy - a -yyy * Y v 4 * ~Y T= m= ~ i -yy ~ -3 Yy- 21. -Y < S-TT ~Y * T ST =T - r W * y y^ a -STT- * T -STS ST Si - Y * MY y si -E* -YYY 22. -STY- Sm', * T -STS Ku , rex , and ►{ i • ph , strenuus, and farther the oft recurring, zu viyi Aurdzdan, ex voluntate Aurainazdis. Of the Per- sian word A nr am a zda h a the last a, as has been shown above, has disappeared together with the mark of separa- tion; hut no other characters would seem to have been here before the following sentence : addmsim gnpdwu nh/dsi’iddydm, as the Median text has only the following groups r -W -T ^ MM IB) -BIB- Of these are ►Y the only ones that this line has in common with the thirty fourth, in which we have to look for the word corresponding to the Persian gdpnm, and these two groups must, therefore, be considered as forming the be- ginning of the Median word which translates both gdpdwd and gitpum. The space between Auramazdahn and gdpdwd must , accordingly , have been taken up by that , which has been rendered by the groups between the correspond- ing Median words Auriizddn and These being only y -yyy , or the mark of distinction and the first personal pronoun , it is evident , that if adumsim were, as Lassen has supposed, a verb signifying Jie subdued”, the corresponding word must have been left out in the Median version. But as this is not probable, we must look in udiimsim for that personal pronoun which the 380 MEDIAN SPECIES OK A H HOW HE A ME I) WRITING. Median text possesses. I am, therefore, inclined to resolve addms/m into adorn, the first personal pronoun, and sim, which then would he an inseparable particle, with the sense of the Latin met, or the Sanscrit eva, and which might easily have been either wanting in the Median language, or merely left out here, as being less necessary to express the precise sense required; and I accordingly translate the Persian addmsim, egomet, even I, or 1 myself. To gdpdwd niydsdddydm, then, correspond the six Median groups ^ ^ MMf C3 >yy, which I separate before 5 because this group expressing the a is otherwise found always in the beginning of words, and because the syllable ivd most probably would have been written either Y^=Y only, or Y^| [ in analogy with the syllable rd which we found above written — f. The Median ^ ^ tkha- phirea , would correspond to the Persian gdpdwd , and >yy • Ada, to the verb n /yds d d dy dm to which words I shall presently return. After the last word we find in the Persian text a chasm and then indistinctly « v m in the end of the thirty sixth, and hmt \ at the beginning of the thirty seventh line , v which Lassen has restored to «TT m HMf but without venturing any explanation of this word. The Median text has immedia- tely after dda the groups, ►►Y Y^ ppo, or the relative pronoun, corresponding to the Persian t/j/'i CMM K-3, and this pronoun it is which I find in the defaced word, considering the indistinct wedges « \ as the traces of Y<~. whereby the chasm must have been filled up, with n -YyY, that is, the mark of separation after niydsdddydm and the first group of the relative pronoun tyd; and as the last groups are distinctly dm , we have the whole MEDIAN SPECIES OF ARROWHEADED WRITING. 381 Persian word ti/rim , quam , tlie relative pronoun in the accusative case of the feminine. This pronoun can only refer to the preceding gnpdnm , which, therefore, cannot he a masculine noun in the accusative case plural , as Lassen has supposed , hut must he a feminine noun in the singular in concord with the following tyam. I consider then gnpdrva , or perhaps rather gdprcd as the instrumental case of gdpu , of which we find 1. 