270 CHAUCER. 



Here the " but " of the third verse belongs at the head 

 of the first, and we get rid of the anomaly of " coveteth" 

 differently accented within two lines. Nearly all the 

 seemingly unmetrical verses may be righted in this way. 

 I find a good example of this in the last stanza of " Troi- 

 lus and Creseide." As it stands, we read, — 



" Thou one, two, and three, eterne on live 

 That raiguast aie in three, two and one." 



It is plain that we should read " one and two " in the 

 first verse, and " three and two " in the second. Re- 

 membering, then, that Chaucer was here translating 

 Dante, I turned (after making the correction) to the 

 original, and found as I expected 



" Quell' uno e due e tre che sempre vive 

 E regna sempre in tre e due ed uno." (Par. xiv. 28, 29.) 



In the stanza before this we 'have, — 



" To thee and to the philosophical/ strode, 

 To vouchsafe [vouchesafe] there need is, to correct " ; 



and further on, — 



" With all mine herte' of mercy ever I pray 

 And to the Lord aright thus I speake and say," 



where we must either strike out the second " I " or put 

 it after " speake." 



One often finds such changes made by ear justified by 

 the readings in other texts, and we cannot but hope that 

 the Chaucer Society will give us the means of at last 

 settling upon a version which shall make the poems of 

 one of the most fluent of metrists at least readable. Let 

 anyone compare the " Franklin's Tale " in the Aldine 

 edition* with the text given by Wright, and he will find 

 both sense and metre clear themselves up in a surpris- 

 ing way. A careful collation of texts, by the way, con- 



* One of the very worst, be it said in passing. 



