the former has undergone no vowel-change in the solitary expression in which it still survives, namely, methinks. These impersonals are now largely obsolete, being superseded by some other form of expression, or by the personal use of the verb. Thus Chaucer's What nedeth wordes mo?" would now be, "What need is there for more words?" The nedeth, or later nedes, being transmuted from the verb need with the termination -es into the noun need with the verb is, and two previously superfluous words being inserted. Spenser's sentence, "It would pity any living eye," would now be, "Any living eye would pity." It repents me has quite given way to "I repent." "It likes me," "it dislikes me," are now "I like," ‚”“I dislike.” “If it please you”—an expression which is the literal translation both of the French "s'il vous plaît" and (as the position of the ist shows) of the German "gefälligst"-is now only used in formal speech: the familiar use makes the you nominative instead of accusative, and the verb personal instead of impersonal-"If you please." This change was already beginning in Shakspere's time, for (as Dr. Abbott has pointed out) both forms exist-"So please him come," and "If they please;" and while the common phrase was "Woe is me," we find in the Tempest, "I am woe for 't." Where a noun is used instead of a pronoun, as in "So please your highness," there is no inflexion of the noun to guide us, but from the antiquated form of the expression one cannot but infer an antiquated syntax also, and that "highness" is meant to be the dative case.

But we have not yet done with this zimmith, or the verb zim, or (as I myself have more commonly heard it pronounced) sim; that is, seem. This verb is used not only in the sense of to appear, but also to think. Can this be explained? Yes; the change of "It sims to me" into “I sim is precisely of the same character as the change of "It pleases me" into "I please," only this latter is fashionable in modern times, the other is provincial and unfashionable. But there is one point more. In Greek (as every Greek scholar knows) there is one verb that bears exactly this same double meaning, and signifies both I think and I seem; but in this case we cannot be sure that the history of the meaning is the same as in the Devonian I sim, for both senses of dokéw are found in the earliest literature of Greece; namely, in Homer's Iliad. In aρк however, as used by Eschylus in Prom. Vinct., 639 for aрkei-"I suffice" for "It suffices that I". we have a clear case of the personal use of the verb substituted for the impersonal, such as one cannot but suspect in

dokéw, and find unquestionably in many of the expressions cited just above.

Let us next turn to a certain expression of time. We all remember how in the Gospels our Lord says, "I have compassion on the multitude because they have continued "—I take the liberty of translating with the true English idiom here, rather than follow the Greek idiom in using the present tense "they continue' -"with me now three days." This word "days" is of course the accus. in English, as indicating duration of time; the rule is the same also in Latin and Greek. But, strange to say, in both Matt. xv. 32 and Mark viii. 2, where these words of our Lord are recorded, the majority of the most ancient MSS. give the "days" in the nom., uépai; not, as in later MSS., uépas. About the meaning there can be no question; simply the phrase is elliptical. We need to supply some part of the verb to be and a relative pronoun: "There are three days during which they have continued with me." Can then our western dialect exhibit any usage at all analogous to this? Yes, imperfectly analogous; that is to say, with a partial ellipsis, the ellipsis being partly supplied. It is a form of expression I have often heard; but I prefer as usual to fall back on the authority of printed books. Nathan Hogg writes

"Last Thesday wiz week, as you naws, brither Jan,

The yung squire ta Tor Abbey becom❜d twenty wan;"

that is, "Last Thursday was a week since the day on which the young squire," &c. Both here and in the Greek in the Gospels a phrase of the nature of a relative adverb needs to be supplied.

One remark leads to another. "Tha yung squire ta Tor Abbey," for "at Tor Abbey." Here we have a genuine Devonianism. "Ur 'th a-been stayin ta Plympton;" "ta lass" for "at last;" "aul ta wance;'

"Thin thare wid turn up zich a rattle,

As ef whole urgmints waz ta battle."

Of this use of to-or in Devonshire ta* [Henrietta]-for at, we have just the converse in Icelandic and the other Scandinavian languages, as compared with A.S. and English, in the use of at instead of to with the infinitive mood of verbs; as to make, or as the Scotch say to gar, is in Icel. at gera, Dan. at gjöre, Swed. at göra [g = Engl. y; ö as in German].

* With the change of vowel here, and in the Dutch and Flemish te, we may compare se for so in Hali Meidenhad.

But in other Teutonic languages to is in many instances employed where we use at, no motion being implied. Thus in German, zu Hofe dienen, to serve at court; das Waisenhaus zu Halle, the Orphans' Home at Halle; and corresponding to the Devonshire ta lass we have zuletzt. So also zu Hause, at home, which in Old Saxon is to hûs, in Du. te huis, and in Old Flemish te huus [goose].

The mention of this Dutch huis compels me to return for a few moments to Pronunciation. For who of us does not know the peculiar Devonshire mode of sounding how, now, cow? The recognized pronunciation of these words is with a sound which it is very difficult to analyse, but it seems to begin with Ŏ [hot] and glide into oo [boot]. The same diphthong (as it is commonly but inaccurately called) in Essex, and commonly in London, begins with ee [there] and finishes with oo. But in London another mode of sounding it is with ǎ [hat] to start with, gliding as before into oo. Similar is the Dutch sound in beginning with a, but it finishes with u [Fr. tu]. But what of our Devonshire ow? As well as I can analyse it, it begins with ou [Fr. sœur] and glides into the Devonshire u. It is the most remarkable sound in our dialect. And in several of these expressions there is yet another point of resemblance to Devonshire-the omission of the article. In current English we say "to court" when we mean "to the king's court," and so in German and the kindred dialects, as in the Old Saxon of Reineke de Fos

"Ok kwemen to hove fele stolten gesellen;"

that is," Also there came to court many proud people," and in Old Flemish te hove in exactly the same sense. But in Devonshire, if a man speaks of going out of a house into the court adjoining, he will say "ta kuart:" "Ha went ta kuart ta vetch tha hood." So also inta houze, into the house; "hom ta vawr dore," home to the fore door; "hur went and kimmitted tha wier ta vlame."