41 the accusative case gripitm , and which apparently has the sense of insurrection , rebellion , ingeniously ascribed to it by Lassen. But then farther we must alter his trans- lation of the word niyasdddydm from an active to a passive sense. Therefore I should propose to read the Persian groups niydsddydm , and to consider this as a passive of the causative of a root sad, or sad, with the prefix ni, and to compare it to the Sanscrit nisad, which, besides the ordinary sense of sitting down and dwelling, also signifies to he vexed , or afflicted 1 ; the causative then would be, to make vexed, or afflicted, and the pas- sive, to he made vexed. We should then have in the Persian the same passive affix ya as in Sanscrit, to which (instead of the termination of medium which is the regular one in the .Sanscrit grammar) the active termina- tion has been joined — a formation of the passive which we also find occasionally employed in the Sanscrit by the ancient poets. The Persian sentence: addmsim gdpywd niydsddydm and likewise the Median : Yo tkhaphiwa dda, together with the following one beginning with the relative pronoun I therefore translate : I myself w as made vexed by a rebellion, which I did wholly put down; regarding, as before mentioned, apdhdrn as an adverb corresponding to the supposed adjective apd/id and with the sense of *) Confer my Radices Lingua: Sanserif under the root sad. 382 >1 R D I A X SPRCIRS ok ARROWHKADKD WRITlVfi. wholly, strongly, or something similar, and giving to the verb awdqnmrdm , or perhaps aim aqunwdrn (as there seems to be place for the mark of separation ^ between *~YHi n'd , and yff, a) the literal acceptation of putting down, destroying. As to the corresponding Median words ter -m< -st T-. v>Mr, a.. Za.poyos , we have the mark of distinction f and, as I suppose, the word that renders the Persian putikaram; and, therefore, in the three groups *5^ IT 5 that stand before the mark of distinction , we must look for the adverb which should correspond to the Persian chiydehiya (chichi). These groups giving naida , or nida , we may compare this word with the modern Persian 5 niz , though this adverb conveys a meaning dillerent from that which could have been possessed by the ancient Persian chiya- chiya', for, the sense of the Median sentence would be: The nations which the king Darius swayed did also bring help against the rebellion; instead of which the Persian has: brought help in every way . The veil) corresponding MKDIAN SPRCIRS OF ARROWHKADKI) WRITING. 387 to the Persian bdrdtiyd (or rather btiratt) we must look for in the groups that begin with (]u , after i^Y in the thirty fourth line. As we found above (1. 15) ^Y^ ; following here after yTX qu, belongs to the root of this verb, but it is not quite clear how many more of the following groups have com- posed the inflected form that renders the Persian barati. After the distinct Y^Y? the traces visible were Y *~7 which, if the space had been left open on account of a natural unevenness, might be the group ^YY *~5 &e: but as there appeared to me to be some indistinct traces visible of horizontal wedges both before and after the last vertical we may suppose these to have been the two groups ►Y HFi Mi’ and as we a ^ ^ 5e CIJ d this inscrip- tion three verbs in the third person plural ending in Hf-, ti , we may venture in this place to separate the groups, so that yY3 ►Y Y»^Y ►Y • qutrcatti , or quiwciti (the ^Y being only an orthographical sign to mark the hard sound of the following consonant) will be the word that translates the Persian bdrdtiyd, they bring. In the following sentence we find in the forty third line of the Persiau text before Pdrsahya martiydhyu the three words addtiyd dzddd bdwritiyd, which are repeated 1 in the forty fifth line before the words Parsd mdrtiyd. As in the Median the words rendering Parsd and mdr- tiyd are in both places transposed, we have to look in the groups that precede Y ^ which would seem to be the same word as the one which 1 have transcribed yiita in the Persian (perhaps writing Y^*~ yyy, yd, erreoneously instead of K- or > what is more pro- bable, HfY~ bikhti, can scarcely compose more than one word , I consider this as a verb in the third person plural with the termination *~Y ^5 £*» which we have already mentioned as peculiar to this person, formed from a root ^=Y*~ *~Yt^ or ►Y"”’ *~YYj£o bikh, with the sense of, to adore. AIRMAN SPKCIKS OF ARROW HRADKD WRITING. 