134 F. B. Luquicus, 



of O's strophes, so in accord with the general technic which we 

 ascribe to x', a strong argument against any change ? 



The second instance of Miiller's failure to apply correctly his 

 formula, is his arrangement of the strophes which constitute The 

 Second Encounter.i In O this chapter is exceedingly clear. From 

 the ver}' first, when, in the poet's words, 



Li arcevesques cumencet la bataille, 



through the series of combats in which now a Paynim, now a Frank, 

 is victorious, until the moment when the Paynims cry w-ildly to 

 Marsila for aid, we are never in doubt as to what encounter is 

 being described ; then, in the last strophe of the chapter, the poet 

 speaks very briefly of a third encounter, merel}' mentions a fourth 

 and a fifth, after which only sixt}^ Franks survive — 



Ne mes seisante que Deus ad esparmiez . . , 

 Einz que il moergent, se venderunt mult chier ! 



There is a very great difterence between this and the corresponding 

 chapter in Miiller's text : the difference between clearness and 

 vagueness, between order and confusion. From beginning to end 

 we are never sure — Miiller himself, b}^ his own confession, was not 

 always sure ^ — just what encounter is being described. The cause 

 of all the trouble is this: Miiller has shifted two strophes 3 far from 

 their position in O. He gives two reasons.* The first is that all 

 the other redactions show his order. Such a reason, however — as 

 explained above — is not in itself cogent.'' The second consists of 

 the following argument: vv. 1671—1676 read as follows: 



Li quens Rollanz apelet Olivier : 



Sire cumpaign, sel volez otrier, 



Li arcevesques est mult bons chevaliers, 



N'en ad meillor en tere ne suz ciel, 



Bien set ferir et de lance et d'espiet. 



Respunt li quens : Kar li aluns aidier . . . 



therefore the strophe immediatel}' preceding these lines must recount 

 a combat in which the Archbishop figures ; In O this condition is 



' 1438-1690. 



■>■ Cf. liis edition (1878), p. 168. 



» 1628-1670. 



* Cf. his edition (1878), pp. 144-145. 



■'• It is probable, in my opinion, that the scribe of x" cbaiig-ed tlie order 

 of his original because he, like Miiller after him, interpreted vv. 1671—1676 

 too literally. 



k 



