^{page 504 note 1} In xxxvii. 26 we find: Mu-ru-ba(-a-ni) ma-ri-ni kar-bi agunu-ni-ma gunusà khaubi ‘Murabas, (and) his [ma] … statues (?) (and) goods for a spoil I captured.’ Here the word occurs without the determinative of ‘stone.’

^{page 505 note 1} So Dr. Lambrino from another fragment of stone.

^{page 505 note 2} So Layard.

^{page 509 note 1} Possibly we should read (ru-ḳu)-si -na.

^{page 510 note 1} So Rassam's squeeze.

^{page 512 note 1} The general sense of the passage will he: ‘The temple-lands of Khaldis (and) the gifts coming from each for these gods (are) the offerings of Mermas, both for sacrifice to Khaldis the …, and for the gods who belong to the shrine, consisting of lambs.’ If kuruni can be taken as a 3rd person singular, the construction will be simpler: ‘The temple-lands of Khaldis he has given for each of these gods as the offerings of Menuas,’ etc.

^{page 513 note 1} The general sense of the clause must be: ‘The … of Menuas Khaldis has marked out (or consecrated) among the gods that are here (and) among the kings hereafter, who assemble the people, as the eating-place of t?e chapel of Menuas.’

^{page 515 note 1} Layard (incorrectly) ra.

^{page 515 note 2} Layard (incorrectly) ta.

^{page 515 note 3} So Layard.

^{page 517 note 1} Can this armtmi-s he the origin of the name of Armenia which first makes its appearance in the Akhæmenian texts, —Armina in Old Persian, Harminuya in the Amardian or ‘Protomedic’ transcript? The uncultured eastern neighhours of Ararat might well have called it Armana-s ‘the land of writings.’

^{page 520 note 1} See Moriz Schmidt.

^{page 522 note 1} There is another possibility, however. Turis may be connected with the root of terubi ‘I set up’ or ‘established,’ and signify ‘a monument.’ Turie and turi would then be the accusatives plural of this word, while turi-ni-ni would be the ace. sing, of an adj. in -nis. But there are many probabilities against this hypothesis. An ace. sing like turi-ni-ni standing by itself is unprecedented, and the place which it occupies at the beginning of the line in which the names of the gods are written is against its being construed with the preceding verbs. Moreover, the parallelism of ainei seems to show that turie or turi is a genitive-dative, which is further confirmed by the use of the form turi.

^{page 524 note 1} Layard has ni.

^{page 524 note 2} So Layard.

^{page 524 note 3} So Layard.

^{page 524 note 4} So Layard.

^{page 524 note 5} Layard has The character is probably ni.

^{page 524 note 6} Schulz and Layard have , which may be ni, ir, or śa. From 1 20 we learn that it is ni.

^{page 525 note 1} So Schulz and Layard. Da, however, is required; see xxxiii. 23.

^{page 526 note 1} For si-hu-da a-da see note on xiii. 2.

^{page 529 note 1} Omitted in t?e duplicate text.

^{page 531 note 1} So Layard.

^{page 537 note 1} So Layard.

^{page 537 note 2} So Layard.

^{page 537 note 3} So Layard.

^{page 537 note 4} So Layard.

^{page 538 note 1} Layard has ‥ .

^{page 538 note 2} Schulz has tar.

^{page 538 note 3} Layard has mu, Schulz a character which resembles mu more than i.

^{page 538 note 4} The copies have .

^{page 539 note 1} So Layard.

^{page 539 note 2} So Layard.

^{page 539 note 3} So Layard.

^{page 542 note 1} Robert has gi.

^{page 542 note 2} Schulz has .

^{page 542 note 3} Schulz has .

^{page 542 note 4} Schulz has du, but xlv. 16 shows that we must read al.

^{page 542 note 5} Schulz has , which looks like the determinative of ‘man,’ followed by the character for ‘ox.’

^{page 542 note 6} Schulz has la.

^{page 545 note 1} Mordtmann makes kuruni (which he reads turuni) ‘dedit,’ but renders karuni ‘he was.’

