As counterplea, the lexicographers may not unnaturally allege that they are pretty much at the mercy of the special students of the French in its old form, and must accept such material as is at hand. And certainly the glossaries extant in English in reference to O. F. are miserably deficient. I am not intending to deny that often enough Frenchmen themselves have failed in their explanation of the old language, but English editions of Old French are generally very far behind the wants of the time. The fact is, it seems hardly to be realized that Old French needs a prolonged and special study before it can be understood properly. It may be stated, in the most unqualified terms possible, that Old French, in its various dialects, is not to be read either by the light of nature or by the simple knowledge of Modern French (no matter how perfect the knowledge of the latter be, and it usually is very imperfect); but that a complete course of special study will have to be undergone before the requisite mastery is attained, as separately as if the language had been Kamtschatkan or Chinese. And this, in case of an intend

ing Editor of O. F. texts, on peril of no less penalty, in the absence of such special study, than the liability to commit errors of the kind which it is the object of this paper to combat and expose.

But if the glossaries on the subject are so imperfect, the published texts, from which glossarial material might be gathered, are mostly untrustworthy. In one point of view, it is intelligible that it should be so. Many texts have been hastily edited, from a desire to give the historical student access to documents, which would otherwise be unavailable for use. But it may be conjectured what fate was in store for Old French MSS., when even the MSS. of our own mother tongue have been, until very lately, so dreadfully maltreated. The complaints of Germans against certain editors of Anglo-Saxon works are well-known (Grimm, Deutsche Gram., p. 186, edit. 1870). In fact, it is only

since 1870' that any grammar has been published in English which showed that the author had a just conception of the structure of the Anglo-Saxon language, and its relation to the remaining Teutonic family. (Of the grammars of the more important Teutonic languages, linguistically speaking, such as Gothic, Icelandic,' &c., the greater portion extant are very unsatisfactory. For instance, Mr. Skeat, in his convenient little Mæso-Gothic glossary, has appended a sketch of Gothic grammar, which is a poor production, even as a sketch. The competent reader may judge what a "valde deflendus hiatus" there is somewhere, from the following remarks of Mr. Skeat, p. 288, viz. :— "To au we may give the sound of aw in law [!]; to ai that of long in 'pride,' but somewhat broader." Mr. Skeat has evidently therefore (or had) not the remotest conception of the absolutely different function of the two diphthongs, viz., ái and aí, áu and aú; and, in any case, to give au the sound of aw in law, was utterly incorrect.)

Old French MSS., therefore, had not much better chance of escape than those of our English mother tongue, in its different periods, and have, in fact, as a body, been inadequately edited. I may have been unfortunate in coming across specimens, but the printed editions which it has been my fate to wade through, in search of linguistic material for a reconstruction of our old Norman tongue, have driven me to the conclusion that many of our existing editions of O. F. will have to be re-edited, to be of

I refer to the Comparative Grammar of the Anglo-Saxon language of Professor March, a book in every way excellent, bating certain etymological crudities, such as the statement in p. 75 (b), "Kan in kan-kan [pankan-panchan] is also like the Hebrew kam-es, five"; also some curious notions as to the origin of the case-endings, ex. gr. of the gen. sg. and plur. at § 63, c. b.; and a few inadequately put explana

tions of certain forms, as in § 158. g. tenima, &c.; see also § 103, a. end.

Helfenstein's Comparative Grammar of the Teutonic languages is a very good compendium, and may be recommended.

For an instance of the scandalous manner in which the confidence of the public is at times abused, see note at the end of this article.

