page 392 note * I use L. to express “Lapidary.” The lapidary characters are numbered according to the list of 287 characters which occur in the Great Inscription at the India House, published many years ago by the East India Company.

^{page 393 note 1} The words are ithkunu and ivnu'u. Both seem to he plurals; as are most nouns and verbs referring to gods or longs. The variation in K 4, is singular.

^{page 393 note 1} This is a mistake. I am now quite satisfied that Flandin has given an erroneous copy of this part of the inscription. He has taken into the tenth line characters which in the original stood below the true ones, in the twelfth line, where they compose the word maduta, “many.” The argument, therefore, which I have given in the text, for this character having the value du, is of no weight whatever. I have, however, no doubt at all that this is its value. It occurs in the name of King Nebuchadnezzar, on Babylonian cylinders, between ku and T.ru; and, as three concurrent r's are inadmissible, it can have no consonantal value here but d. The Median character of like form begins with d or t, which affords a strong presumption that it does so too; and that it ends with u appears, independently of the Van Inscriptions, from the above-mentioned word maduta; which is sometimes written with, and sometimes without, a u after it. (April, 1848.)

^{page 397 note 1} It is most probable that the Median value is nar; the final r, which had very much the nature of a vowel, being often assimilated to the consonant which began the following syllable. See the following note. (April, 1848.)

^{page 398 note 1} The value of this character, and of its equivalents signifying “king,” can, I think, he proved to be nir or nil; but the terminal sound was, as in many other instances, assimilated to that of the consonant which commenced the following syllable. Thus it should be read nis in A.khama.'n.nis.si.ya and nin in the word before us, nin.nu.wi. In Median, we must either suppose that this kind of assimilation prevailed to a great extent, or that ar, an, am, and ang were undistinguished. I incline to the former supposition, for reasons which it would be out of place to bring forward. I think that the final r sometimes retained its sound and sometimes took that of the following consonant; but I cannot admit that it took that of a nasal, when a nasal did not follow. I read ad.da.ra, not an.da.ra, for Gadára; the proper value of the first character being unquestionably ar; and if this be the first character of the name of Cambyses, as I have been told, I should there read it ab, not am. It seems to me highly probable that the terminal r might be sounded with its proper value before any consonant. We have for examples, Hardastana, Harbaya, Harmimya, Karka, and Parsa, in all of which the r was probably sounded; and it must have been so in the name of Xerxes, when written, as it often is, with an s between the ar and the sa; for a concurrence of three s's is quite inadmissible. There can be little doubt that terminal n; was also assimilated, when it preceded r and probably s; but it was certainly sounded in the verbal termination nli, where it is characteristic, and I presume before all consonants but those which I have mentioned. I believe these facts respecting Median and Babylonian assimilation have not been stated; and it will be of importance in our present enquiries to keep them in view. (April, 1848).

^{page 402 note 1} Also with the Persian postposition á, which is placed after various locatives, as wité-á, Mádéshu-á, &c. It is stated by Benfey that this is equally the case in the Vedas.

^{page 402 note 2} It is very possible, however, that it is thu, in conformity with the Persian transcription, and with the of Abulfeda. [I have now positive proof that it is thur; the name of the God is Atthur; that of the country Ath.thur.ru or Athuru. (April, 1848.)]

^{page 404 note 1} I am disposed to read this title ni kâmuin. The value of probably begins with k; for this character closely resembles that which at Nakhsh-i-Rustam terminates the word corresponding to Katpatuka. The vowel of this character and the consonant of are only conjectured.

^{page 405 note 1} See § 36.

^{page 408 note 1} In confirmation of this, the compound character (see § 39) may be compared with the Median which has the same value. The first four wedges in each correspond; and the remainder are what we have now before us. (April, 1848).

^{page 412 note 1} The supposed genitives plural nabinam and âluwinam occur in clauses which contain one other word only, and which are found in various contexts:—sometimes at the beginning of the inscription, followed by the clause “the king says.” The second word in this clause terminates in na, and is, I presume, an instrumental, the clause signifying “permissu deorum” or the like. It cannot, therefore, I think, admit of doubt that this case is really the genitive plural.

^{page 416 note 1} In most instances in the Babylonian and Assyrian inscriptions, L. 77 and its equivalent are determinative prefixes of names of districts of country, smaller than those denoted by In the Median, two characters resemble that before us; but neither corresponds to it. One is ka or ga; the other t.It is to be observed, however, and the observation may be of great importance, that this last character is often used, like L. 77, as a determinative prefix. It takes the place of before characters which the addition of a wedge like this would convert into different characters; as, e. g. before or which a prefixed would convert into or I believe that these two derivative prefixes and are exclusively used before words which literally or metaphorically denote place. (April, 1848.)