^{page 546 note 1} The passages are: xxxiii. 14, (khau-)ni Puteria aśuni ‘who has conquered of the city of Puterias the neighbourhood (?)’; 1. 15, kudhubi pari Musanie Zapsa aśuni ‘I departed out of the land of Musanis (and) the neighbourhood (?) of the city of Zapsa’; xli. 18, Bikhurani aśunini Bamni ‘the city of Bikhuras (and) the county of Bam belonging to the neighbourhood (?).’

^{page 550 note 1} Perhaps e. See line 14. The next word may be nuna-bi ‘I attacked.’

^{page 550 note 2} Layard has li.

^{page 550 note 3} Layard has se.

^{page 550 note 4} Layard has i.

^{page 551 note 1} Layard has li.

^{page 551 note 2} So Layard.

^{page 551 note 3} Layard has kid.

^{page 551 note 4} Layard has an.

^{page 551 note 5} Layard has ni.

^{page 551 note 6} Layard has li.

^{page 551 note 7} Layard has za.

^{page 551 note 8} Layard has ma.

^{page 552 note 1} Layard has ir.

^{page 552 note 2} Layard has li (?).

^{page 552 note 3} See note on this line.

^{page 552 note 4} Probably an error for some numeral.

^{page 552 note 5} Layard has .

^{page 552 note 6} Layard has kid.

^{page 552 note 7} Layard has .

^{page 552 note 8} Layard has kid.

^{page 552 note 9} See note on this line.

^{page 552 note 10} Here expressed by .

^{page 552 note 11} Layard has li.

^{page 553 note 1} Layard has li.

^{page 553 note 2} Layard has li.

^{page 553 note 3} See lvi. iii. 3.

^{page 553 note 4} Layard has li.

^{page 553 note 5} Layard has li.

^{page 553 note 6} Layard has lu (?).

^{page 553 note 7} Layard has li.

^{page 553 note 8} Layard has me.

^{page 555 note 1} So squeeze and Layard.

^{page 555 note 2} Written

^{page 555 note 3} Schulz has tar.

^{page 555 note 4} So squeeze and Layard; Schulz has kab instead of tar-khi.

^{page 555 note 5} So squeeze and Layard. Schulz has hu.

^{page 555 note 6} Written .

^{page 556 note 1} Written ; apparently for kab , though perhaps for ḳar.

^{page 558 note 1} So Layard.

^{page 558 note 2} Layard has (?)

^{page 558 note 3} So in line 14; Layard has bad (?).

^{page 558 note 4} Perhaps la; see note.

^{page 558 note 5} So in line 14.

^{page 558 note 6} So in line 9.

^{page 558 note 7} The printed text has ma, but the original copy reads na.

^{page 559 note 1} So in line 10.

^{page 559 note 2} Layard has (?).

^{page 559 note 3} Layard has bad (?).

^{page 559 note 4} Layard has the impossible . The last character is plainly da.

^{page 559 note 5} Layard has (?).

^{page 559 note 6} Layard has .

^{page 559 note 7} Layard has a (?).

^{page 560 note 1} Layard has da.

^{page 560 note 2} Layard has a.

^{page 560 note 3} Layard has i (?).

^{page 560 note 4} Layard has .

^{page 560 note 5} Layard has ma.

^{page 562 note 1} Tsolagerd or Zolakert is called Tash-burun by t?e Turks. The first aud last lines of the inscription were published in “Ararat,” Feb. 1870.

^{page 563 note 1} My restoration of these two characters is uncertain.

^{page 563 note 2} Text I. has , text II. .

^{page 564 note 1} The text has a.

^{page 570 note 1} For the inscription of Menuas in the pass of Kelishin see No. lvi. It is possible that No. liv. also belongs to Menuas.

^{page 572 note 1} So Layard and Robert.

^{page 572 note 2} It is yery doubtful whether a character is really lost here.

^{page 573 note 1} Robert inserts the numeral 〈 before ‘palaces.’

^{page 573 note 2} So Layard and Robert.

^{page 573 note 3} So Layard and Robert.