Again, in current English we use the verb tell always as a transitive verb with the thing told (that is to say, the information communicated) as its direct object, expressed or understood (very often in the form of a noun sentence), and with the person to whom the thing is told as the indirect object, expressed or understood. But in Devonshire tell is often intransitive, as, "Go owt an yer min tul;" i.e. talk. And in the Devonshire Dialogue, "Her used to tell to her flowers." Precisely analogous to this is the use of Aéyw for λaλ in late Greek; as for example in the Gospel of John, c. xiv. 10, in

the best MSS. Aéyw is so used, accompanied indeed by a cognate accusative, but intransitive. So also in the forensic use of the word, as in Soph. O.T. 545, and Acts xxvi. 1.

There remain yet a few points that ought not to be passed over in silence. The origin of prepositions is one of the most difficult problems the philologer has to deal with. In seeking to trace their earliest history the investigator soon finds himself involved in a thick mist, where it is impossible to see any object clearly and well defined. In such a mist it avails little to throw the reins on the neck of imagination, and gallop madly along the path of wild conjecture. The species of legerdemain practised by some etymologers really justifies Voltaire's sarcasm when he wrote that Ki and Atoës were names of an ancient emperor of China, or rather they were different forms of one and the same name; for a philologer would simply change the K into A, and the i into toës, and the transmutation is complete. Now suppose a reader of the Ayenbite of Inwyt meets with the sentence, "pis we bezechip toppe alle ping," he may see clearly that the meaning is, "This we pray above all things;" but how can toppe come to mean above? He may go to Donaldson's New Cratylus, and learn the force and significance of every letter of every preposition in Greek, Latin, and Sanskrit, and may perhaps believe what he reads; and he may plod through the chapter in which the use of certain nouns as prepositions is discussed, where he will find it proved to demonstration that dikŋ and Xápis are only different forms of the same word, like Voltaire's Ki and Atoës, and much truly will he have learnt about the Old Kentish phrase "toppe alle thing." Suppose now, when he has recovered from his bewilderment, he betakes himself to Nathan Hogg by way of distraction (as the French say) after his Donaldsonian toils, he will find on the first page, "Tha Daysy tap tha Grave," and will recognize the word at once. Elsewhere he will find "pin tap the hadges," and again "pin tap uv tha vier," the phrases tap, pin tap, and pin tap uv being evidently equivalent to one another and to the thrice-recurring toppe of the Ayenbite of Inwyt. But moreover he finds pin used without tap, as "pin me wurd;" "pin axin tha vally;" and he also notices apin similarly used, as in "apin crassin tha strayt." He now has no difficulty in discerning that apin is the Devonian form of upon, that the definite article is understood, and that tap or toppe when used as a preposition is really the familiar noun top with an ellipse-upon the top of being the full and complete phrase.

Another elliptical expression is outel dores. Here, as with tap, of is understood, as it is also in out doors and indel doors. But what does outel mean? It means "the out deal

of;" that is to say, "on the outer side of." In outel the d has disappeared, absorbed in the final t of out, while in indel it remains. Deal, from A.S. dælan, to divide, signifies primarily a part, and from this radical notion all the other senses of the word can be easily traced. In this case a part of the door comes to signify one side of the door as opposed to the other, and then the portion of space which is on the one side or the other side of the door.

Certain verbal forms are worth a brief notice. From the French ho-là we have derived the verb to holla, which is sometimes confounded with the adj. hollow on the one hand and with the interjection and verb halloo, A.S. calá, on the other. Such is Professor Skeat's view. But in Devonshire holla is cut down to holl. From the adj. stiff we have the verb to stiffen. In Scotch the adj. has the form steeve, as in the lines

=

"A fiery ettercop,
A fractious chiel,
As het as pepper,
An as steeve as steel;"

and this very form steeve is used in Devonshire as a verb, the verbal termination being dropped. So in Scotch in the adj. and verb deave deaf and deafen. We have all heard how the notorious Viscount Dundee found time one day amidst his cruel persecution of the Covenanters to call, as a matter of curiosity, upon an old lady whose age far exceeded the ordinary limits of human longevity. His name Claverhouse was commonly contracted in the Scottish mode of speech into Claver'se, and clavers means noise or din. So when he asked the old lady, who seems to have entertained as little affection for Presbyterian zeal as she had respect for persecuting fury, what difference she observed between the days of her childhood and those of her age, she replied, "Then there was ane Knox that deaved us a' wi' his clavers, an noo there is ane Claver'se that deaves us a' wi' his knocks." But we can find an analogue to this steeve for stiffen without travelling north of the Tweed. For do not our poets familiarly cut off the termination of open and use ope as a verb? In two other instances indeed a termination is dropped by which we change a noun into a verb, and the Dev. Dial. uses the noun itself as a verb. Thus hap is used for happen, and carr for carry. But while we now think these vulgarisms, they are more