PROGRESS OF ROMANCE PHILOLOGY 243 



also of that branch of it which he had specially studied, Provencal 

 philology. One by one, however, professors of Germanic languages 

 Adalbert von Keller, W. L. Holland, Konrad Hofmann, Karl Bartsch 

 offered courses in Old French and Old Provencal. On the other 

 hand, the teaching of English philology was often coupled with that 

 of Romance philology, with the result that in 1859 a magazine 

 was founded by Ferdinand Wolf and Adolf Ebert " fur englische und 

 romanische Litteratur." 



It may be noted further that in Germany the study of litera- 

 ture was decidedly more popular than the study of language. The 

 example of Diez was followed rather slowly. As regards Old French 

 literature, for example, an influential initiative was that of L. 

 Uhland, who, in 1812, in an essay now famous, directed attention 

 to the French epic poems (chansons de geste). 1 Acquaintance with 

 the Old French epics and with the romances of the Round Table 

 was soon recognized as indispensable to any one who intended to 

 study thoroughly German medieval poetry. 



The second edition of Diez's grammar showed that great progress 

 had been made since the first, but it owed little to works published 

 after the date of this edition. The various changes and additions 

 made by the author were the fruits of his own researches: what 

 others had discovered was comparatively little. The third edition, 

 which appeared from 1869 to 1872, is less personal. In the period 

 between 1860 and 1870, centres for Romance studies had been formed 

 in Germany and elsewhere; works of value had appeared, and the 

 utmost the master could do, enfeebled as he was by age, was to 

 introduce into this third edition a portion of the results obtained 

 by his successors, all of whom might have called themselves his 

 pupils, although few of them had been actually present at his lec- 

 tures. 



If, in the year 1904, nearly half a century after the second edition 

 and thirty-five years after the third, we examine Diez's grammar 

 from the heights now reached in our knowledge of the Romance 

 languages, there are two facts which will strike every impartial 

 observer. The first is that the rules established by Diez are still for 

 the most part valid; his doctrine remains practically whole and 

 sound. The second is that there are serious deficiencies in the work. 

 Certain very important questions of a general character are not 

 dealt with at all; various Romance territories are incompletely 

 explored; the geographical extension of linguistic phenomena is 

 not indicated with precision; the notation of sounds is often too 

 vague, and the history of their changes is at times neglected. 



To pass these deficiencies in review is to realize the fact that the 

 greater part of them could hardly have been avoided half a century 

 1 In Fouqu6's Musen, first year. 



