[43] Du Cange, preface, pp. 13, 29. Rusticum igitur sermonem non humiliorem paulo duntaxat, et qui sublimi opponitur, appellabant; sed eum etiam, qui magis reperet, barbarismis solæcismisque scateret, quam apposite Sidonius squamam sermonis Celtici, &c., vocat.--Rusticum, qui nullis vel grammaticæ vel orthographiæ legibus astringitur. This is nearly a definition of the early Romance language; it was Latin without grammar or orthography. The squama sermonis Celtici, mentioned by Sidonius, has led Gray, in his valuable remarks on rhyme, vol. ii. p. 53, as it has some others, into the erroneous notion that a real Celtic dialect, such as Cæsar found in Gaul, was still spoken. But this is incompatible with the known history of the French language; and Sidonius is one of those loose declamatory writers, whose words are never to be construed in their proper meaning: the common fault of Latin authors from the third century. Celticus sermo was the patois of Gaul, which, having once been Gallia Celtica, he still called such. That a few proper names, or similar words in French are Celtic, is well known. Quintilian has said, that a vicious orthography must bring on a vicious pronunciation. Quod male scribitur, male etiam dici necesse est. But the converse of this is still more true, and was in fact the great cause of giving the new Romance language its _visible_ form. 27. It may render this more clear, if we mention a few of the growing corruptions, which have in fact transformed the Latin into French and the sister tongues.--The prepositions were used with no regard to the proper inflexions of nouns and verbs. These were known so inaccurately, and so constantly put one for another, that it was necessary to have recourse to prepositions instead of them. Thus _de_ and _ad_ were made to express the genitive and dative cases, which is common in charters from the sixth to the tenth century. It is a real fault in the Latin language, that it wants both the definite and indefinite article; _ille_ and _unus_, especially the former, were called in to help this deficiency. In the forms of Marculfus, published towards the end of the seventh century, _ille_ continually occurs as an article; and it appears to have been sometimes used in the sixth. This of course, by an easy abbreviation, furnished the articles in French and Italian. The people came soon to establish more uniformity of case in the noun, either by rejecting inflexions, or by diminishing their number.--Raynouard gives a long list of old French nouns formed from the Latin accusative by suppressing _em_ or _am_.[44] The active auxiliary verb, than which nothing is more distinctive of the modern languages from the Latin, came in from the same cause, the disuse, through ignorance, of several inflexions of the tenses; to which we must add, that here also the Latin