399 The next words: me Auramazdes me tuere , are known. Then follow the six groups, ^ corresponding in the Persian to the two words It ad a kdr . . . , of which latter , the kdr is only indistinctly visible, and the rest has totally disappeared. Now of the Median groups the three first, occur again in the forty seventh line where they can only correspond to the Persian huwdtiyd or huwdtiyd geistd, and the three other, compose a word, which otherwise renders the Persian /idea, and to which we have attributed the sense of help , assistance. The same acceptation answers apparently in this place, for in the Median word might have been implied the literal sense : according to one’s own will, from whence may he derived both the sense of self- willed , w'ilful , which the word undoubtedly has in the forty seventh line, and that of absolute, sovereign, or willing , gracious. The Median sentence in the forty second line would then imply: Protect me, Ormuzd, by thy gracious help. The group being found in this word only, its value cannot be determined. In the Persian the word luidd , which Lassen regards as an adverb heic, seems rather to be a preposition, cum; in the defaced kdr . . . therefore , we must expect the word which renders the two Median that I have translated gracious help, and which perhaps was the same word as that which has stood in the forty seventh line correspon- ding to the Median kkhabe. The fol- lowing words : et gentem hanc et hanc regionem, we have already explained. In the two last sentences we meet with three new words. The beginning is known: illud ego Auramazdern. 400 MEDIA* SPECIES OF ARROWHEADED WRITING. To the Persian zacltiynmiyd udutudiya correspond the words composed by the six groups at the end of the forty fourth and in the beginning of the next line, which most probably must lie separated so that the three first ►►yy compose the word that renders the Persian zdcliiynmiyn ; and the three last ► the word with the sense that is implied in the Persian dddtddiyd , which Lassen has translated, bene moderatas mentes; but as the Median text inserts a r =m> ro, before the last verb, the Avord has rather been a vocative in analogy with the amdsuciyS found in the inscription C, with the sense of, who possesses wisdom (Sanscrit nttadlii), allwise. In the Median text then follows: Aura- mazdes me, Avhich last word is not found in the Persian. The last word in this passage , which is preceded by the mark of distinction ZW <- =n * z ti, I suppose that the third verb, rendering the Persian puduyri likewise terminated in the *~Y^- i ti, found in the preceding line, the more so as the mark of distinction ► following after shows that the next word must be the one which corresponds to the Persian pupim. Of the six groups , therefore , between yy and the mark of distinction , I suppose that the three first YY< <- -Y> •nikha, compose (he word that renders the Persian 404 MEDIAN SPECIES OF ARROWHEADED WRITING. huwdtiyd gdstd and which, as before mentioned, I consider either as an adjective with the sense of selfwilled. wilful, or perhaps as an adverb, wilfully. The three next groups, then , ^ -Y- would compose the word that renders the Persian