^{page 417 note 1} As I understand that doubts have been entertained as to this being the true value of this character, I think it right to add that there is positive evidence that its Babylonian value was In the contract published by Porter, the names of several parties occur in the body of the inscription and again under their respective seals. One of these names is written in one place with and in the other with and that this last is there can be no doubt. Whether or not the character sometimes stands by abbreviation for rada or lada is a distinct question; I believe it does. (April, 1848.)

^{page 418 note 1} I have since satisfied myself that this verb must signify “honour” or the like. There are reasons for giving the second character the value pâ or phâ, and there are others for giving it the value shâ. If the initial character of the Median name for Zharanga be the same as this (as Westergaard's statements render probable) the balance of evidence will be in favour of shâ; otherwise, I should lean to pâ, connecting the noun tapâsina (as the instrumental of a verbal in si) with the same root.(April, 1848.)

^{page 419 note 1} It is yâ-un, the consonant of the second character being unknown. Can the root be akin to the Sanskrit yachh or yam? The second character is also found as medial in a verb pa-umi which certainly corresponds to the Sanskrit pâlayâmi or the Gothic fôdja. The Sanskrit l is often a corruption of d; and is probably so here. This would give for the value of du or tu; and indeed I have observed a place in the inscrip. tions where it seems interchanged with and its undoubted homophone Probably, therefore, the word is yâtun. [It is yârun, i. e.ya ârun, the first being the relative pronoun in the nom. plural neuter, and the other a verb agreeing with it in the imperfect. The meaning is, I presume, quæ dedêre. (April, 1848.)]

^{page 420 note 1} This translation must be modified; yet I believe it to convey the true meaning of the original. It should be observed that the leading verb tashâmi (or tapâmi) here terminates with while in the initial sentence previously quoted, it terminates in the latter was probably the ending of the transitive, the former of the middle voice. To honour one's self on a vanquished enemy may have been an idiomatic expression for imposing terms on him. The Greek which is “to honour” in the active, admits in the middle the sense of “imposing a penalty.” That the verb here occurring has the sense of “honouring,” or something very similar, is evident from a sentence in XLII. 5, 6, before the commencement of the king's speech. Tashâmi Kinuâm Spuwina-kanan uru tashâmi nabin, “I honour Kinuâs, the son of Spuwinus, as I would honour a god.” This is a complete sentence. The form of the potential mood which it gives is interesting, and is I believe unique in these inscriptions. (April, 1848.)

^{page 420 note 2} This genitive is often abbreviated The ideographic prefix is here used as equivalent to as the corresponding character is in the names of Babylon and Bactria at Nakhsh-i-Rustam; while the initial short a is dropped.

^{page 424 note 1} This passage is erroneous. The word is âvibi, or the like; the second character being See note to § 37. (April, 1848.)

^{page 429 note 1} Since the above was written, it occurred to me that perhaps I confounded two distinct characters; and a close examination of the inscriptions has satisfied me that I have done so. The character which has the power of vi, bi, mi, or pi, is properly it is so represented by Schulz in the majority of instances, though he has repeatedly confounded it with which is only used in these inscriptions as a determinative prefix, or as an abbreviation of the word nabi, “god.” In the Berlin copy of Mülhbach's inscription, the distinction is always marked. The name of the king there speaking is written with the former of these characters after s, Arrasvis; that of his father with the latter, and I suppose it should be read Nabiriduris; though it is possible that it was Arriduris, the having its Babylonian value an, which is here assimilated to the following r. It follows from this, that the reading wanâmi in § 25 is inadmissible. The word is uviâmi, or the like; the first consonant may be m or p. From the termination, I presume that it is an optative or potential. “May I —“ or ” I would —.” The root is perhaps connected with the Sanskrit vap, and may signify “ to offer.” (April, 1848.)

^{page 430 note 1} In one place, indeed, (VI. 9,) it would seem as if this character were used as an abbreviation for the whole word svina; but only part of the characters remain, and I suspect that, though not so marked, the defaced part of the rock must have been large enough for the rest of the word.

^{page 431 note 1} M. Botta says that the final character in this name, which occurs in no other proper name that I have seen, is interchanged with If so, it must be a u, which will suit the present word as well as s; for the final sh of the Persian may be omitted here as well as in Babirush. is also a good value for the affix of the 3rd sing., which this character represents. (May, 1848.)

