criterion I have decided to give the benefit of the doubt to them all. On the contrary, verses which AC or AD omit, and which occur in B or BL alone, are not likely to be genuine; the few that were inserted in my text of Book II before I had access to D remain there on probation, as their removal would have caused much inconvenience [1]. Generally I have followed the text of one MS. (C for Book I and D for Book II). This was the only practicable course, for most of the variants are of such a kind that any selection must have appeared capricious. The readings of C and D have been corrected where necessary, but have seldom admitted a variant reading on the ground that it would express the meaning more clearly or elegantly, since that is just the ground on which many variations have been introduced by copyists. Their attempts to improve the Mathnawí will not be imitated by those who recognise that Jaláluʾddín, great poet as he is, cares more for the substance than for the form, and that the smooth perfection of highly finished art is incongruous with the character of the poem, the manner of its composition, and the power, fullness, and rapidity of thought which it so frequently displays. Almost every correction of the C and D readings has been derived from A or B, the two MSS. which come next in ago. I have occasionally ventured to make emendations of my own, as at I ٢٠٤١, where I read (MSS. and edd. ve 2), and II ٢٣٧٩, where MSS. and edd. have instead of .
In its orthography my text differs little from the MSS., but I distinguish and a 4 and 5, and ignore the post-vocalic 3. The archaic spellings are nearly always retained.
Orthographical differences.I follow the MSS. when instead of 5 and they write Sand scanned as long syllables before consonants or as short syllables before alif; when they write and j, scanned as short syllables
- ↑ The verse in question are ٨٣٩, ١٢٥٩, ١٩٦٧, ٢٢٢٤, ٢٥٠٤.
