34 THE INDIAK ANTIQUARY. [Februart, 1875. as well as many other inflections, owes its existence to the degradation of entire -words gradually incorporated into the terms they are intended to inflect. The external plural seldom occurs an the abso- lute state ; it is indicated by the terminations », •, and p. In the names for the decades the • oc- curs constantly, e.g. nzo 20 (Wr. 5), raw (JT. Q. 1, 10) or vasv (ff. 199, 1) 40, too 70 [H. 3, 4). The letter 3 is probably the characteristic sign of the plural in the other words : p^m {H. 3, 3) • mercifnl (gods).' The * of the plural does not disappear before another termination, e.g. (jn)iiS 'the houses' {tl. 667, 2; Ab. i. 11), (p)^ {H. 373, 4). The plural in -dt tc occurs even in words not terminating with n in the singular (o)nV»j {Os. xxxv. 6), (4-m {H. 63, 5), rrpoo {U. 16 (ao)mrfo (H. 4S4, 9), and with internal modi- fications : noTQ {Os. xxxi. 3), rom* (H. 51, 7), from apn {Os. xx. 9) andtrtu {Os. xi. 3). The reduplication takes place in nW« the AUlat of Herodotus, originally h» Ilos, the Semitic Kro- no8 t then by extension f god.' The Mimean texfl often show nrr, eg, (>)nnv {H. 666), {j);tto {H. 361, 2; 362,2,3), (o)™ (S. 395, 2), (l)rvron (io. 403, 2). The various forms of the internal plural are not distinguishable in the consonantal writing • the form occurring most is hstn ( Jy& 1 3 JUih (a)iw {Os. xxxi. 3), (n)ccEK {H. 468, 3), Smh (*V. XLV. 2), (j)-k* (0*. it. U), («n)Tvw (J xni. 8) ; tp s (0», xnii. 5) probably = ^tss, asmtf. There are also examples for the plural of the plural: (pjxtns (#• GGG), fan)n-rrM (Os. xijj. 8), B)rr3pM {lb, XX. 3). The ijod is the characteristic for the status conslncctivus of the external plurals, so that gra- phically the plural and the dual arc both the same, e.g. q {Os. xvm. 3), (m)-™ (0*. jx. I), ~fa {Os. xxkt. 5) wfjw vrw 'the gods and goddesses of...' {Os. xxix. 6). The T/od is sometimes sap- planted bya I, e.?.i:a(0 S .ix.2 ? &c.),(v,^^rf S o (.46. i. 11-12). It may be seen that no regard for cases exists, contrary to the usage of the Arab*, ft seems also that the use of the form vn is limited only to the names of tribes, like ditto «b {Os. 1.1; IV. 1), UTM , :a (ft. , x , 2 ; XI. 3), pm ra (io. xviu. 2), &c. The Arab grammarians, who were struck by the termination in m of many Hemyaritic and indigenous proper nouns, hare justly considered it as the apocope of 'fj>> ■•<'!.,' thus imparting to the name to which it is added an indefinite sense ; in short, the m is a sign of indetermi- nation. The Sabasan mimmation in general follows the same rules with the Arabic tamvin, e.g. c cjv s w (cDScnr, Os. 5. 1)= <_r*-», crrr {Ab. r. 5) = Bit? 5 (I - Heb. re; cmi {0$. I. 1 1) =jjJ } cm — <-*a^ - crfia (-ff, 478, 16), = J^ ? ( roi , = /j/j.) ^ ^ DpTti (l"6.) = OjA , C3T5Q (/£,) = ^j*> f CTOr3 {a. C81, 8) = <u*J, ctscq = •>;*-«, and the diminutives bits = «±±*» , nnp (Cto. xin. 1) = &lj* , orta = a±*w, and the internal plurals The following do not always receive the m in conformity with the Arabic tatnrhi: 1st — Proper nouns terminating in h and i: e.g. ks *Saba,* tc? 'Kane, 5 TrOT3 'Kaminakum,' trp 'Karnon.'and the divine name mfiff, the Semitic Astarte; %
- tut
The elative ttoh = «**-r , cstm, dstx ; 3r(^— Pro- per nouns resembling one of the inflections of the Imperfect, or rather the third person of the Perfect : ttct or yir. "At, ;i-^. -err, ^ r? nTO nacp, nnm, ni'ia, rnn ; 4f/r — Proper nouns termin* ating in y, pay, pnp, 'Katabani,' -:,■ 'Gedta- nita>' jM3j 'Gebuutn.' These rules nevenlic- less have many exceptions, and the use or omission of the m appears to depend on lot ul usage. Thus we meet with rt«, - : - ; nn-nan ( Hadramau t' by the side of bw, ; the omission of m is so frequent that it is .supt i tluous to give further examples. As a sign of indetermmation the «n must natu- rally fall away in tho statw .- ,,, w IVVv the first word is closely connected with the following one, and thus obtains a detenu i- sense: nar» m {U. 5»7, 1) ' I he house, temple uf Attar,' *«? x?cm ' the peoples of Saba/ nsra -jH, «tho kings of ilain, i.e. of the Minieans ;' nor can the m occur before the per- Bona] suffixes mar, vnpa &c. Tho linguistic problem here presents itself: Does the Sabamn language possess a definite ar- taole, like nearly all the northern Semitic idioms, or has it none, like the Ethiopic ? Osiander after a minute investigation decided that the Sabnean
