ON ETHNOGRAPHICAL PHILOLOGY. 
219 
2. TheEuphonic Concord. —In respect to the value of this character a single 
remark is sufficient. No scholar who admits the Celtic languages, with 
peculiar system of initial changes, into the list of Indo-European lan¬ 
guages, can, on the mere strength of the euphonic concord, isolate the 
Ca^ tongues. 
What has just been written as to the value of these two important cha¬ 
racters of the Caffre languages, the Euphonic Concord and the System of 
Prefixes, has dealt with the question, as a general question of philological 
criticism. It has assumed them to be found in the C'affre languages exclu- 
atv/y, viz. in the CafiVe throughout, and in nothing else anywhere. Yet 
this is, probably, the fact in neither case. 
Tht Eiqthonic Concord in odter Langvoges besides the Cajfre. —It is stated 
that the euphonic concord exists in the A fact mentioned to me 
byDr. Tutsehek induces me to expect it in the Tumali. In the WolofF its 
operation is undoubted, and has long been recognized. The definite article 
in Woloff is the numeral one. According to the initial of the noun with 
which it agrees it changes its own. 
66 nne 6aye 
a father. 
rfhienne ^uaba 
a lion. 
yui^one Aabousc 
a pistol. 
fenne radaa 
ajar. 
wieune mpithie 
a bird. 
senno safara 
a fire. 
venoe/asse 
a horse. 
On the other band, the definite article is postpositive; it, nevertheless, 
Inhes the euphonic letter of the substantive. 
6 aye-fia 
the father. 
<iftaba-<fhia 
the lion. 
Aabousse-yun 
the pistol. 
wdua-6a 
the jar. 
wnithie-nia 
the bird. 
safara-sa 
dtcfire. 
^ssc-t»a 
the horse. 
The System, of Pn’fixen in other Languages besides the Cajfre. —Three 
wets, put together, suagGst resparchea npon this point. 
ff. The monosyllabic root ^+a vowel=ir/w; in several of the languages 
onWtern Africa — Kru, in ; Pessa, Ishu ; Ashaiitce, iduia —Karaba, itu ; &c. 
p Caffre for tree is the root /+avowel,-f-tlie prefix m, mi, or imi — 
Uffre, wtrft’; Suaheli, wd; Wakamba, Boyce writes, “the twelfth 
wdension contains jdural uouns, heginning with the prefix tni*;” and gives 
instance the word in question, iniUi=^t)rts. 
IheGonga for tree is mitto; the Kaffa, wiitsa; the Wolaitsa, rntssa; 
«ic Woratta, w»‘Ui. In order to get at the root, it is l>cUcvcd that the ne- 
^ity for this doconipositioii is gcjieral in the African languages, and that 
the application of it will give a partially moiiosyUabic character to more 
than one language of Africa, in tJiLs sense some remarks of d’Avezac upon 
the Yebu language are borin* out.~.Sf’e .Mrmoires de la SociiU Eihnologiqvc, 
Tol. ii. frart 2. p. 52 ; ami Clussieul Museum, No. xii. pp. 197 and 201. 
3. The Position of Injkxiond Elnnmts.—In tJic Galla language the in¬ 
flexional elements generaUr come before tbo root; in other words they are 
P«fixes. 