^{page 574 note 1} Layard inserts here e (?).

^{page 574 note 2} So Layard and Robert.

^{page 574 note 3} So Layard

^{page 574 note 4} Schulz has 70.

^{page 574 note 5} Layard has 15.

^{page 574 note 6} Omitted by Robert.

^{page 574 note 7} So Schulz, Layard and Robert.

^{page 574 note 8} So Layard and Robert.

^{page 574 note 9} So Robert. Layard and Schulz have , perhaps ki.

^{page 574 note 10} Robert has ni; perhaps we should read śa.

^{page 574 note 11} So Robert. Layard and Schulz have ge.

^{page 575 note 1} Schulz has e.

^{page 575 note 2} Layard has e.

^{page 575 note 3} So Layard. Schulz has lu for me(?)-ku.

^{page 575 note 4} So Schulz and Layard. Robert has me-i-a.

^{page 575 note 5} Layard has .

^{page 575 note 6} So all the copyists; but ni is required.

^{page 575 note 7} So Layard; Schulz has zu.

^{page 576 note 1} Ga in xxxix. 62.

^{page 576 note 2} So Layard.

^{page 576 note 3} So Layard; perhaps the character is da.

^{page 576 note 4} So Layard.

^{page 576 note 5} Perhaps śudhuḳubi ‘I despoiled.’

^{page 577 note 1} So Layard; Schulz has , contrary to analogy.

^{page 577 note 2} So Layard.

^{page 577 note 3} So Layard.

^{page 577 note 4} So Layard.

^{page 578 note 1} Can the phrase really be: “Among the people of the king and among the priests to Khaldis, Teisbas, Ardinis, and all the gods of the inhabitants,’ Khal-die-di, etc., agreeing with abaibadi? Adabidi will then be the simple substantive adatis ‘service,’ from which adoba for adabia is formed. Ada has the idea of ‘company’ in adaieśi li. 4, as well as in the conjunction (‘and’ = ‘along with’). Argistis says that he had built the citadel “for the people of Khaldis” (Khaldia) he might therefore consider that the spoil he brought home was brought back for his subjects. Moreover the words with which the sentence usually concludes (khasi-al-me ) would favour this interpretation.

^{page 582 note 1} So Layard.

^{page 582 note 2} So Layard.

^{page 582 note 3} Schulz has la.

^{page 582 note 4} So Layard.

^{page 582 note 5} So Layard.

^{page 582 note 6} So Layard.

^{page 583 note 1} So Schulz and Layard, but we oug?t to have ni.

^{page 583 note 2} Layard has .

^{page 583 note 3} Schulz has la.

^{page 583 note 4} Schulz and Layard have da.

^{page 583 note 5} So Layard.

^{page 584 note 1} Layard has .

^{page 584 note 2} So Layard.

^{page 584 note 3} Layard has VII.

^{page 584 note 4} So Layard.

^{page 584 note 5} So Layard; Schulz has i.

^{page 584 note 6} Layard has LX.

^{page 585 note 1} Schulz by an oversight inserts al here, which is omitted by Layard.

^{page 586 note 1} Schulz has na.

^{page 586 note 2} So Layard.

^{page 586 note 3} So Layard. Perhaps a character is lost between is and lu.

^{page 587 note 1} So Layard.

^{page 587 note 2} Layard has DC.

^{page 587 note 3} Schulz has i.

^{page 588 note 1} So Layard.

^{page 588 note 2} Layard reverses lines 52 and 53, making line 53 precede line 52.

^{page 588 note 3} So Layard doubtfully.

^{page 590 note 1} So Layard.

^{page 590 note 2} Schulz has i.

^{page 590 note 3} Schulz has ra.

^{page 590 note 4} Schulz has di.

^{page 590 note 5} Schulz has la.

^{page 590 note 6} So Layard; it is omitted by Schulz.

^{page 590 note 7} Schulz has ma.

^{page 590 note 8} So Layard.

^{page 591 note 1} Layard has gis.

^{page 591 note 2} Layard has ri, and omits the determinative .

^{page 591 note 3} Layard has 〈〈 “king.”

