chap, m STRUCTURE OF MON-KHMER GROUP 447
though in some cases it seems to be possible to discern in them an original monosyllabic root, yet this does not stand out as clearly as in the Mon-Khmer languages; it is no longer (as a rule) capable independent existence, and the normal type of the language is dissyllabic. If it was originally monosyllabic, it has long since passed out of that stage, and its dissyllabic words are now treated as independent roots for all purposes of composition by means Af the addition of prefixes, infixes, and suffixes.1
The point, however, which mainly concerns us, is that a great part of the constituents of the Sakai and Semang dialects agree closely with the Mon-Khmer languages both in the monosyllabic character of their root-words and in the method of composition by Prefixes and infixes.2 When we exclude the Malay and other Malayan elements, we find that the words which are common to Sakai and Semang are in the Main monosyllabic, or capable of being reduced to Monosyllabic roots. Thus, in Sakai, from the root cflAa (Semang chi'), ato eat,a we get the verbal formations am-cha', en-chd, ka-cha' (and in Semang nte-chi'\ and the substantival dncha (Semang inchi) aiid chana', a food a 8; from yut, a to return,a i.e. to go hack, the transitives tyut and tengyut, a to return,a x-ea to give back, and so on, quite in accordance with Mon-Khmer methods.
1 As to the probability of an earlier System of monosyllabic roots in the j^layan languages, see especially **jnappel, a Over de Wortelwoorden in Maleische Taal,a and Vreede, aOver ^ Wortelwoorden in de Javaansche ^1)a in Actes du Sixihne Congris nternational des Orientalistes tcnn en **83 e Leide, 1885, and Brandstetter,
Tagalen und Madagassen, pp. 48*57, 1902.
2 See especially Schmidt, op. cit where this fact was first pointed out. Skeat (in the Selangor Journal) had already drawn attention to the prefixes in Besisi.
3 Probably an infix formation ( = ck-an~aa).3 Probably an infix formation ( = ck-an~aa ).