verb pdddyd, and conveys the same notion. To the Persian pdpim tyc’nn rastnm correspond, lastly, the eight groups between the mark of distinction and <- in the last line. The first of these is half effaced, hut the traces that remain visible show its having been either £^Y*~? P*> or ►YT P e > am ^ ^ rom con ‘ gruity with the Persian syllable pd I suppose that the Median word here occurring was the same as the Persian, and therefore give, as mentioned before, to the group Y3 the value of thi, or perhaps si, th and s being those of the Median letters which approximate nearest to the Persian p. The Median word corresponding to the Persian pdpim is then £eY*~ pi thi (pisi), or ►YT"-? pat hi (pesi). The next word ►^Y Y^i we know, renders the Persian pronoun tyd, and therefore the four groups ►YT*^ Ysr— Y ►Y’ path kha, must compose the adjective that translates the Persian rristrim. 46. ? £atagus, Haruwatis, H- 18. ithus, Gadara, Saka, Miika. 3?atiya 19. Daryawus khsayajiiya yachiya 20. awama mauiyahya haca aniya- 21. na. ma darsam imam Parsam karam paclii- 22. ya. yachiya kara Parsa patahiitiya. hya 23. thuwistam siyatis akhsata huwaci- 24. ya Aura nirasatiyii abiya imam vijiam. 1 Ego Darius rex magnus , 2 rex regum , 3 rex regionum harum 4 multarum, Vistaspis 5 filius, Achaemenius. Gene- rosus (sum) Darius 6 rex. ex voluntate Auramazdis. 7 hae (sunt) regiones, quas ego 8 tenui — hac in re Persae ac- tores (fuere) — 9 (et) quae adorationem igni , mihi tributa 10 attulere : Cissia, Media, Babylonia, 11 Arabia, Assyria, Gordyaei 12 Armenia, Cappadocia, Sparda, Iones 13 turn qui terrestres, turn qui 14 maritimi ; turn regiones hae: 15 Parutia, Asagarta, Parthi, Zaranga, 16 Harii. Bactria, 408 MEDIAN SPECIES OF ARROWHEADED WRITING. Sogdia, Chorasmia, 17 5atagus, Arachosia, India, 18 Gan- dara, Sacae, Maca. Generosus 19 Darius rex venerandus. 20 Talis mihi (sit) adoratio consecrata. 21 ne conspiciam hunc Persani actorem huinili (conditione). 22 Venerandus (est) actor Persa (utpote qui) tutor sit. Sit 23 (in) Ion- gissimum (tenipus) tbrtuna incolumis, o celebrande Divine juste, apud hanc gentem. It may lie observed that as the Scythians, Squdra, which in the similar list on the tomb of Darius are enumerated among the tributary nations, are not mentioned here, we may perhaps from this circumstance infer that these inscriptions were cut in here, before the great king undertook his expedition against the tribes of Scythia. The Median inscription H, that stands at the side of the Persian I, is preserved complete without the least indistinctness or mutilation , the unlettered spaces which now and then occur being attributable to an unevenness of the stone at these places. The Median inscription is not a mere translation of either of the Persian, hut is of a quite original tenor, which we shall now venture to explain. The first seven lines contain only some of the usual titles which we have already explained : 1- r -m * r nr -m< e=a tit < = Y YY luik , in the eighteenth line , contains for the most part known words , the contents therefore are not difficult of explanation. As we shall insert below the whole passage in the original character, we here only mark the known words with the corresponding Latin terms. The con- cluding passage begins : IS. Ille Dari- 19. us Rex generosus. Me o Aura- 20. mazdes me o tuere cum di- 21. is. Then follows in the twenty first line the pronoun yy luik , after which the conjunction Mf quttu, is tw r ice repeated. It is therefore probable that the two words following after qutta, both of which are accompanied by the demonstrative pronoun sa, are nouns governed in the accusative case by the prece- ding tuere. Darius asks in the inscription NR the pro- tection of Auramazda for his gens and