266 BANTU PLACE-NAMES IN AFRICA. 



the Eastern Province, and hence phonetised with more precision, 

 by the missionary grammarians, than it was the fate of the names 

 to be, which had, in earlier times, been learnt from the Hotten- 

 tots by the Colonists. Thus the place which appears on our maps 

 as Komgha is spelt with more precision by Kaffir scholars as 

 Qumra. and our Kln-Klu is really Xu-Xuwe ! 



It will be convenient to deal, at this point, with the western 

 limit of the southern Bantu names, which we shortly sketched 

 at the beginning. We shall thus complete the southern sector 

 in which Bantu and Hottentot names overlap, a sector stretching 

 roughly from North Zululand to the right bank of the Keiskama, 

 and from North Basutoland to the southern sea. Our western- 

 most southern Bantu names seem to reach to the Umtata. and 

 the Umgeto stream, just east of the Great Salt River, and 

 Bokana's River, west of the Bushman's River. Further north 

 we find Umtlouli's and Kwenkezile's. Matomela's and Jokweni's 

 kraals in the Peddie district (the last name, however, is, of 

 course, not Bantu, but from the English "yoke"), (joing north, 

 we meet with Dcbe, near the Keiskama, and Gonzana, west of 

 the Tyumie. Lastly, east of the Zwartkei source, we get Ntaba- 

 kaMoses, obviously a recent name, since the Kaffirs have lieen 

 penetrating the interior of the Colony, and Musa Cmtwakazi, 

 with Umaleki and Lovani east of Aliddelburg. ( loing north- 

 ward again, we find a huge gap in our native names, owing to 

 two invasi(Mis, the one from the east, due to results of Tshaka's 

 rule, and one from the south, less violent but more effectual 

 and permanent, of the Dutch X'oortrekkers of the '30's. 



The Orange Free State contains almost no native names on 

 the map, and was, no doul)t, but sparsely populated at the time 

 of the (ireat Trek, having been largely depopulated by the earlier 

 raids of Moselekatse, stretching from Thaba Bosigo, in Basuto- 

 land. to Marico, near Zeerust, his earlier settlement, and far 

 beyond. Most of the town sites, however, have native names 

 {e.g.. IHoemfontein is McnujaiDuj, where /arr/r -a'ildcats nrt' ) . and 

 also the rivers, e.g., Linyana is the \'et at Menschvretersberg.* 



Before we find Bantu names again in an\- strength on (.ur 

 majis, we must ])ass Uj) once more cast of the Drakensberg, till 

 we reach the district on the border of the I'.ast Transvaal and 

 Swaziland, where we get a western limit of Bantu names on mv 

 large map (in \\'akkerstrt)oni and Piet Retief districts); thence 

 ])\- .Ma\iriestad. northward by the Cjodwan River and Marola to 

 tlie Zout])ansl)erg. .\orth-west of Marabastad, the ri\ers still 

 maintain tlieir nati\e nanu'^ 1 ui tlie l\ailwa\- ma]), but Lclioe is 

 tiie only station 01- siding thai j)re>er\es one on tlie Pietersburg 



* It has l)crn tlmiiKln ll'^it H" I'-nntii had ()i.i.u|)ii(l llic prcsciil l-"rce 

 State lieforc tlie Dutch, hut this certainly does iioi :i])i)ly ti) the eastern 

 part hetween the L'l^per Vaal and the Caledon. A casual perusal of Mr. 

 EllenherKcr's excellent "lli';tor> of the i'.asuto" will show thai these 

 trihcs often settled there on tlie way from the Xorth to the present " ( on- 

 (|uerr(l 'ierritory." 