^{page 591 note 4} Layard has XM.

^{page 591 note 5} So Layard.

^{page 591 note 6} So Layard.

^{page 591 note 7} So Layard.

^{page 591 note 8} Schulz has ma.

^{page 591 note 9} So Layard.

^{page 591 note 10} So Layard doubtfully. Schulz has .

^{page 591 note 11} So Layard.

^{page 591 note 12} So Layard.

^{page 591 note 13} So Layard. But we must read ti; see xl. 54.

^{page 591 note 14} So Layard.

^{page 591 note 15} So Layard.

^{page 591 note 16} So Layard.

^{page 592 note 1} Schulz has at.

^{page 592 note 2} So Layard.

^{page 592 note 3} So Layard.

^{page 592 note 4} So Layard. Schulz has IV instead of VII.

^{page 592 note 5} So Layard.

^{page 593 note 1} Schulz has ap.

^{page 593 note 2} Schulz has la.

^{page 594 note 1} So Layard.

^{page 594 note 2} Perhaps we should read for .

^{page 594 note 3} So Layard. Possibly ḳar-bi ‘stones’ or ‘monuments.’

^{page 594 note 4} So Layard. If we read i with Schulz, we must correct a into si and so get si-i-hu-bi.

^{page 594 note 5} Layard has .

^{page 594 note 6} So Layard.

^{page 594 note 7} Layard has .

^{page 594 note 8} So Layard.

^{page 594 note 9} Layard has , which can hardly be correct. For Satiraraus see xl. 54.

^{page 594 note 10} Schulz and Layard have kha doubtfully.

^{page 594 note 11} So Layard.

^{page 595 note 1} So Layard.

^{page 595 note 2} So Layard.

^{page 597 note 1} So Layard.

^{page 597 note 2} So Layard.

^{page 597 note 3} Layard omits; Schulz has .

^{page 597 note 4} So Layard.

^{page 597 note 5} So Layard.

^{page 597 note 6} So Layard. See xxxvii. 26 (ma).

^{page 598 note 1} So Layard.

^{page 598 note 2} So Layard.

^{page 600 note 1} So Layard.

^{page 600 note 2} So Layard.

^{page 600 note 3} So Layard. Schulz kas a miaformed ki.

^{page 600 note 4} So Layard. Schulz has .

^{page 601 note 1} Layard has XM.

^{page 601 note 2} So Layard; Schulz has DC … VI (?).

^{page 601 note 3} So Layard.

^{page 601 note 4} Line 35 shows what has to be supphed.

^{page 603 note 1} So Layard. Schulz has Ni-ku.

^{page 603 note 2} So Layard. Schulz has .

^{page 603 note 3} So Layard. Schulz has hu.

^{page 603 note 4} Schulz has ma.

^{page 604 note 1} Schulz haa mu.

^{page 605 note 1} So Layard.

^{page 606 note 1} Schulz has la.

^{page 606 note 2} So Layard.

^{page 606 note 3} So Layard.

^{page 606 note 4} So Layard.

^{page 606 note 5} So Layard. Schulz has tar.

^{page 606 note 6} Layard has ra. The character may be da or ta. See line 80.

^{page 607 note 1} These two characters probably represent only one, perhaps uz.

^{page 607 note 2} .So Layard; the word is omitted by Schulz.

^{page 609 note 1} So Layard.

^{page 609 note 2} So Layard; Schulz omits.

^{page 609 note 3} So Layard. For the completion of the sentence see 1. 25.

^{page 609 note 4} So Layard.

^{page 609 note 5} So Layard.

^{page 609 note 6} So Layard.

^{page 609 note 7} So Layard.

^{page 609 note 8} More probably ‘in one year.’

^{page 610 note 1} So Layard. Schulz has mu.

^{page 610 note 2} So Layard. Schulz has śi.

^{page 610 note 3} So Schulz. Layard has ku.

^{page 610 note 4} So Layard. Schulz has a.

^{page 610 note 5} So Layard.

^{page 612 note 1} So Layard.

