720 ABORIGINAL PLACE NAMES. 



of the proper name, then the letter ,to be capitahsed is that next 

 following after the crasis. Thus, NgoTukela (along the Tukela), 

 neNkandhla (and the Nkandhla)':, . . Further on, in his notes 

 upon these above-mentioned rules, he proceeds : — ' After much 

 study ... I see no reason to change or modify the foregoing 

 rules. . . . They are in the best possible accord with good 

 taste, do not mar or break up the word, or disturb its integrity. 

 . . . . Much less do they involve any arbitrary, difficult, or 

 needlessly minute division in a name.' 



" By the above eulogy of his system, Gjrout seems to 

 have unwittingly stated to us some of its most conspicuous 

 defects. For just what it does involve is an arbitrary, compli- 

 cated, and needless ' chopping about of the names ' ; just what it 

 does do is ' to mar and break up the words and destroy their 

 integrity,' at least in so far as it makes the name of a particular 

 river appear at one time as though it were Utiikela, at another as 

 though it were Otukela, and at another as though it were Tnkela, 

 and so on with all names written according to this changeable 

 method — the capital letter, which most people would take to 

 indicate the commencement of the actual name, would be inces- 

 santly altering. ... at one time Inkandhla, then Enkandhla, 

 and finally neNkandhla!" (Introduction, pp. 92-93.) 



The main point, then, is to have a system, and to adhere to 

 ir as far as possible. But while Mr. Godfrey is consistent 

 throughout, and rightly so from the strictly philological point of 

 view, yet place-names which are commonly current have a way 

 of getting down to a working level, and so for general ])urposes 

 adherence to prefixes is not practicable, the whole tendency being 

 to eliminate them as far as possible. With all this in mind, we 

 have ventured to fix the place-names as we found them, the 

 elifort being to get at the most generally accepted form, consistent, 

 of course, with accuracy, and, in any case, the acceptance of 

 the single neutral form is sufficient in itself and avoids the over- 

 burdening of the lists. 



The name Indwe, for instance, if we adhered to the strict 

 rule, should be written in-Dwe, and be classified under the 

 letter D — an arrangement that would effectively disguise an old 

 friend. At the risk of some inconsistency, which we fear is 

 unavoidable in the circumstances, we have adhered to the better 

 known forms in all similar words. 



II. — Aboriginal Place-names Lists. 



I. — Distribution. 



A further point that has no slight value and interest is that 

 of the Distribution of Place-names, especially as between the 

 Hottentot-Bushman names and the Bantu names.* So long ago 

 as 1833 Kay committed himself to the statement that the Kei 

 was the dividing line, a statement that has come to be accepted 

 in a general way. He tells us that: — • , 



