^{page 49 note 1} m and s will be used as abbreviations throughout.

^{page 49 note 2} ā as the prolongation of ặ; ī, e and ai, as the prolongations of ĭ; ū, o, and au as those of ŭ.

^{page 49 note 3} Except before a pronounced cæsura.

^{page 49 note 4} The formulation is faulty, but the elaboration which follows makes the meaning clear.

^{page 50 note 1} A knowledge of the Devanagri script is very helpful to the student in mastering Urdu scansion. In this script, of course, the aspirated consonants are represented, as scientifically they should be, by a single letter.

^{page 50 note 2} We may anticipate a little at this point to remark that it is convenient to treat m as short () and ms as long (—).

^{page 52 note 1} Grahame Bailey's lists (cf. below) cover 176 metres. He is certainly correct in his assertion that a proper classification would reduce very considerably the number of metres which can reasonably be regarded as distinct.

^{page 52 note 2} “A Guide to the Metres of Urdu Verse,” S.O.A.S. Bulletin, Vol. 9, Part 4.

^{page 53 note 1} Pybus, too, had recognized this but, unlike Grahame Bailey, he made no attempt to analyse the metres in terms of quantity.

^{page 53 note 2} Stated thus, “consonant” must be interpreted in the sense appropriate to the Urdu script. Stated without reference to this script the rule would be that “short” = consonant followed by a short vowel, or initial short vowel of a word or syllable—i.e. cv, or (initially) v. Long = cv;c, cv, or (initially) v or vc.

^{page 54 note 1} The inadequacy of his prefatory statement is puzzling, for his own analysis of the metres in quantitative terms must itself have necessitated the discovery of most of the points made above.

^{page 55 note 1} The word “may” is important. The poet is free to choose whether he will scan the word long or short.