^{page 612 note 2} So Layard. Schulz has za.

^{page 612 note 3} Should probably be śi.

^{page 613 note 1} So Layard.

^{page 613 note 2} So Layard.

^{page 613 note 3} Layard has a.

^{page 613 note 4} Schulz has ma.

^{page 613 note 5} Layard has ni.

^{page 613 note 6} So Layard.

^{page 613 note 7} Schulz places a lacuna after a; not so, however, Layard.

^{page 613 note 8} So Layard.

^{page 613 note 9} So Layard. Schulz has pa. With this reading the translation would be ‘out of what belongs to the country of Gurḳus.’ But there is no other example of the adjective in -ue being used thus absolutely.

^{page 613 note 10} Schulz has tar.

^{page 613 note 11} So Layard.

^{page 614 note 1} So Layard.

^{page 614 note 2} Comp. line 78.

^{page 614 note 3} So Layard.

^{page 614 note 4} So Layard.

^{page 614 note 5} Omitted by Layard.

^{page 615 note 1} Layard has se.

^{page 615 note 2} So Layard.

^{page 615 note 3} Layard has e.

^{page 615 note 4} So Layard. Schulz has i. We should probably read ‘in one year.’

^{page 616 note 1} So Layard.

^{page 616 note 2} So Layard.

^{page 617 note 1} Layard has i.

^{page 617 note 2} Comp. xxxi. 11.

^{page 617 note 3} So Layard.

^{page 617 note 4} Schulz has ma.

^{page 617 note 5} So Layard.

^{page 618 note 1} So Layard.

^{page 618 note 2} So Layard.

^{page 618 note 3} Layard has 〈〈 .

^{page 618 note 4} So Layard.

^{page 618 note 5} So Layard.

^{page 619 note 1} So Layard.

^{page 619 note 2} So Layard.

^{page 619 note 3} So Layard.

^{page 619 note 4} So Layard.

^{page 619 note 5} So Layard.

^{page 620 note 1} Layard has ni. Perhaps we ought to read .

^{page 620 note 2} So Layard.

^{page 620 note 3} Schulz and Layard have .

^{page 622 note 1} Schulz has la.

^{page 622 note 2} Schulz has .

^{page 622 note 3} Schulz has .

^{page 622 note 4} Layard has ‘seed-seed.’

^{page 623 note 1} So Layard.

^{page 623 note 2} For the country of Bias see xxxvii. 8.

^{page 623 note 3} See the same passage.

^{page 623 note 4} Layard's copy marks no lacuna here, hut the verb is required by both sense and grammar.

^{page 625 note 1} Or ya.

^{page 626 note 1} Layard has se.

^{page 626 note 2} Layard has .

^{page 626 note 3} Layard has se.

^{page 627 note 1} So Layard, but xliii. 48 shows that we must read .

^{page 627 note 2} Layard has khi, but see xlix. 17.

^{page 627 note 3} See p. 568.

^{page 629 note 1} Layard has e.

^{page 629 note 2} Layard has (ab or dha).

^{page 629 note 3} Layard has ba.

^{page 629 note 4} Layard has khi.

^{page 630 note 1} Layard has It may be hu.

^{page 631 note 1} Layard has e.

^{page 632 note 1} For this restoration see xxix.A., etc.

^{page 633 note 1} So squeeze and Layard; Schulz has e.

^{page 633 note 2} Schulz has la, Layard se, the squeeze ti.

^{page 633 note 3} So Layard.

^{page 633 note 4} So Layard.

^{page 633 note 5} So squeeze ana Layard; Schulz has lu.

^{page 633 note 6} So Layard.

^{page 633 note 7} So Schulz; Layard has khi.

^{page 633 note 8} Schulz has la.

^{page 633 note 9} Schulz has tsi.

^{page 633 note 10} For the restorations of the text see Ko. xliv.

^{page 634 note 1} So Layard.

^{page 634 note 2} So Layard; Schulz has e.

^{page 634 note 3} Restored from li. iii. 4.

^{page 635 note 1} Schulz has gis, Layard na.

