112 F. B. Luqntens, 



lence. For this thesis may often be used as a decisive weight, 

 where former investigators have laid an equal amount of evidence 

 in either scale. To its proof, then, will be devoted the first part of 

 this article. 



Let us call to mind the interrelations of the various Roland re- 

 dactions. The original poem (whether it existed in manuscript form, 

 or only orally, we need not attempt to decide) may be called x. 

 From X derived, perhaps through lost intermediaries, a lost manu- 

 script, which may be called x'. x' was already augmented b}- 

 interpolations, especially by that of the Baligant Episode. ^ From 

 x' derive, through lost intermediaries (except, possibly, in the case 

 of O), all the extant redactions. 



That X was a poem of marked and consistent technical excellence 

 is practically 2 proved by the following two facts: (1) one of the 

 extant manuscripts, O, is of marked and almost consistent technical 

 excellence ; (2) the few technical faults of O may be plausibly 

 attributed to copyists.^ 



These facts, however, need detailed exemplification, for very few 

 nvestigators have realized them. If Petit de JuUeville has con- 

 tended,^ " On pourrait . . . montrer . . . que I'oeuvre a une contexture 

 bien plus serree qu'on ne I'a dit quelquefois " ; if Ten Brink has 

 written,^ " Wie gewaltig ist die Konzeption des Ganzen, wie ein- 

 heitlich und geschlossen, in alien Teilen von der herrschenden Idee 

 durchleuchtet ist die Komposition ! " ; if Professor W. M. Hart, in 

 his recent Ballad and Epic. ^ frequently ascribes artistry to the poem; 

 the great majorit}^ of writers have thought otherwise. For instance, 

 Becker, in his Grundriss der altfranzosischen Literatur (1907),'^ denies 



^ Cf. Miiller, in the introduction to his edition of the Rol^ind (1878), 

 p. V ; and Stengel, in the introduction to liis edition, jd. ix. 



^ I say ' practically,' because it is possible, though very improbable, 

 that the technical excellence of O was due to some lost intermediarj^ 

 between x' and it. Cf. infra^ p. 127, Note 3. 



' It is hardly necessary to say that we are speaking of O's technical 

 excellence as a poem, not as a manuscript ; and that therefore we do 

 not consider such eiTors as those of vv. 197, 562, 604, etc., as teclmical 

 faults. Such errors are merely typographical errors — if we may be allowed 

 the anachronism. 



* La Chanson de Rolm^d (Paris, 1894), j). 57. Cf. also Geddes, La Chanson 

 de Roland (New York, 1906), p. 175. 



^ Geschichte der englischen Litteratur (Sti'assburg, 1899), p. 145. 



' Studies and Notes in Philology and Literature, Volume XI (Boston, 1907). 



' pp. 40-41. 



