THE MASARWAS AND THEIR LANGUAGE. 223 



The accent is always on the final syllable of a word, and is 

 very marked, the stress rising towards the end of the word, as 

 guirihc, to perish, amatho, a small pot, gahe, a giraft'e. The 

 union of accent and stress has rather a pleasing efifect on the ear. 



The great majority of words in the language are dissyl- 

 labic, a number trisyllabic, and a very few polysyllabic, most of 

 them evidently derived from monosyllabic roots. Most of the 

 words end in a vowel, the remainder in a nasal 77; or n, but never 

 in any other consonant. In this respect Sesarwa resembles 

 Sechuana and the other Piantu tongues round about, but other- 

 wise there is no resemblance between them, except loan words, 

 of which a considerable number exists in the language, princi- 

 pally from Sechuana. 



There are no prefixes in the language. Most of the relations 

 of time, place and number are expressed by the addition of suf- 

 fixes, or by the use of separate words. Some of these, especially 

 the pronouns, are prefixed, not postfixed. but are not prefixes in 

 the proper sense of that term. 



There are but two numbers in Sesarwa, singular and plural. 

 The plural is expressed by the addition of the suffix re or ra, as 

 kolia, a zebra, koharc, zebras, klunii, a lion, kJianirc, lions. Bor- 

 rowed words usually take the suffix ra, as koko, a fowl, kokora, 

 scperc, a horse, scpcrcra, horses. If, .however, the word 



ends in i, as gaicti, a servant, the form assumed is gaictiare; 

 shotsi, a dreamer, shotsiarc. It would appear that here we have 

 the original form of the plural of which re is a worn-down form, 

 while ra is a later development, especially as it occurs most 

 largely in loan words, and in words expressive of things not 

 familiar to the old Bushman life. Reduplication to form the 

 plural such as is given by Bleek, Bertin and Planert in their 

 grammars of the various dialects known to them do not occur in 

 Sesarwa. The plural is always formed by the above suffixes, and 

 never, to my knowledge, in any other way. 



Abstract words are very rare in the language. Where such 

 have to be employed, they are formed from the infinitive mood 

 by the addition of the suffix 0, as \\khaba, to form a habit, 

 \\khabao, a habit. But it is doubtful whether such forms are 

 really abstract nouns or not, as 777,00, sight, and many others 

 that might be named, are the same as infinitive mood. 



Gender does not exist in the language, at least, in the way 

 it can be expressed in Xamaqua. Thus Bleek's contention that 

 Bushman would turn out to be not a sex-denoting language is 

 correct, so far as Sesarwa goes. The gender can only be in- 

 ferred from the context. The agent is expressed by the suffix 

 thi, as \\gan, to build. \\gantJii, a builder. There is no article 

 in the language, nor anything corresponding to it. Thus aba 

 may mean a dog, or the dog, du an eland, or the eland. 



There are no cases in Sesarwa. The genitive, dative or 



accusative can only be expressed by the aid of prepositions, p<\\j^*|^^ 

 inferred from the structure of the sentence, as : Hie r^nu /<oSB~/>j 