^{page 635 note 2} So Layard, Schulz has e.

^{page 635 note 3} So Layard aud Robert; Schulz has ma.

^{page 635 note 4} Schulz has śi.

^{page 636 note 1} So Layard and Robert.

^{page 636 note 2} Schulz has la.

^{page 637 note 1} So all the copies. But analogy seems to require mu ‘their.’

^{page 637 note 2} So Layard; Schulz omits.

^{page 637 note 3} So Layard and Robert; Schulz has ma.

^{page 637 note 4} So Layard; Schulz has nu. Perhaps we should read 〈〈 nu.

^{page 637 note 5} So Layard; Schulz has ma.

^{page 637 note 6} So Layard; Schulz aud Robert omit.

^{page 637 note 7} So Schulz; Layard has tu (?).

^{page 637 note 8} So all the copies, but we should probably read ga.

^{page 637 note 9} So Layard.

^{page 637 note 10} Schulz has se.

^{page 638 note 1} Layard has e.

^{page 638 note 2} So Schulz and Layard; Robert and Sarkisian bave ni.

^{page 638 note 3} So Schulz and Layard; Robert has e-ni, Sarkisian ni.

^{page 638 note 4} So Schulz and Layard; Robert and Sarkisian have ni.

^{page 638 note 5} Layard has du.

^{page 638 note 6} So Robert and Sarkisian; Schulz and Layard have ir.

^{page 638 note 7} So Layard; Schulz has za.

^{page 639 note 1} So Layard.

^{page 639 note 2} So Layard.

^{page 640 note 1} So Layard.

^{page 640 note 2} So Layard.

^{page 640 note 3} So Layard; Schulz has i.

^{page 640 note 4} Schulz and Robert have XXXM.

^{page 640 note 5} Schulz has e.

^{page 642 note 1} Omitted by Grotefend.

^{page 642 note 2} The copy has da.

^{page 643 note 1} So Grotefend.

^{page 643 note 2} Omitted by Grotefend.

^{page 643 note 3} Perhaps we should read bi.

^{page 643 note 4} Grotefend has si.

^{page 643 note 5} Omitted by Grotefend.

^{page 643 note 6} The copy has la.

^{page 643 note 7} Grotefend has za.

^{page 644 note 1} Grotefend has lu.

^{page 644 note 2} A line seems to have been omitted.

^{page 644 note 3} The copy has da.

^{page 644 note 4} The copyist has probably omitted here.

^{page 645 note 1} The copy has da.

^{page 645 note 2} Grotefend has ku.

^{page 645 note 3} The copy has ma; but see xxx. 15.

^{page 646 note 1} Omitted by Grotefend.

^{page 646 note 2} Probably an error for ni. The character rat elsewhere occurs only in Schulz's copy of v. 14–52, where Layard more correctly has ar; and I have, therefore, not ventured to include it in the Vannic syllabary.

^{page 646 note 3} Omitted by Grotefend.

^{page 646 note 4} The copy has an, but we do not find this character used elsewhere phonetically in Vannic.

^{page 646 note 5} The copy has

^{page 646 note 6} The copy has im.

^{page 647 note 1} The copy has an.

^{page 647 note 2} The copy has se-ri.

^{page 647 note 3} The copy has la-me. For the restoration of the text, see xliv. 8.

^{page 647 note 4} Omitted by Grotefend.

^{page 647 note 5} The copy has .

^{page 647 note 6} The copy has .

^{page 650 note 1} Schulz has .

^{page 650 note 2} Schulz has pa.

^{page 651 note 1} Schulz has se.

^{page 651 note 2} Sandwith has at-it.

^{page 652 note 1} Sandwith has te.

^{page 653 note 1} Clearly for ulgusiyani-e-di-ni; see p. 655.

^{page 659 note 1} The copy seems to have ga.

^{page 659 note 2} This character must be wrong. Perhaps e should be read.

^{page 661 note 1} This should probably be i.

^{page 662 note 1} Perhaps mu.