any real critical value to the linguistic student. It is impossible to work on the data extant without falling into serious error. For instance, misled at least to a certain extent by the printed texts extant, Mr. Payne (in the Trans. of Phil. Soc., 1868-9), p. 381, says :-"Take, first, therefore, cuer (Norman F.) Mod. cœur), which was, probably, pronounced kuur [koor], &c." To this, at p. 585, Mr. A. Ellis, in a note, states his own incompetency to speak on this point. For myself, I can only say that Mr. Payne seems to have adopted the very sound that the Norman had not. I may be permitted to quote here the results of my own observations on a poem in Old French, which I am editing. Referring to this identical word: "It is somewhat difficult to decide which is the form of the word (in 1. 104) that should be adopted. The MS. has fully written out five 'quoer,' eight 'quor.' The contraction qr, which occurs 15 times (together with four times in the word reqr) should be written 'quor,' of course, though in the only case where any test can be applied, viz., at the end of a line, it is written qr, but rhymes with per. In pronunciation there is not any reason to suppose that it was at all different from what it is now. (For this rhyme with per cf. the German rhyme entbehren and schwören, to be noted in the spelling queur in 158.) In the compound verb where the forms occur, it is to be noted that in the present tense (where alone it is met) the 1st sing. has reqr (all the four times), but the 2nd sing. reqrs (= requers), and the 3rd sing. also reqrt (= requert), while the 2nd plur. has requerez, written fully. [The simple verb i sg. has quoer (1. 1761) fully]. Besides 1. 104, where it is qf, though rhyming with per, the other four cases where it rhymes with er (1. 205, 685, 1348, 1632) are written out quoer. It is never rhymed o, but there occurs a play on words thrice in the poem, showing that the sound was very close to the open o:—

365. Ke hem vus face au cors u au quor maufé.

1350. Cist sunt martir de cors, cist de quor duluser.

1470. Si cors est las, mais sis bons quors tut frais est e nuveus.

Vie de St. Auban."

What Mr. Payne means by coupling together such works as "the Life of Edward the Confessor," and "the Conquest of Ireland," in the same breath, as pure Norman texts, I am at a loss to conceive. Anything more impure than the French in "the Conquest of Ireland," it would be difficult to find, whereas the forms in "Edward the Confessor" (as may be seen in spite of Mr. Luard's editorial misdeeds) are remarkably accurate and free from foreign admixture. But, as the works on which he (Mr. Payne) mainly relies are these editions, it may be inferred from the following critique on Mr. Luard, that there is but small chance of any really scientific induction on such data. The editors he has followed are M. Francisque Michel, Mr. Wright, and Mr. Luard. With the former, as being a foreigner, I have nothing especially to do (though his few notes and explanations are given in English); one specimen shall suffice to show that he is not, however, an infallible guide.

In 1837, M. Michel published [Pickering, London] a little poem in Anglo-Norman, "On the Conquest of Ireland," with introductory essay by Thomas Wright, &c., &c. Among other curiosities contained in said poem, here is

one:

P. 33, 1. 670. Maurice Prendergast is signifying to his companions how it would be better for them to fight with the Irish; he says:

Kar armés eymes le plusurs,
Vassals hardis e combaturs;
E les traiteres sunt tut nues,

Haubers ne bruines n'unt vestues;
Pur ço, si turnum ...
N'averunt-il de mort garant.

For we are armed the most (of us),
Vassals hardy and fighters,

And the traitors are all naked

With no armour,

Therefore, if we, &c.

They will have no protection from death.

Here the opposition is plain between the armed English knights and the naked Irish caterans, and whatever may

be said of the rest of the text, armés eymes, we are armed, is plain enough. But in his notes M. Michel actually translates this

"For the most part of you, you like arms" (!)

So then, eymes ought to mean you like, which it most assuredly does not.1

From M. Michel we proceed by easy transition to Mr. Thomas Wright, who, in 1841, published, for the Historical Society of Science, "Popular Treatises on Science in Anglo-Saxon, Anglo-Norman, and English." Of these we are concerned at present only with the Anglo-Norman. This is a treatise by De Thuan, called "Li livre des Creatures," and is accompanied by a translation. And this translation is certainly very bad.

Here are a few specimens of the errors in Mr. Wright's edition and translation, culled from amongst the three or four hundred that are committed within less than 3000 lines !

al buver.

L. 65. Coe [sic] dit en repruver li vilain "So the vilain said in reproof to the drunkard" (!) (v. 276, also.) Buver has nothing to do with drinking, it is simply bouvier, "neat-herd."

L. 108. It may be seen here how poor a basis Mr. Payne has for speculation as to the termination rie, supposed equal to ire in rhyme. Cf. also 1. 264, 693. The line in Mr. Wright's text is—

Pur cel me plaist à dire, d'iço est ma materie.

Now, as dire cannot be otherwise than dire, the accompanying rhyme must be mateire. To edit materie is inadmissible; both measure and rhyme are flatly against it.

The poet says he will instruct both clergy and laity,

For another instance in which eimes (the same word) has been misunderstood by this savant, the reader

may consult Burguy's Grammar, Vol. I., p. 270, note (1).