his populus ; Xerxes in his inscriptions for his regnum and his own palace with that of his father. As the inscription H is found on the principal platform , it is not unreasonable to suppose that one of the tw’o words following after 410 MEDIAN SPECIES OF ARROWHEADED WRITING. qutta has reference to the building and the other to the kingdom or the people of the king. As the word after the last ►► in the twenty second line begins with the same two groups as ►f ►ff*" ? khabe , and moreover possesses in a labial corresponding to the last syllable of the other, we are led to suppose that the word here found )[► ►f is an ad- jective derived from s=n- help , assistance, and having the sense of who gives assistance, assistant. Proceeding from this supposition we have in the five groups, V f ►y 5^, that precede the last to look for the word expressing gens , or populus. It is true that we always meet dahyus with the sense of people, or nation, but as this is evidently a Persian word in a Median garb, it is not at all improbable that the Median language possessed another and native word of the same import, and that this may be found in these five groups, V y^Hf M *~i ^y 5 sa.tkhat. The three last of these, ^ ^ i^y? that, being found likewise in the seventh line before what seems to be some case of the pronoun , I conclude that these compose the word with the sense of people, and that we have in the two groups V y^Hf an adjective implying good, brave, or the like. If this be true, we should in the groups Eff- TT= have the word that had a reference either to the palace , or perhaps rather to the very platform on which the inscription is found, and on which all the buildings of the palace were erected; and accordingly I ascribe the sense of platform to the word ►► ^►5 .hero , which occurs 1. 8, 9, 14, 23, preceded by the mark and 1. 11 by the mark This word is 1IRDIAN SPKCIKS OF ARROWHKADRD WRITING. 411 apparently formed from a root ►► by the affix yfC:, ro, in the same manner as we above found HK yutro, formed from the root yat. Lastly the A futk, in the twenty" first line I take to be the accusative case referring to the preceding T -I'll ro, me; and the whole sentence then would run: 20 me tuere cum 21 diis, ilium (me) et fundamentum 22 hoc et nobilem populum hunc opiferum. The four last groups of the inscription 1SI e -Sib compose apparently a word that is derived from the same root as •> which we found NR 47 corresponding to the Persian piidayd, dere- Iinquant, the form of the word merely showing that we have here another tense or mood. The known words then in the two last lines give : illud ne — quod homines — derelinquant. The whole there- fore might be conjectured to have been: may it never happen that men leave the good , or something to that effect. Now as Darius in the Persian inscription H im- plores the protection of Auramazdes against ddruga, the vice of lying, and as telling the truth was, according to Herodotus , one of the three things which the Persians taught their children , we may venture to suppose that the Median word ►Y i orikhkha , signified the truth. And for the other unknown word HEK- *~tt HAM we may perhaps take from the Persian inscription I the sense of ddrsdm,\ ideam, so that the two last lines would imply : may I never see that men leave the truth. The concluding passage of the in- scription H from the eighteenth line is consequently : MEDIAN SPECIES OP' AHROWHEADED WRITING. -e- * r nr -m< e=,\ tst < ii- 23. feram. Illud ne videam quod ho- 24. mines vcritatem derclinquant. MEDIAN SPECIES OF ARROWHEADED WRITING. 413 As in the middlemost passage we find the word efh »T~ platform, four times repeated, it is evident that something must he stated there touching the erection of the platform. We find farther two words, each twice repeated, ►f, qusikha , 1. 