^{page 663 note 1} Rawlinson (Journ. R.G.S. x. (1841) p. 21Google Scholar) says that the pillar of Keli-shin, “which is upon a little eminence by the side of the road, and nearly at the top of the pass,” is “a pillar of dark-blue stone, 6 feet in height, 2 in breadth, and 1 in depth, rounded off at the top and at the angles, and let into a pediment, consisting of one solid block of the same sort of stone, 5 feet square and 2 deep. On the broad face of the pillar fronting the E. there is a Cuneiform inscription of forty-one lines, hut no other trace of sculpture or device is to be seen.” “On breaking away the sheet of icicles with which the surface of the stone was covered, the upper half of the inscription was shown to be irrecoverably obliterated, and the lower half also to be much destroyed.” Keli-Shín signifies in Kurdish “the blue pillar.” At the distance of five hours from this pass, “there is a precisely similar pillar, denominated also Keli-Shín, upon the summit of the second range, which overlooks the town and district of Sidek. This also is engraved with a long Cuneiform inscription; and as it is said to be in far better preservation than the one at Ushneï, it would be very desirable to examine and copy it.” As for the pillar of Keli-Sipan, or “white pillar,” Sir H. Eawlinson describes it as “a rude column of white stone, 12 feet in height, 3 feet in breadth, and 1½ feet in depth, fixed in a pediment, and differing only in size and colour and the want of an inscription from “the pillar of Keli-Shín Ushneï.” “It faces also W.N.W. instead of due E., like the Keli-Shín. There are some rude figures like a horseshoe engraved upon different parts of it, which had been taken by the Kurds for writing. Between this pillar and the village of Legwin, there is another, which is also called Keli-Sípán, but it has been thrown down, and is of smaller dimensions even than the Keli-Shín; this, likewise, on the three sides which are exposed, is without inscription.” The first-named pillar of Keli-Sipan is two miles from Khorenj, overlooking the plain of Lahijan, at the source of the Little Zab, and due south of Lake Urumiyeh. The pass of Kelishin, on the other hand, leads eastward from Rowandiz into the plain of Ushneï on the south-western shore of the lake.

^{page 665 note 1} Or ri.

^{page 665 note 2} Perhaps al.

^{page 666 note 1} Perhaps us. The upper portion of this line is cut off in the cast.

^{page 667 note 1} Must be i.

^{page 667 note 2} Or perhaps ri.

^{page 667 note 3} There is a fracture here, and the two characters may be te-ru.

^{page 668 note 1} Possibly da.

^{page 668 note 2} Probably ri.

^{page 669 note 1} The last sentence of the note on v. 20 (p. 486) must accordingly be erased.

^{page 669 note 2} The dental, however, it must be remembered, is omitted in iulie or iudae in another inscription of Menuas, xxxi. 28.

^{page 673 note 1} See also W.A.I, i. 36, 27.

^{page 675 note 1} It may be meant for sum or even tur, or perhaps for the two characters tab and ru (du).

^{page 676 note 1} See Burnouf, , Mémoire sur deux Inscriptions cunéiformes, p. 176.Google Scholar

^{page 677 note 1} The text as copied by Schulz is published in the Journal asiatique, 1840Google Scholar. Another copy of the Persian text by Boré has been given by SirRawlinson, H. in the J. E. A. S. Vol. X. p. 334Google Scholar. For Westergaard's copy of the “Protomedic” text as far as line 21 see Norris, , in the J. R. A. S. Vol. XV. p. 156Google Scholar. For the Babylonian text see Oppert, , Expédition scientifique en Mésopotamie, ii. 2Google Scholar. It reads: a-na e-bis liv-su vâ tsalam-ma ina mukh-khi ul is-dhu-ur ‘(he gave command) to make his tablet and an image; thereon he did not write.’ Liv () is the lav ‘tablet’ of the older Assyrian period. Oppert has already recognized that is here the ideograph of tsalamu ‘image,’ as at Naksh-i-Rustem (26), ma being the phonetic complement. See Bezold, : Die Achämenideninschriften, p. 76.Google Scholar

