892 LANGUAGES 



Bantu, and died at the end of the 'sixties of the last century leaving a 

 great comparative grammar of these languages only one-third written and 

 published)— Dr. Bleek, studying the full forms of these Bantu prefixes, 

 made a remarkable suggestion deduced from very little evidence. He 

 pointed out the curious want of correspondence in some cases between the 

 prefix and its particle. He noticed that the "31a-" prefix never, except 

 in certain degraded West African languages, has a corresponding "-ma-" as 

 particle, but on the contrary is followed in the sentence by " Ga-" ("-ga-"), 

 " Ya-," or " A-." He also noticed that the " Mu-" (third) prefix generally 

 has as a corresponding particle "Gu-" or some degraded form of " Gu-." 

 Finally he summed up his researches (he had very limited evidence to 

 go on) by declaring that the original form of " Mu-" was " Ngu-," and of 

 'Ma-," "Nga-"; and that it would be found that the preliminary vowels 

 of the full form of the prefix were thus but the vestiges of a former 

 reduplication of the syllable. Thus (according to Bleek), the old form 

 of " Umu " would be " Ngungu-," of " Ma-" would be " Nganga-" etc. 

 Now when we go over again all the material which Dr. Bleek had before 

 him, we can only admit that this was a very remarkable and clear-sighted 

 theory ; and it is a theory which is increasingly supported by subsequent 

 investigations. At the same time there still remain several points of 

 difference to explain away. Although it was unquestionable that in most 

 of the Bantu languages the (third) "Mu-" prefix had " (xu " as a corre- 

 sponding nominative prefix, while " Ma-" was likewise followed by " Ga," 

 still the fact remained that "-mu-" often answered to " Mu-" (very seldom 

 to the third prefix, but almost always to the first) as an objective or 

 accusative particle, and nearly always did so in an adjectival capacity,* 

 while " ma " was sometimes the accusative and always the adjectival 

 particle of " Ma-," even though the nominative form of the particle might 

 be "ga" or " ya." Moreover both "Mu" and "Ma" seem to have been 

 amongst the oldest prefixes, and even to have existed, and to exist still, 

 in West Central African languages to the north of the Bantu line, which 

 are the only African tongues offering any resemblance whatever to Bantu. 

 " Mu," indeed, in early African forms of speech seems to have been the 

 sound meaning "one" and also "a person," "a man," while "Ma" not only 

 indicated water or liquid, but was frequently used in non-Bantu (as well 

 as in Bantu) languages as a kind of collective plural prefix indicating "a 

 company," "a collection," 'a flood of things." 



* Thus, in Luganda, omu-ti is " a tree " (third prefix). 



Adjectival prefix omu- (omw-lungi = handsome [tree]). 

 Numeral > 



Nominative ! prefix "gu-" (otnuti f/wmo gugwa ; o^teme). 

 Accusative ; (Tree one falls ; cut it). 



