[811] Transactions, xxii. 25. In this case, however, it is so. Hincks afterwards instanced an, which, followed by ac, he reads, not ‘anac,’ but ‘nabu.’ Athenæum, Sept. 21, 1850.
[812] He explained this matter with great clearness in 1850, when he showed that the sign for an is used (1) as a simple phonograph in some words, as in ‘zarangu’; (2) elsewhere it occurs alone as an independent ideograph for ‘god,’ and forms the plural ‘gods’ by the mere addition of the plural sign. (3) Again, it is found before the proper names of gods, as before Aurmuzd. Here the name is phonetically complete without it, and it is therefore simply a non-phonetic determinative. (4) Elsewhere it forms part of a compound ideograph, and may entirely change its phonetic value; and (5) it may be used ideographically for ‘god’ in Semitic proper names, where its value is not an but ilu, or sometimes Assur.—Transactions, ib. pp. 27-30; Athenæum, ib.
[813] He admits in Sept. 1850 that he had not yet seen Longpérier’s paper (Athenæum, Sept. 21, 1850).
[814] He explained, however, that his short a corresponds to the Greek epsilon (Transactions, p. 10). In his list the consonants followed by ā are really those followed by a, and those followed by a correspond to the consonants followed by i. His view of the four vowels dates from the paper on Van, Dec. 1847 (J. R. A. S. 1848, vol. ix.).
[815] Athenæum, Aug. 24, 1850, p. 908.
[816] Essai de Déchiffrement, p. 11; cf. Journal Asiatique, 1848, xi. 246.
[817] Thus Löwenstern, writing in 1847, before the third column was taken, says: ‘Rawlinson a, durant nombre d’années, interdit au public savant la vue des trésors dont il s’était réservé de faire un usage si utile à sa gloire.’—Exposé, p. 10.
[818] J. R. A. S. xii. 404.
[819] Mr. King has given a list of 329 signs (First Steps in Assyrian, p. cxxxii.). Conder reckons about 550 in all (‘On Hittite Writing,’ J. R. A. S. 1893, p. 829).
[820] Menant, in 1864, reports 6,000 words, Ecritures, p. 256.