^{page 678 note 1} The only character not found in the Amardian texts is dku in its later Babylonian, form. The inscription is circular, and reads: Ku-'ar-hu-man (‘the king’) Man-dhu-mas. The sound kw seems to have been known in Kappadokia and Kilikia. Among the proper names found in the inscriptions of Hamaxia are Κούας (C. I. G. 4401, etc.), Όβρανγούεις, Κούαλις, Κιδαμούασις, and Έпιούασις, and at Selindy we have Κούαλες (C. I. G. 4424). The inscription probably belongs to the time of Kyros, when the system of cuneiform writing employed in Khapirti or Amardia (the plain of Mai Amir) was likely to hare been carried beyond its original boundaries, as Khapirti formed part of the ancestral dominions of Kyrps, and its language and writing accordingly were those of the conqueror of Western Asia.

^{page 681 note 1} I ought to have mentioned in my sketch of Vannic grammar that participial adjectives in -śis resembled those in -khinis in discarding the accusatival -ni. Indeed they went further, as neither the substantive which preceded the adjective in -śis nor the adjective itself in some instances took -ni. Thus we find inani nuśi ‘royal city,’ nu adaiśi (li. 4). On the other hand, we have ini asê baduśie, Menua-ni … aluśie, and in the dative plural maśi-ni. In my list of Vannic characters, also, kab must be expunged. I now feel convinced that in xxxii. 7 represents ḳar. Accordingly, the word for ‘stone’ will be ḳarbis and not kabbis. I have to add, too, that in one passage (xxxi. 21) tu is replaced by . In the account of the verb, moreover, I have omitted to notice the causative conjugation formed by the suffix su, sa (as in kharkhar-su-bi, kharkhar-sa-bies). Su probably signified originally ‘to cause,’ ‘make.’ The noun-suffix -tsi, I am now inclined to think, contains the idea of ‘all.’

^{page 681 note 2} Hu and u are arranged alike under u, and ś and ts follow s. Otherwise the order of the letters in the Latin alphabet is followed. Kh follows k and ḳ. Dh comes after d.

^{page 698 note 1} M. Guyard believes inanida to signify ‘spoil,’ and would translate the stereo, typed phrase: inanida arniusinida . X zaduni ‘the spoil and tribute in one year X has made.’ Da is used to form abstracts, and as the phrase occurs at the end of the account of a campaign and an enumeration of the booty carried off in the course of it, it would seem naturally to be a summing up of what had been done. Moreover, in xlix, line 29 appears to be parallel to line 26. The repetition of the statement that the king had built the citadel of Van is unmeaning, and arniusinida could easily be derived from the root ar ‘to bring.’ If the interpretation of M. Guyard is adopted, inani in the execration formula would signify ‘property.’ On the other hand, zadu means properly ‘to build,’ and arnuyada seems to be ‘castles.’ But the expression aluśi inani, parallel to aluśi Dhuspae, in lvi. iii. 5, really settles the question. See also xlix. 11. We must further remember that -ni is the phonetic complement and that the root of inas and inus is probably the same, while no verb from a root in, signifying ‘to spoil’ or ‘carry away,’ has been found in the inscriptions. In the execration formula the ascending scale is naturally: the offender himself—his family—his name—his city. For his ‘property’ to come at the end, instead of ‘city,’ would be an anti-climax.

^{page 724 note 1} In one or two cases, it will be noticed, the pronunciation given on pp. 421, 422, has been corrected.

^{page 725 note 1} It is possible that aris is the full word, from which arani is formed, -ra would stand to the suffix -ri as -na to -ni and -ka to -ki. If we compare gissuris, e-uris, the root would be a ‘to speak,’ so that ‘man’ would be ‘the speaker.’ The legend of Ara (Aras) seems to show that the word aras actually existed in Vannic (pp. 414–416). The solar hero of Van might as well be called ‘the man’ as his Phœnician representative Adonai, Adônis ‘the lord.’ Compare also the name of the king Arame(s) or Arave(s), perhaps Ara-hu-e-a (p. 405).Google Scholar

^{page 730 note 1} The compound ideograph “heaven” must be excised, and “citadel” substituted for it.