122 F. B. Luqiiicns, 



Before leaving- the strophes, we must discuss two special uses to 

 which O puts them. The first is a device which might be called 

 the ' short -strophe group.' The average number of lines in O's 

 strophes is fourteen ; therefore ten lines or fewer would surely im- 

 press the hearer as a ' short strophe.' There are sixty-six such 

 strophes in the manviscript. Of these, forty-two are scattered, but 

 the remaining twenty-four constitute four almost equal groups. ^ 

 When we examine the content of these groups, we must conclude 

 that in each instance the choice of short strophes, rather than long, 

 was intentional. In one instance, the last part of the first encounter, 

 the device obviates monotony ; in the other three, the quarrel of 

 Roland and Ganelon, the swearing of the treachery, the duel of 

 Pinabel and Thierri, it helps to obtain the rapidity requisite to 

 dramatic effect. 



The other special use to which O puts its strophes is the device 

 known as ' laisses similaires.' ' Laisses similaires ' ma}' be defined 

 as two or more successive strophes presenting the same matter in 

 similar manner. It is obvious that here technical perfection will 

 consist partly in a proportioned combination of similarity and dissim- 

 ilarity — there must be as much of the latter as is possible without 

 lessening the prominence of the former ; and partly in the attainment 

 of what may be called an ' incremental ' 2 eifect — each successive 

 strophe must either advance the action, or augment the intensity, 

 of the preceding strophe. As an illustration of these points let us 

 analyse the three ' laisses similaires ' in which Oliver begs Roland 

 to summon aid. 3 In each strophe Oliver begs his friend to wind 

 his horn, so that Charles may turn back ; Roland refuses, giving a 

 reason for his refusal ; adding that he prefers to wield his sword, 

 and ending with the prophecy that all the Paynims shall die. Note 

 that in this summary there was one indefinite phrase— ' giving a 

 reason for his refusal.' This reason, the most important part of each 



1 317-341 ; 596-660 ; 1261-1319 ; 3850-3933. lu three of these groui^s 

 are interspersed some sti'oplies of more than ten lines ; the latter are not 

 of sufficient length, however, to prevent the groups from being felt as such. 



- Although I am not discussing the origin of the device of ' laisses 

 similaires,' I choose the word ' incremental,' because it points to the 

 possibility that their repetition is somehow allied to the ' incremental 

 repetition' of popular ballads. As to the origin of the device, cf. Die- 

 trich, Romanische Forschjingcn^ i, 1—48 ; Nordfelt, Les Coiiplcts Siimlaires dans 

 la vieillc Epopee Fran^aisc. Stockholm, 1893. 



3 1049-1081. 