8, 15, and ^yy, qusiya, 1. 11, 15, derived apparently from the same root. The latter, qusiya, it is true, occurs in the inscription NR as the name of the Cossaeans, hut of these of course there can he no question here. On the contrary, the one , its ght be supposed to mean the several buildings erected upon the platform, and to be a noun in the plural number. The following, *-£ T2T 5 sawa , we found D 13 corresponding to the Persian nominative plural ana, these, and might here also be considered as the nominative plural referring to the just mentioned preceding noun ►►Y ►Y j but as Y*^Y? Sana, from its occurrence in the seventh line would seem to be a more general inflected form of the pronoun, it is more advisable, to regard sana as a loca- tive case, and to refer it to the follow ing £^YY*~ yY^-j the platform. The two groups *-yy , lastly, would seem to be an affix , or postposition , implying above, upon, being in its use similar to the before mentioned ( qqa ). This sentence then means: buildings were erected on this platform. In the next sentence all the words are known and have been explained. It concludes with yY3 KY"“ MEDIAN SPECIES OF ARROWHEADED WRITING. 415 qusiya, in the eleventh line and is rendered: e voluntate Auramazdis hoc fundamentum ego exstruxi. That the following sentence, which extends to the relative pronoun ^ in the fourteenth line, contains an invocation to Ormuzd, is evident from the words which render the Persian Auramazda and lu'idn bag ibis ; the whole sentence kci» g = /, -3- 3 — t < -ar- t- s-tt * -= ETTT -ST tT= E£ *-yy "“►f 5 rovi.n , must then form the verb with the probable sense of prosperare, or propitium esse, amare. The first word Y j /<«£ , might be either the de- monstrative pronoun used here before the name of Aura- mazda in the same way as before the name of Darius, or, what would give a somewhat better sense, might be considered as an adverb meaning , therefore , on that account, and connecting this sentence with the one prece- ding; these two may consequently be translated: According to the will of Auramazda I erected this platform , may therefore Auramazda together with the other gods be propitious to this that needs his favour. And the follow- ing sentence, *- *-*- * - I * ~MT ' -2- * — T < -SYT- T- S-TT * -E * ETTT 13. -ST * tT = -I > <- iya wazarka khsayii- 13. Jriya khsayaj>ivanam , 14. khsayaf) iya rfa'Ayunam , 15. khsaya|)iya ahvaya 16. bumixa. Adam Artakhsathra kh- 17. saya|)iyp puthra. Artakhsathra 18. Daryawa-y khsayajnya 19. puthra. Daryawus Artiikhsa- 20. thru khsayajuya puthra. Artii- 21. khsathra Kh^yarsd khsaya- 22. jpiya puthra. Khsyarsa Dar- 23. yawu6 ! khsayaf)iyd‘ puthra. 24. Daryawus Fastaspahy 25. a nama puthra. FastaspaAy- 26. a Arsama nama puthra; Ha- 27. khamanisiya. 3>atiya Ar- 28. takhsathra khsayajrhm. 29. im dm iistasnam aj)a gd- 30. nam mam upa mam 31. kartd. 3>atiya Artakhsathr- 32. a khsayaj£>iya. mam Aura- 33. mazda uta mathra baga pa- MKIII4V SPKC1KS OK AKROWHKADKD WRITING. 425 34. buwa uta imam dafw/nm 35. uta tyd mum kart//. 1. 2. buwum. 4. asmam/m. 5, siyntim. 6. mart/ywhya. 7. Artakhsathmwi khsaya[)iy«ni. 13. khsayafuydnam. 14. The syllable ddh is written or 10. <<< } V a - 278. — , 290. EE^- i 296. Hf, «. 293. we. 286. EY- EY, pa- 279. *~EY/ n, na. 280. ~Y> P . 307. EYYY, 355. *-Ei sa. 288. iE> q- 301. 414. rim (pa). 311. s-YY, ««• 278-79. ■"^YYY i as. 291. -!Y, 295. 294. 399. =Y< *=Y» «■ 282. Efc, M. 293. (re, la). 302. if, kha. 287. EYY, 345. a. 292. /»• 300. E• 280. Bff> 376. ~YK~, 4ii. «*• 284. K> 324. ~yy, 336. t|, ////. 282. fi, po. 346. ~YK, 344. ) 328, 353. T--T. 355. *TK, oi. eH, 316. T=TT-> 377. HffE, M. 288. ^ , Mm. 305. f=m, «• 306. HTCMJ, 4io. r ?