^{page 375 note 2} In the systematic transcription so-called emphatic or velarised consonants are written , (respectively breathed and voiced denti-alveolar fricatives), , (breathed and voiced dentialveolar plosives) and (voiced alveolar flap). The symbols , j and y are used respectively for a voiced pharyngal fricative, a voiced palato-alveolar fricative and a palatal semi-vowel (see also note 2, p. 380). The acute accent is used where necessary to indicate the prominent syllable.

^{page 375 note 3} The use of ‘cannot’ is necessary since a few nominals, e.g. manwu ['mammu], do not perforce require d. nətta dmanwu and nətta manwu are both possible.

^{page 376 note 1} Other tenses could be used.

^{page 378 note 1} For discussion and application of the concept of ‘commutation’ (also ‘substitution’), see ‘The Phonology of the Nasalized Verbal Forms in Sundanese’ by Robins, R. H., BSOAS. xv, 138Google Scholar, and ‘Glottalization in Hausa’ by Carnochan, J., TPS., 1953.Google Scholar

^{page 379 note 1} The term ‘absence of …’ belongs to the related concepts of mutual exclusion and complementary distribution and is not to be equated with ‘zero’ as a term in a system. See also p. 381 and Addendum (1). Underlining and the symbol ζ are used for ‘absence of…’ and ‘zero’ respectively.

^{page 379 note 2} i.e. having the same number of terms and similar applicability to the language material.

^{page 380 note 1} Subscript numerals are used with C when reference is specifically to the consonants of N (excluding t initial).

^{page 380 note 2} Here and elsewhere below where special emphasis is placed on pronunciation, a transcription is given in square brackets using the symbols of the International Phonetic Association (revised to 1951).

^{page 382 note 1} See ‘Sounds and Prosodies’ by Firth, J. R., TPS., 1948.Google Scholar

^{page 382 note 2} []. Of. buday following—['bu:di]. Related forms justify the transcription, e.g plural iumyən [], ibudyən []. In general, [i] and [u] in Berber are frequently the phonetic forms of what are more systematically written əy, y, yə, əw, w, wə according to context.

^{page 383 note 1} Cf. elsewhere, for example, the singular and plural (suffixed ən) forms of

asəgrəs [] isəg(ə)rsən []

asəywən [] isəy(ə)wnən []

^{page 384 note 1} See, for example, Eléments de grammaire berbère (Kabylie-Irjen) by A. Basset and A. Picard.

^{page 385 note 1} Also swaənnátt. aənnátt s not, by definition, a noun.

^{page 385 note 2} For example, plural patterns (Arabic loans) in which w is infixed, e.g. əlqaləb-ləqwaləb [] but əlfuət-ləfwai [], əlbabur-bəwabər [], əlmusi-ləmwas [].

^{page 385 note 3} Cf. af [] + ayu [] = [], am [] + ayu + []; also ləmwas [] in preceding note.

^{page 385 note 4} The symbol ~ is used for ‘successions’ and the hyphen for correspondences.

^{page 386 note 1} Also the monosyllabic ti, and, with initial , ijjəlt and ijjal.