с

In Assyrian itself an interchange of c with k and g sometimes occurs; e.g. cabru and gubru; kinnātu (“female slave”) is on Michaux's stone cinātu. The latter was apparently the vulgar pronunciation common in Babylonian. In fact, in the Babylonian, g commonly takes the place of k, e.g. gatu for katu, sangute for sankute, "chains." This pronunciation began to prevail in Assyria in the later days of the empire. Dr. Oppert remarks that c seems to have had a softened sound, which assimilated it to the Hebrew ; thus Tukulti-an, S'arru- cinu = 1, Sacanu = D. Tiglath, however,

סרגון

סגן

answers to the Assyrian tiglat or tigulti; Sargon is not the Assyrian S'arru-cinu, but the Accadian original S'argina; and Sacanu, by which, I suppose, either Sacnu or Saccanacu is meant, was non-Semitic. On the other hand, c and to one another in Nebuchadnezzar, Calah, and Accad.

answer

L is the pronunciation generally, though by no means necessarily, assumed by a sibilant before a dental e.g. khamistu and khamiltu (“five”), istu and ultu (“from”) asdhur and aldhur ("I wrote"), astacan and altanan ("I fought"), lubustu and lubultu ("clothing"), mikhiztu and mikhiltu ("strong"). L, however, never becomes a sibilant. This change of consonant, peculiar to Assyrian, must have been effected through r into which the sibilant first passed. Compare the mutations of final s in Sanskrit. Before a second hard sibilant, & may also become l; as in ulziz for usziz. In common pronunciation / seems to have been somewhat mouillé ; thus verbs ending in 7 generally have a attached even in cases which would hardly permit the conditional suffix, e.g. aslula; while on the other hand the case-terminations are sometimes improperly dropped before a following 1, as in ana gurunit lū agrun, "to a heap I heaped."

M usually, but not always, becomes n before a sibilant, a dental, or a guttural: thus we have dhen-su from by, khansa and khamsa ("five"), khandhu (DD), tsindu and tsimdu (“a yoked-chariot "), muntakhitsu for mumtakhitsu ("fighting"), dumku and dunku ("lucky"). In this way is explained the change of the plural-ending into », like the change of mimmation into nunnation. So in Ethiopic, m before dentals and labials passes into n. A double b or p may be replaced by mb, mp (e.g. inambu for inabbu), and a double dental by nd, nt, (e.g. inandin for inaddin) just as in Ethiopic. M first changed to n can be assimilated to a following consonant, as in ikhkhar from makhiru, takhatsi (for takhkhatsi) from

.

N, as in Hebrew, is assimilated generally to the following radical. This is the rule with verbs, though we meet with inandin for inaddin ("it is given "). Contrary, however, to Hebrew, n is assimilated (regularly) before t and 8; e.g. limuttu for limuntu (“injured"), libittu for libintu (“ brickwork "), maddattu or madattu ("tribute") for mandantu (mandattu is found); cissu ("much," "collected") from

for אַף אֲמֶנֶת for אֶמֶת So in Hebrew .(כנש .Targum) אֶנֶף

S' rarely represents a Hebrew as in khurśanis from , śi'amun, siba, "seven" (W.A.I. ii., 19, 66). Where the Hebrew has and V, Assyrian also has έ and 8; e.g. sarru and saru (“king”), cabiś and cabisu, "trampling." S' seems to have been preferred by the Assyrians, ≈ by the Babylonians (see suprà). Just as the example of = Sarru-cinu shows that the Assyrian pronunciation of s was hard, so the fact that t-s is frequently expressed by s points to a similarly hard pronunciation of the latter.

Thus

sarrut-su ("his kingdom ") is also spelt sarrusu, kat-su (“his land") becomes kaśśu for ka-su. The difference between ś and ts in Assyrian was probably that between t-s and t-z. Hence a final dental followed by the sibilant of the third personal pronoun is very commonly represented by ; e.g. dannuśu for dannut-su, illaśu for illad-su. Conversely, & followed by the dental of the secondary conjugation is often written (like st pronounced sh in the tenth conjugation of the Mahri), as aśacan for astacan, aśicin for asticin, aśarap for astarap. Probably, however, the sibilant in vulgar pronunciation changed the place of the dental, just as, conversely, in Hebrew the dental of Hithpael followed the sibilant. In the inscription of Khammurabi z takes the place of έ in tsirrazina tsirrasina (tsirrat-sina). This hard pronunciation of s would once have been universal among the Semites, as is implied by the Greek pronunciation of samech (§). The interchange of s and s in Assyrian (mostly in the later inscriptions) would show that a softened pronunciation was becoming usual. Similarly in Babylonian we find usalbis-su, "I covered it" (for su).

