TOANSACTIONS OF THE SECTIONS. 
tm ihr Ifcrber verb Meskar, fecit. Yet the translations cannot be absolutely trusted. 
rWpoiiit of interest is, the apparent change from an active to a 
Xto A r™ instence. of thi. Mr. Ne.man g.thcrod m h.. 1 r- 
wiRabul) mminar; and we have hero a hint, that the Ghadamsi may 
'Jiuswitrpun’randoldcrn.ethodBof conjugaUon than the 
^ Arabised. All, however, that appears of the development of the ver » m the 
W^ier of Matthew (to which Lokmana fuhle may b« added), cloatly agreea 
•all rlttt is known of the other dialects. • . <* 
Ibf columns of sliort sentences given by Mr. Ilicbardaon yield one point of eUino- 
'.^1 inwrwt. The Berber ^iherto accessible has failed to impart a posiUvo 
i 'il^dsv of the native pronouns of interrogathm, owing to the cxtciisivc borrow ing 
•:irtluc, even for thi-*e necessiiry ideas. By a process of rather refined re wning, 
-taWunude out that one priuiUive dcmoustradTc particle of Berber Ji Jja, Vjti- 
7i(tkicb b seen even in the Arabic DhaUk, ille, TeUk, ilia); that a sccotid is H «. 
fca. Ta; aliicb sre often used with the sense of a relative. Analysts likewise shows 
ro.v* where the clement ff'a entering words gives them an interrogative foree; wd 
sWiiidiYuIiul yet clear instances exhibit B"i in isolation as an interrogaim*.^ Uc 
Hmwj Bin has Nila ihillun, illud cujus est? where th may seem to W the inter* 
r*-:* nr clement; but this needs fuller elucidation. But there U a third parlich-, 
i-Ki, which enters the system ns a relative. Mr. Newman had already 
Wcbtioii lo ibe formula in DclnpOTtfl's ' Tale of Sabi.' Kon»i k-ayy-tdifkoit, \ ri 
"i-diiu; or, Voi u< mihi dantes; and in the Ghadamsi third cliapter of Matthew 
^ rf yi'Idiun. qui est ventimis ; and apparently also eski means I n quo„w hlcli 
**' T neliU ^i=quj. Wo have thus a sutlicicnt account of the letter K, which nms 
‘'■sp a whole class of Kabail (lemonstrativo and relative adverbs. But the (Iha- 
wntmees now obtained distinctly exhibit as a pure interrogative, Aoy, qulsT 
% qna ? (quid ?) ; if nt least we may presume on the distinction of gender. 
Ghadamsi. 
Arabic (translated). 
Probably more accurate. 
Kay tefrul ? 
Quid Ubi (res est) 1 
Quern tu vis ? 
'fay iefrid ? 
Quid opus tibi ? 
Quid tu vis? 
Tay yinna ? 
Quid dixit? 
fay hadata? 
Quid evenit? 
fuy lakhbur ? 
Quid nfivi ? 
fay tetfad? 
Num quid babes? 
Quid tenes? 
• here comes once only, and the translation is unsatisfactoryf, a* it 
tiwT**? means Amas, Desicleraa; yet analogy puts the sense of Kay bc- 
J s doubt. With this wc must couple the phienomfiia of the Hiinssa language, 
I^ominal system of which has otlier strong veseniblances to the Berber. In it, 
*mtn ‘r°' to this, the Galla has 7'uui, for the Berber 
>1'^. ‘l|a- On the whole it results that the Berbers have the following 
’'e system (besides other demonstratives which need not here he mentioned), 
K T QnisZ qincf 
W T Qui, quae, 
y j DT Ille, ilia. 
^ '2*'^nay fulfil all three functions; AV fulfils the two lust, and iomptimes 
j »wle Oc Heiiiso gives the sentence J)i Icfridt Quid vis? aa (lUadamsl; 
His 1) also as an interrogative. 
ivoid comparing the Indo-Kuropean elementa ko, mi, uho, n, 
*-•1 uwu-!!. j L identity with these of the Berber; especially since the 
Art°i ” L'* Person 5» in Berber and Haussa s, ]*1. sew. in .Arabic A, pi. 
(> j 4 ^ 1 * •’clatrd as Sanscrit san, *a to 6, n- Such phenomena indicate 
connection of these very distant languages, which it remains for 
“"l^rew to trace out. 
f 
fTa/iAm, qiiis ? and it Is quite ^certo, 
^ "Hut which* • fonus, that tliis is to be analysed into jro-/-iAin, rt* euro* w. 
mn;... “ P^cn as the Tuarvk equivalent is pure iVrabic, aud means Quid iu tit} 
»he use of Kay and 2’ey indifferent. 
