and neglected not only in Arabic phrases (e. g. وطيبات للطيبين, scaned --|---, {{{1}}} (
Arabic characters), scanned ---|--, an example which illustrates the author's bold use of the metrical tashdíd), but also perhaps in Persian words now and then [1]. For our immediate purpose, however, his practice in regard to the rhyme is especially instructive. Here, again, he takes great liberties. It has been asserted [2] that such rhymes as ذورjy (mír) and گور (gór) are inadmissible in the Mathnawí, but they are quite common (e. g. I and {{{1}}}; I and ; (Arabic characters) etc., etc.). On the other hand, so far as I have noticed, yá-yi maʿrúf (i) is hardly ever rhymed with yá-yi majhúl (é); [3] in AC the latter often rhymes with (á).
There is a second class of imperfect rhymes, in which the short vowels are different, e. g I وشور and گور I۳۰۹۰ ;گور and شور I۳۰۸۹ ;صور and شور and ثور ۴۹۵۷ Ir۹۴۰ and ;النشور and and منکری رتی and ژنی ,آمدن and شدن . , F , J or in which a short vowel corresponds to a nímfatḥa, e. g. .ملك را and مر مرا
Early alterations in the MathnawíSome examples occur in all the MSS., but the significant fact is that their number varies according to the age of each MS. C has more of them than A, A more of them than B. The youngest of the three MSS. is the most correct from a prosodist's point of view. In other words, we can here trace the process and method by which the copyists have altered the text of the Mathnawí, since in nearly all these cases the motive which induced them to alter it is unmistakable. The number of verses affected (in Book I alone) is about sixty. Without giving a full analysis, I will set forth the evidence in some detail, as was promised above.
- ↑ E. g., I think the readings of AC Bul. at I ١٧٠ and of AD Bul, at II are probably genuine, though I have not ventured to adopt them.
- ↑ By Prof. C. E. Wilson in his Commentary on Book II, 1. 949.
- ↑ The only instance I can recall just now is I ١٦٠٥, where (jarí) rhymes with (miré).