/«• 280. ^YY>4 278. YEYY- (?)• 353. -fff, ra ‘ 288. Xl.tM(si). 317.404 428 MEDIAN SPECIES OK ARROWHEADED WRITING. lv ’ ,ra - 278. y\ 5 h. 279. V, «a. 288. ro. 284. yj, 2 cr. 283. rWW323.414. o. 311. yf = . qu. 285. fff-. Am. 321. <, m. 278. 7 «*. 285. <-Y~, 344. ®*. 279. si. 285. «> e . °> and the Mediau language would in this respect he richer than the Persian ; hut I will not deny the possibility r that the language had only , like the Persian , the three vowel sounds n, i, u, and that the alphabet possessed separate signs to represent the different length of the sounds; and if this should prove true, I would arrange the six characters as follows : a ’ *» 7 - ll - In representing consonant sounds the Median alphabet is much poorer than the Persian, having signs for only sixteen consonants: 1 q; k, ^ • kh; 7 t, >=Y^7 th; p, m? pk; 1 V’ W7 ' r >* y »-?• */ ui -j, ^yy- w - The language might also have possessed the sound of /, although I have not ventured with certainty to give that value to any of the groups. Farther , we have the two signs and Y^yYY:' representing single consonants and apparently the very same sounds as ►f; t. and th. The alphabet lias moreover signs for syllables con- sisting of a consonant and a following vowel , but with this restriction, that syllables terminating in the short a (a) are expressed either by the signs of the single conso- nants, or by the signs that represent syllables ending in the long a (a). Now as each consonant, with the excep- tion of j 71, which is not found in the beginning of syllables, might form five syllables, we should expect to find seventy live, or if the language possessed the sound of /, eighty 430 MEDIAN SPECIES OF ARROWHEADED WRITING. signs for syllables terminating in a vowel. Three of these syllables ya, tea, ra, however, are evidently expressed hv the same signs that represent the single consonants y, tv, r; and of the other seventy two (or sev entv seven) we have decipl tiered the followin ig thirty 7 three (or thirty five): rM- qa; i qu. — ku. — kha ; khu; , kho. -jHf, ta; t! ; tu. ■ -K& till ; -a, thu. — Bit pa; pi; P e * po. , — pha; phi- — HfK> yu; vo. — MU- i ra; : -m v e z II, wu; wo. - - B- s* i; su. 11 i •> sa; ti -TMvH I »-T n *~tt ’ *-rr »- ►TT? - -.thuphti, 290. — ►YY- )(Y-i .bero, 410. ~Y -Y V Y=H(-, akha. sa . . . , 382. — Y aua p> 307. G--~Y~j anappi, 307. . • • Ef- > anappitu , 309. ...-Y- ~< -CY> anappitun, 308. . • • EfY*" ~< m HY, acap- petuthta, 309.) ~T -E-< TE sithus, 297. (?) ~ -Y -TTT< ~YYe- --kb, 393. ~YYY» Asa, 305. -E-TTT -STT- ~T -T, Asrakha, 295. “YY< <~ -Y- •nikha, 399, 403. ~Ye -YYY< E-/p Ariya, --T S¥= S-TT EE= -T=> athtiti, 403. ~T < -STT- T- S-TT- a». razda, 312 ( — —_Y? Au- razdan). -y saw; *~E YSY EE rY-Y; sawaq.; 305. ~EYe -YYY< m Aiiwa, 292. ^eYe -yyy YIY Hf= rfN, Aruvvatis, 295. *~eYe eY E— tp A bay a, 301. — eYe E-YY ^ ada, 380. -ETE S-TT E?T S-TT -EL Adastana, 292, 303. ^eYe Ye HeE -IY) Asina, 296, 303. ~eYe -7*- eY • kkbalu, 380. MKOIA.V SPKCIKS OK AKKOWHHADK It WIUTING. Hf ~Y~Y HE- Hf=, phaphphiyiti, 403. Hf ~yyy, Pba Sa , 291 . G-EYY~ Ph&sar, 291. . . ^ -3f, Phasiirn, 391.) Hf ~m yey , Phasawa, 292. CHf~3f»3» Phathawa?) "f ^TTT • pbaru, 283, 333. -T =rir -T= E->', ~T, Pharutiyab, 304. Hf <«, Phasa, 3G8. Hf= -yyh hyy~ -y < hy Tikhrakhudap, 300. Hf=<- M? HfY~<< 411. ~yy~e <-y- y- -m, 309. -YYE ~e -yyy (rffY) v, Khsarasa or Khsarassa, 288, 320. -YYE ~Y^h khkhahe, 308. -yye -y hem? y- -y, khkba.pokha, 410. -yy< yutu, 310 . HPK -Y~Y- yutuplil, 350. ~YYK HY, vutuda, 351. -yy< -y- yut, 323 . o-hy- yutia , --HY HY~, 435 yutlar, • . HY Y~< T3 yutta.wa, . ~5Y -Y E~YY, ynt.tta,. -yY— Y- yut-,- -yY-Y EHfY’ yut.ta) ~YYK -Y tY- (-Y.), yutro, or vutrot, 354. ~YY< -HYY<, yophri, 358. ~YY< IY- Y- ^n 7 - yuthzra, 391. ~YYK Y^TTY yutiitu, 3io. ►m ~ OYYH) duzzatbuida, 394. -< HTT -ST ^TT S-TT, tuthihusta, 310. lot) MEDIAN SPECIES OF A RROWHK A DK D WRITING. : -tt< -< , vcvutu, 400. =t -p -m< -in-, ptirir, 370. :if ™ if. p.kha , 414. -if EE- ? pi, 373. :if3fi-fffsa., 355 . :iffififff, ppoth (?), 370. :E ™fff ^ qasraya , 301. :E -Iff -TT> qru., 315. ~Eh TT if j wuzakha, 311. E~ff E~ff j ttata, 295. -f — f fi, khaphas, 377. -f if if ~Ef -f 5 Khat- pathukha, 302. -f -f f~ -W -ff-> zruin, 392. if if iif ? tkhap , 385. (. .ifif f^fj tkhaphiiva, 380.) if if iif. tkhat, 410. if- 4 -ffE -fi, bikhti, 398. if- Jji Ki— ? biovi, 390. if- f iffi- pliinim, 402. (-Ht -f -fc-Eff-, phininidattir, 327 .) -ft- -fff< -Eff- f- V. berirzza, 307. -ff- -fff< ^ff- beril, 372. (— E-ff* berista, 385.) -ff- -fff< Hf, berida , 355. OT- IP P~T 0, pea- .kha, 404. iff- fl, pethi, 404. ifff, yo> 331. (,-fft -rr -Et, y°- n ) ifff if yfif =-ff. yotta, 323. -fff Hf. yota, 398. if iif, khukhab, 305. ^ 5 ff -f E-ff' sphada, 302. — ^ff J-ff — Ef, stana, 358. £ .kha, 335. Takhapharatu, 303. -fV^Tr^Tr^phsata. 3-yy -y ji^yy^yy-, .kha 9 33a. tadzaram , 283. e-A TETT- a==n-?3, >yy -^y ^yy^? datam. 353. 372. ► IT 009 ►in :ny y~ ^*~yy, tasta, sio. ^Sr edakha, 345. >yy a -yy< ^yy, da. Yun, Yuna, hyus, 280, 333. (. . 302-3. -Jf 5 dahyustun , &A 3T ya-ta, 400. — :y syy^ dahyu. BB~ >•> 414. nam) B1 *~W 7? Parasa, 288, :hy a -yyy ^yy -< 354. (^*-^yj , dahyostu or ^y -riB -m< dahyostun, 280, 333. Bakhthris, 293. 2 -TI A -A sTsf, B1 P*™» 283, 333. dahyus, 279. ^yyy -a. 382. ^y DHY’ suguda, 293. ^yy v, •-, 305 . ^y- yy -yyy< b^\ m < rar, . . wothirarra. ^y 5 Dariyawus, 278, .. qufwatti, 387 , yy^= -y= £ zana- stun op zanastun) H << zu, 343 . CH, zal {} <^y^ y^ ,yyy ^yy, za.poyos, 386 . 11= o.r, 382. K=-m<^=f } orikh- klia, 411 . K= C-3-) ,T = <- IrfffD <- B=A, Okhavenisiya or Okklia- venisiya, or Okhavenis- siya, 285 . h= =r=y- « P hi, 339 . U- y^y -^-, owak. 382 . MKDIAiV SPKCIKS OP AKROHHUADKI) WRITING. 439 I jj £:|, ozakha, 3 11. - ^YY V, vissa , 334, « s=Y - YYY' odo. uqbetvo, 299 £ v iy i , 343. ■^YMY~ ^YY si®- 0 , • Y- “ Y^ ? ■ sathi, 343 . (•m\ Y^Y, .sathi . wa , sis.ni) 363.) 2.Ti!^a>a J|] \ ' S^i , /AyA/z-s/ ' YBty.^K ^<^hh%JbZ-V-^h. -Vi. - mi-.m E . ♦ XHfT. r.EYYY< i Y /6 -!K-< -IK ^Ie-IIeEM -"H-- W3f£*YYT< Y9. Y, 3YHTf£. St- At -YYK ^Y. ft*/ rjHY Y Y^YY^YY<^ YYI^YY.^YY— Y.4 Y. /(/ -BY^ri YIY -Yt ^YY.Y-m. ^11.r1^11htl^bl11-ldB £ =. '{(' -tf^rr t v-r-^ p< -Mr- Et=!tt te-m- ■u -T.< sHTKf.lt. E»- ft. r fj ,0-aH-ffi.Mft*. ?2 -If J-r-tV t-ETE < E= <> ; -,V M eI TWTT.-U-1T. I. 73 £**.<#. Y. V>ev *Y. «* fcHfrSK ,Y£.<^Y.**Y. z.^1. Y. Vi *YY. ^ iY-tYY- Y.HrvV ►H-EYY— < Y- swMk //V ^YYYY <*>. .7 *»/ ^.r/ i Ilf *~ I'ab.XTl o 2J 26 27 HTT.Kt.sm. ~ ~ .srfrElWr^V=lf sKWHIt 26 2.9 mi.n^'r^.«|^.E£EK<^.v.xHira.itnitr.4 30 ►y-t. y st.-h r= iHfrt -y. y= y mf-T.Ht hyk^y y- -rryy.YY Y^K-.^YY-TfS. = YKT-YE-w.-EYY-.eeY. YE.--Y- *1 1 11 mr ^ , :> . M => .W.W.M. V: ~y ~y* !3UmK^ \v 11 T hmK I. wt M n (fc.-rT . ill <1- -<■ V *ui.+t*. yi.s r Jff<1 Y S? . 7// s’ S'//** / I'fy'/Jf V/SSd SZc/ ^ |f . 3f I A. lh *ffft. - =>. !<«.!!. I ski II -ft. :! = cl If If *H<^. 1. If ! H a 4 YY X ^ ^ C N v K \\ 4 fit HI \ 1^1 V *$s x 1 -% u I V>w ifii ?m ^ ^ *•$ ^iK X -x x S 3>§ \ T\ ft \ V§ ff t¥t Ifi V •N v ^ ftf* %Ji *N I Tab. XV U lab. XVI W/ Tab.XVI I). Tab XVI. aa 1111 aa u - A 1 ¥ [ V£ V" ' AAA ii Y AAA i Y ; /‘V 11 tT A A A A =- — a 1 V aa:: if Tt Y Y j|U ' A A I AAA 't i f ill! A r= /r Y 1 A A A + A i-i VW' vw 4-' Y v A — A Y w A ' AA iIT| 2 AAA TF A A 1 A A A . M £ ? ' . lujl II i Y M A • liil 1 W v'v' w \7‘ » ; a li - • • rm AAA liii n7 w nil Tab. XV ; I u \ •I \ 1 § I I K 7*- I v \ V' II gf. a 1H1, iff. £l W < =K!.#J. flff IH .^A-WtfA HI E! !*W -*r\m &\r *' & w Cfr I 1 J jfc m 9 .