126 



F. B. Liiquiens, 



luckily, however, the two possible stemmata are radicall}' different ; 

 it is imperative that one of them be proved the better. Our thesis, 

 if conceded, solves the question. 



Before proceeding to the discussion of this point, however, let us 

 recall the chief features of the manuscript dilemma. Here are the 

 two possible stemmata (for the meanings of x and x', cf. supra, p. 112 ; 

 x" and x'" = lost intermediaries ; V 4 = the manuscript of the Li- 

 brary of San Marco in Venice) ^ : 



The Oxford Stemma 

 x 



The Redactions Stemma 

 x 



O V4 



all other 

 extant redactions 



all other 

 extant redactions 

 (inclusive of V4) 



There may have been intermediate manuscripts, now lost, between 

 X and x', between x' and O, etc. ; but the question of the existence 

 and number of such intermediaries is not so important as that of 

 the amovmt of difference between x and x', between x' and O, etc. 

 — these various amounts of difference are represented by the various 



introductio2i of his second edition of the Roland, Gottingen, 1878 ; (b) Zeit- 

 schrift f. roman. PhiloL, iii, 439—452 : (2) of the ' redactions stemma ' : 

 Ram.beau, tjber die ah echt nach-weishare^i Assonanzeii des Oxf order Textes der 

 Chanson de Roland, Halle, 1878, pp. 10—17 ; Toerster, Zeitschrift f. roman. 

 Philol., ii, 162—180 ; Perschmann, Die Stelhing von O iti der Vherlieferung 

 des altfranzosischen Rolandsliedes, in Stengel's Attsgaben und Abhandbingeii aus 

 dem Gebiete der roman. Philol., iii, 1—48 ; and especially Stengel, (a) Ver- 

 handlungen der fiinfiindvierzigsteii Versanimlung Deutscher Philologen rtnd Schtil- 

 mdnner in Bremen vom 26. his 2g. September j88p, pp. 132—133 ; (b) in the 

 introduction of his edition of the Roland, Leipzig, 1900. 



^ For complete stemmata, cf. Zeitschrift f. roman. Philol., ii, 163—164; 

 and Fassbender, 1. c, p. v. 