^{page 431 note 2} While this sheet was passing through the press, I received a new cuneatic datum, which appears at first sight inconsistent with what I have here advanced, but is in my judgment capable of being completely reconciled with it. As the inconsistency may occur to others, being of a very obvious nature; I think it best to state it myself, and to explain it as well as I can. The new datum is a line of one of the smaller Babylonian Inscriptions at Behistun, which M. Botta has published in the Journal Asiatique for May, 1847. This is evidently, as M. Botta has remarked, the commencement of the inscription marked b in the key plate. It contains two words, answering to Iyam Gomâta of the Persian inscription which corresponds to it. Now the last of these words begins with the character which I have here valued as agu. The five characters of which it consists are, according to my values, Agu.ma.â.ta.â. Here then, it will be said, I have committed an error; for the initial character must be read gu or go, without a preceding a. I, nevertheless, persist in my assertion that the value of the character is agu in this place as well as in the word Sa.at.agu.s' The explanation of the apparent inconsistency between this reading and that of the corresponding Persian name is this. When the value of a character terminated with a vowel, it was customary to repeat the vowel at the beginning of the following syllable. Thus, in the word last quoted, Sat is expressed by Sa.at; and in like manner, in the name of Ormazd, mas is expressed by ma.as; in that of Hystaspes, tas by ta.as; in that of Achæmenes, man by ma.an; and so in instances unnumbered. This mode of writing was extended to the case of two words, when intimately connected. Thus, the word, which, when it begins a sentence, or follows a word not ending in a, is written stu (or sthu,) is written astu, when preceded by a word which terminates in a. It occurs frequently in both forms, both in the Great Inscription at the India House and in Botta's Plates. Compare, for instance, VII. 45 of the Great Inscription with the 21st line of Rich's barrel. They are identical, save that the former, which follows L. 260, du, begins with L. 90, s; the latter, which follows L. Ill, ma or va, begins with L. 222, as. Now, it is in strict conformity with this rule of Babylonian writing, that the proper name Gumâtâ, when preceded by a demonstrative pronoun terminating in â, should take the character agu as its initial, in place of gu which it would have in a different position; as, for example, after anku, “I am,” as the name of Cyrus (Kuras) is found at Murgâb. I feel confident that the difference between L. 113, and L. 62, is not that the former is ku and the latter gu; but that the former is ku or gu, and the latter aku or agu. I believe that the best mode of transcribing Babylonian words is by using apostrophes for supernumerary vowels. Thus, I would transcribe the two before us in the following manner â.na.â'Gu.m'â.t.'â. The three apostrophes mark the places of vowels, which might under other circumstances constitute syllables, but which do not do so in the present instance. I read the first word as a trisyllable, but not with confidence. The second character begins with n, and it seems to me that it terminates with some breathing, which would prevent its coalescing with a following vowel. The word “says,” in which it also occurs, seems to me to be the Hebrew the plural used for the singular, as in the generality of instances where gods or kings are referred to by others. It should be observed that in the first person singular, which occurs both at Nakhsh-i-Rustam and in the great lapidary inscription, the final u is omitted, as it would be in the Hebrew am disposed to read these words i.na.am.mu and a.naam, although the second and third characters in the former are often replaced by a single one; which it would be more natural to read nam than naam; but which may, nevertheless, have had the latter value. On points of minute accuracy of this kind, as well as with respect to the value of a few characters which are yet unascertained, the proper names in Major Rawlinson's possession may be looked to, as likely to afford information. The values of the great majority of the characters are, however, in my judgment, already settled beyond the reach of criticism; and I confidently expect that the other Behistun names will be as consistent with these values as I have shown that of the Magian to be. (May, 1848.)

^{page 433 note 1} I find, however, that in the name Warasmis, at Khorsabâd, this character is used after that which denotes ras. The value ami is scarcely admissible here; and I am, therefore, disposed to retract this last observation. (May, 1848.)

^{page 434 note 1} In M. Botta's Paper in the Journal Asiatique for May last, which (though a year published) I have but just received, this is remarked as a possible coincidence; yet M. Botta seems to reject this reading of the word before us. He has, however, correctly taken the initial character of the Khorsabâd word for a determinative prefix. (May, 1848.)

^{page 435 note 1} Since this was written, I happened to observe that the initial character in this word is almost identical in form with that which terminates the name corresponding to Zharaka in the Third Persepolitan inscription N.R. The preceding character has disappeared from that inscription; but I confidently restore it as on the authority of one of the Khorsabâd inscriptions, where the name also occurs. This name I read Zhâranga; and, accordingly, I give the value ha or ga to the character before us. It is probable that p and v were expressed alike; and thus the word may be read gavâ-gana, “the cow-kind,” being a compound of two well-known elements, which occur in most of the Indo-Germanic languages, including our own. (April, 1848.)

