When we pass from comparisons of general scheme and spirit to those of metrical form, the matter presents greater puzzles. As observed above, the earliest French _chansons_ known to us are written in a strict syllabic metre, with a regular caesura, and arranged in distinct though not uniformly long _laisses_, each tipped with an identical assonance. Further, it so happens that this very assonance is one of the best known characteristics of Spanish poetry, which is the only body of verse except old French to show it in any great volume or variety. The Spanish ballads are uniformly written in trochaic octosyllables (capable of reduction or extension to six, seven, or nine), regularly assonanced in the second and fourth line, but not necessarily showing either rhyme or assonance in the first and third. This measure became so popular that the great dramatists adopted it, and as it thus figures in the two most excellent productions of the literature, ballad and drama, it has become practically identified in the general mind with Spanish poetry, and not so very long ago might have been described by persons, not exactly ignorant, as peculiar to it. [Sidenote: _Ballad-metre theory._] But when we turn to the _Poema del Cid_ we find nothing like this. It is true that its latest and most learned student, Professor Cornu of Prague,[195] has, I believe, persuaded himself that he has discovered the basis of its metre to be the ballad octosyllables, full or catalectic, arranged as hemistichs of a longer line, and that he has been able to point out some hundreds of tolerably perfect verses of the kind. But this hypothesis necessitates our granting that it was possible for the copyists, or the line of copyists, of the unique MS. in the vast majority of cases to mistake a measure so simple, so universally natural, and, as history shows, so peculiarly grateful to the Spanish ear, and to change it into something quite different. [Footnote 195: I have not seen Professor Cornu's paper itself, but only a notice of it by M. G. Paris in _Romania_, xxii. 153, and some additional annotations by the Professor himself at p. 531 of the same volume.] [Sidenote: _Irregularity of line._]