or

E is always a vowel. Occasionally, however, it answers to a Hebrew ; e.g. ecilu ("place")=Aramaic Sp, rēcutu in Babylonian. There must, therefore, have been a time when the guttural pronunciation of was known to the Assyrians. E sometimes replaces (as in erinu=); conversely we have ra-‘a-du (“thunder”) =, aggullu ("wheel"). Hebrew is also sometimes represented by u or i, e.g. uzalu (“gazelle”)=J'jċ Istar = . Its pronunciation differed but slightly from that of i, as is shown by the interchange of the two

.עשתר

vowels (see suprà), and the fact that many characters have indifferently e and i as their vowel-sound. At the same time the presence of radical e was always observed; verbs with e radical are full. E with u fell away; thus from

we have ul-la-'a in Assyrian, and u-'ul-la-'a in Babylonian. Babylonian sounded it more clearly than Assyrian (so ci-e-nu for ci-i-nu). So from y, Assyrian has epsit, Babylonian e-ib-sit. On the other hand, generally in Babylonian and Achæmenian e was assimilated to i, while in Assyrian the converse took place; thus Assyrian ebusu Babylonian ibusu. In both the weaker sound a was lost before e; e.g. ebus, “I made” (for a-ebus); but a following assimilated even a radical e; e.g. isma'a for isme'a from yo. Occasionally e is interchanged with a in roots, owing to the guttural aspirate

=

.in Hebrew עגם and אגם גאל and געל common to both, like

Thus agu'u, "crown," is given also as e-gu'u in a syllabary, from the Accadian ega (compare ), and eliah and aliah are used indifferently; so erzituv for irtsituv in Babylonian, which often replaces by e an Assyriani, where this has been weakened from an original 'a. In the Babylonian recutu (Assyrian rukutu) u has been lost before e, which here replaces kh.

=

P prevails in Assyrian where b appears in Babylonian, and (often) in the cognate dialects (e.g. Assyrian epis=Babylonian ebis, pursu=). Conversely we have bislu DD. The two sounds interchange in Assyrian itself; thus we find iśkhupar, “he overthrew," śikhubartu,, "overthrow," paldhuti, "surviving," baladhu, "house" (5). In one instance u seems to replace; etstsuru, "a bird,"=Arabic,

.צפור Hebrew

R, though, like 7, sometimes used to form quadriliteral roots (e.g. parsidu, palcitu, iśkhupar), is much more scantily employed than in the cognate dialects. Thus we have cuśśu‘u, not, annabu (“hare”), not

.דשא

.

S was never aspirated, as in ancient Hebrew and Phoenician Samsu. Dr. Oppert gives a long list of words where Hebrew has and , but Assyrian simply s in both cases: sumilu, siptu, sarru=, paskupi, dussūtu, distu = Nvi¬. Already in the seventh century B.C. the Hebrew pronunciation seems to have inclined towards an aspirated ; this would explain the transcription of Sargon, etc., by D. In Assyrian itself we have a word like bis-śu, bis-śate, and in Assur-bani-pal's inscriptions taśbusu is a variant of taśbusu. In Arabic (and Ethiopic for the most part) 8 (6)=Northsemitic sh, and sh-Northsemitic 8 (§). Before a dental, ts might become s, as in marustu (and marultu) for marutstu. So bislu - Hebrew DD, isid=TD". Hebrew,

T servile, in the secondary conjugations, is assimilated to a preceding,,, and (e.g. itstabat becomes itstsabat, "he takes," iztacir becomes izzacir, astacan becomes aśacan. After a guttural, t servile may change to d or dh, e.g. igdamir for igtamir, ikdhirib for iktarib, ikdhabi for iktabi. We find even amdhakhits for amtakhits ("I fought"), according to Dr. Oppert through the influence of the following, though after m or n t more usually becomes d. Bd in Assyrian, again, was regularly changed into pt; e.g. captu, "heavy" (1), aptati, "ruins" (7). There is one instance of d in Assyrian and Babylonian replacing an of the other dialects: is always nadin. T replaces dh in Babylonian in tub-dhub (so in Ethiopic cadana, b, damana=