^{page 435 note 2} As there is another representative of the Lapidary 249, and another character, besides that, in the Van syllabary, having the value yâ, I have no doubt now that the true value of is râ, which combination would otherwise have no representative. It is the Lapidary 95, of which seven variations are given in the table; it is also formed in several ways at Khorsabad; and one of these is almost identical with that before us. In the Third Persepolitan it is written That the value of these characters is ra is quite certain. The last occurs at Nakhsh-i-Rustam in the names Zharanga, and Phrat; nor is it any objection that it is the penultimate character in the name of Cyrus; for the Median equivalent is certninly to be read Kuras; the final character, to which I formerly gave the value rus, being of like form to that which at Khorsabâd represents ras in the well-marked name Warasmis. The word before us is thus asbârâ; which is the old Persian word for “horsemen.” This would lead to the inference that some of the preceding nouns signified soldiers of different sorts. With respect to dârâ, as I now read it, I can say nothing satisfactory. (April, 1848.)

^{page 438 note 1} If the following word signify “horsemen,” (see note in § 40,) it is more likely that what I have here enumerated were “foot soldiers (men of the camp?).” (April, 1848.)

^{page 440 note 1} I find I have committed an error as to this last compound name. It is not “the dwelling of Sennacherib,” (Sankirib, as I would read the name,) but the dwelling of some god, whose name I cannot yet read with any confidence. The vertical wedge which distinguishes names of men is not found in this name. (April, 1848.)

^{page 440 note 2} Within the last few days I have received the Numbers of the Journal Asia-tique for 1847, containing six papers'of M. Botta on the Khorsabâd inscriptions. From these I have endeavoured to deduce something satisfactory as to the character which I have here considered. M. Botta says that he has found it interchanged with the character which he numbers 82; and this again has been interchanged with which at Nakhsh-i-Rustam by itself denotes “man,” and which is also used as a determinative prefix before wan, which has the same meaning. Now, (though I must protest against a general admission of M. Botta's alleged equivalences of characters; they being not observed facts, but inferences from such, and being very often rashly drawn, and even palpably erroneous,) I believe that these characters are really used alike, and that the meanings are nearly the same. On comparing the two parts of XVII, I find that where was used in one, a character almost identical with M. Botta's No. 82 was used in the other. I have not produced it in the table at the end of this Paper, because I have confined that to characters which occur in the later inscriptions, from which, owing' to its great size and intersecting wedges, it seems to have been rejected. The conclusion, then, at which I have arrived is, that is identical in meaning with or at least so similar to it that it may often be used for it. It by no means follows, however, that it was phonetically equivalent to it. The value of the latter was ba or wa, and it is probably used as an abbreviation for the known word wan; but the other may be, and probably is, as distinct from this in sound as ath or as is from in', with which, however, it is interchanged in the great lapidary inscription, as well as at Khorsabâd and at Persepolis. I have as yet no clue to its phonetic value. M. Botta speaks very confidently of the identity of the two names of kings which I have mentioned above, which, however, in opposition to all other inquirers that I have heard of, he denies to be names. He says that both forms occur “ in the same monument,” (vol. X., p. 315). In none, however, of the Assyrian monuments that I have seen is there any such repetition of the king's name in connexion with the word “says,” as we find in the Achæmenian and Van inscriptions. As in the great lapidary inscription, the king's name occurs at the beginning of each monument and not again, except it may be required in some statement of fact; and here the pronoun of the first person would more naturally be used. The name of Nebuchadnezzar occurs six times in the great inscription; but at least four of these belong to the first king of that name. We have in one place, “of me and of Nebuchadnezzar king of Babylon, my father,” and similar expressions occur elsewhere, where the name appears. Now, as in'all the inscriptions from Khorsabâd that have been published in the Journal Asiatique, the king's name is only found in the initial formula, “The palace of —;” and in the compound proper name above-mentioned, I own that I receive with great distrust this statement of the two forms being used as equivalent. If the shorter form be used in this formula, not merely on the bricks, nor on “the reverse of the plates,” but indiscriminately with the larger one in the finished inscriptions, I must then admit the phonetic equivalence of the two. Even then, however, I should suppose the king mentioned here to be as distinct from the father of Sennacherib, as the son of Arrasvis in the Van inscriptions is from the father of Spuinas. The notion that the father of Sennacherib was king of Babylon, to say nothing of the other countries mentioned in the inscriptions, is so inconsistent with the established facts of history, that I cannot admit it for a moment. (May, 1848.)

^{page 442 note 1} I was not aware till this was in type, that M. Longpérier had anticipated me in reading the name of Assyria in the beginning of the Khorsabâd inscriptions. I have just seen his letter in the Journal Asiatique for December, 1847, in which he mentions having so read it in the Revue Archéologique for 1847, p. 504. It appears, too, from this letter that he noticed in the same paper the similar relations between the single and double hieroglyphic leaves and the single and double which I mentioned in § 24. (May, 1848.)