10 



ETIINOLOGT OF THB IKOO-PACIPIC tSLAK^DS. 



of the ftndont use of tho double ciefiuitive as tbe unit. The Darfnr d-ik 

 is a varmtion oi the aainc form ur of the connate (d). I'he Pi^alagaisy 

 i-iii, i-io, i-ai corresjjonifj with ttie Goii^a sibilant form and the Se- 

 mi rio liij ta kc. In Aaomna several vanetiea are |)reserved, e-aa, i-ae, 

 a-si, a-saj i-cba, ji, i-tn, ta» tu-ttf, la-bl, ta-ka, sa-ila, sa-ra, i«-ra, 

 ai-ti, 7Tte-[svL, w«-wa, m-Bui, sn-mo-^\ SiC. Nome of the Nigerian lan- 

 guaee* retaiu tbe nasal jircfix in iiif^her imnibers, and prove that it was 

 prenxed to tUe i^umerala througU«iut in aome 9ji sterns. Walofhasit 

 in 2} d and 4.1 Some other Nigt-rian tjyatatna had tlie kbbi ftrefix. 

 It ia pre^^erved ia some of tbe A^bantl and Uubun diulectij mi, 

 mie, ba &c.) lo the Ziiubian systi^ma subatuntival prefixes are used 

 in tbe subfitaritivd lorm of tbe auinerals, the ftojisessive reixlering them 

 ordinal When nsed as cardiiia^a or qualitivea tiiey take the iie> 

 finitive, prefix of ill© connected suhgtantive. Thus in Kosnh 

 1 has the substantive form wi-nje (so is/-biiii 2, id-tatu 3 &c.) and 



the qualitive forms um-nye, i*-nye, in-uye, lu-ii\t% Lit* 



ijye, Au-nye, and higher nujnbera take the plurals aba—t ama — , e- 

 , en'—, mi—. In some at the publi-shed Jists the preHxes are 

 coniounded with the roots, particularly in the common Lonlracfed 

 ibruisof tbe prefix as in mirye 1, mbint 2, ritutu 3 Stc, In other cases 

 the eonauuaut of tbe prefix is elided and I be vowd only pn served. 

 \1 hen it h recollected ibt*t in the archaic Seinitico-Libyfin formaiion 

 definitives which tii their variatians embraced tbe whole ran tie ot con- 

 fonanta, might be u?ed either as prefixes or poslfixes, and that both iire 

 fotiail in many wordjj, the diffieuity of analysing and comparing the 

 Semiticu-Atn^an avMerns will he understo.id It i* only where a con- 

 aiderable number of concurring Ifnuta are obtainable that satisfactory 

 conclusions can be arrived at« 



The common Ztmbian term appears to preserve the same form of 

 the unit, with the labial definitive aa ibe initiHl, although the latter haa 

 beiome a 8ub»tantive part of the root. The moBt common 

 form of the uhiuiate root is si, variable to jt, ji &c, and cor- 

 responding with the jMalagtiBV si and Artibic bi| [comp. the £^., 

 Galla and Malagasy def. si, zi] but larms in a and o also uccur as 

 in Semitic, (ionga ajid Malaipat^y. In some languBges & become t 

 and iit utberii r or 1. Thelabiitl preRxaal element ha» generally the 

 form mo. The same combinatton is fuund in other formations, nud 

 wbetltet' both tbe definitives arc to be consitlered m being primarily a 

 def. cnuipouitd used m the numeral, ur a mertdy servile function is to 

 be ascribed to the postfix or prdiv, must be doubtrul in most coses, the 

 rein live pojitiotf of the priticipaf and accessory definitives having va- 

 jri«»*l even in (he same formation* If the original Jorm ol A were 

 wa-hl, wa-kha &tj., it would f*dk>w that all tbe i'eniitico- Libyan Ibrjna 

 ot tbe sibilant, nspinite^ dental and guttural definitive and unit might 

 at one time preHic tha labial. But in tbe archaic non*co«creted con- 

 dition of tbe <|]os3ary it in drnf that each deft ui live had a separate 

 currency and was capabie of being used as the unit. The coiubtnaiiouai 



f Mf. KoeHe'd Poli/glotta Africante, received nince the text w»ft writ- 

 ten, eri&hlet ma to muke Motne udhiiionft. He eiveA ya bp the Wob'tT 

 pr«fix. lu Ballotn m*(euph nin-, ntm 1 ii prefixed to the 5 iioiple nu. 

 mertla. 



