The differences which exist in the spoken language of France in different districts have been caused by two sets of influences, the one arising from a condition already existing before the Latin language was spread over the Gallic territory by the conquering Romans, and the other due to events which occurred after that language had gained a firm footing there. From the variety which exists to-day in the language of any speech district, we are justified in assuming that a like variety existed in the Celtic language of France before the Romans arrived there. The Latin language, coming in contact with a language thus variegated, was necessarily influenced in one part, in a different way from the influence it felt in another part. In this way the Neo-Latin language was broken up into dialects.

‘La Naissance du Chevalier au Cygne,’ published by Henry Alfred Todd in vol. iv, Nos. 3 and 4 of the Publications of The Modern Language Association of America, a publication deservedly valued and supplemented as to the introduction, with many important additions by Gaston Paris in the Romania, vol. xix, pp. 314-335, is the earliest version of the legend of the “Children changed into Swans,” contained in the French Cycle of the Crusade. It is not without any interest that the three French versions of this legend have been published in an inverse order of their origin. The younger version, published by Reiffenberg, was followed by Hippeau's publication of the middle version, both connected with the other branches of the Cycle of the Crusade. That now the earliest French version has been published, is of great importance to the scientific understanding of this interesting legend.

The name ‘Gerund’ was borrowed from Latin grammar, and has for a long time been in use in English, though with varying application according as it covers more or less ground. It is found in grammars of the last century and of this, where it is often confused with the ‘verbal noun.’ This name is employed here because it is the most convenient that has been suggested; ‘infinitive in ing,' used by Whitney and others, while suggestive of its relationship and real function, is too unwieldly for practical grammar. A far more objectionable name, because the objection is based upon a radical difference in functions, is ‘participle,’ which is given to the verb-adjective also, thereby causing confusion. Etymologically, ‘gerund’ is just as misleading as the majority of its fellows. Participles, for instance, would by this test include infinitives, for infinitives partake of two natures; and infinitives would include participles, which are just as unlimited as infinitives. The origin of words, however, does not confine their use with unyielding bonds, and their use defines them.

William Heyse places as a motto on the title-page of his ‘Mecklenbörger Burhochtid’ the following lines:

Certain languages and dialects all over the world—Catalan, Italian, German, Algonkian—relieve themselves from the charge that each is “a bush where no birds sing”, by presenting a peculiar development of diminutives, which gives the speech a noticeable “Klink un Klank”.

It is with much hesitation that I venture to speak to you on a subject which many of you understand better than I do; the only excuse I offer is, the profound regret with which I notice the indifference concerning French literature.