PROGRESS OF ROMANCE PHILOLOGY 249 



greatest detail, the historical phonology of the French language. 

 The edition of the French versions of the Vie de Saint-Alexis (1872), 

 the essential parts of which are his personal work, shows how far 

 he was then in advance of his contemporaries in his knowledge of 

 Old French. But he was never satisfied with this work of his youth. 

 For many years he treated in great detail the subject of French 

 phonology at the College de France. The fragments of these lectures 

 which he published in Romania and elsewhere (on the " close " o in 

 French, on the development of Latin c, etc.), are sufficient evidence 

 as to the depth of his researches on this difficult subject. I know that 

 certain portions of the great work that he planned on Old French 

 grammar were ready to print at the time of his death (1903); per- 

 haps it will be possible some day to publish them. We recognize the 

 impress of the master's method in the works of several of his pupils. 

 I shall only cite, because they are among the earliest, the book 

 of Charles Joret on Latin c in the Romance languages (1874), and 

 that of Arsene Darmesteter, a scholar prematurely lost to science, 

 on the formation of compound words in the French language (1875). 

 These are works which completely replace the corresponding chapters 

 of Diez's grammar, but which nevertheless cannot be considered 

 definitive, so abundant is the material ready to the hand of him 

 who has eyes to see it. When M. Joret's book appeared, with more 

 than four hundred pages of close print devoted to a subject which 

 in Diez's work occupies a few pages, it might have been supposed 

 that the material was exhausted. Not so; more recent researches 

 have developed and completed in various directions the work of 

 M. Joret. 



In Italy, the establishment of Romance studies on a scientific 

 basis dates from the foundation of Prof. Ascoli's Archivio glottologico 

 italiano (1873). It was a rare piece of good fortune that these studies 

 were then undertaken by a scholar who was a veteran in linguistic 

 research, who was entirely at home in the various fields of Indo- 

 European philology, and who moreover was endowed with a breadth 

 of view and a power of expression which would have placed him in 

 the first rank in any other field of human knowledge. Prof. Ascoli's 

 Saggi ladini, which occupy the first volume of the Archivio glotto- 

 logico and overflow into later volumes, are a model description of 

 an idiom whose infinite varieties cover a considerable territory and 

 which has left traces in regions where to-day it is extinct. It may be 

 said that in this section of Romance philology, aside from a few 

 useful remarks by Diez, nothing had been done. The limits of the 

 language spoken to-day by the populations of the southern parts of 

 the Grisons, of the Tyrol, and of northeastern Italy, had not been 

 determined; still less was there any suspicion of the existence of a 

 former wider extension of these dialects, whose territory is now greatly 



