for no very obvious reason. Such a word as yonder is common enough still; but its corresponding adjective yon, as in the phrase “yon man,” is usually relegated to our dialects. Though it is common in Shakespeare, it is comparatively rare in the Middle English period, from the twelfth to the fifteenth century. It only occurs once in Chaucer, where it is introduced as being a Northern word; and it absolutely disappears from record in the tenth, eleventh, and twelfth centuries. Bosworth’s Anglo-Saxon Dictionary gives no example of its use, and it was long supposed that it would be impossible to trace it in our early records. Nevertheless, when Dr Sweet printed, for the first time, an edition of King Alfred’s translation of Pope Gregory’s Pastoral Care, an example appeared in which it was employed in the most natural manner, as if it were in everyday use. At p. 443 of that treatise is the sentence—“Aris and gong to geonre byrg,” i.e. Arise and go to yon city. Here the A.S. geon (pronounced like the modern yon) is actually declined after the regular manner, being duly provided with the suffix -re, which was the special suffix reserved only for the genitive or dative feminine. It is here a dative after the preposition to.

There is, in fact, no limit to the good use to which a reverent study of our dialects may be put by a diligent student. They abound with pearls which are worthy of a better fate than to be trampled under foot. I will content myself with giving one last example that is really too curious to be passed over in silence.

It so happens that in the Anglo-Saxon epic poem of Beowulf, one of the most remarkable and precious of our early poems, there is a splendid and graphic description of a lonely mere, such as would have delighted the heart of Edgar Allan Poe, the author of Ulalume. In Professor Earle’s prose translation of this passage, given in his Deeds of Beowulf, at p. 44, is a description of two mysterious monsters, of whom it is said that “they inhabit unvisited land, wolf-crags, windy bluffs, the dread fen-track, where the mountain waterfall amid precipitous gloom vanisheth beneath—flood under earth. Not far hence it is, reckoning by miles, that the Mere standeth, and over it hang rimy groves; a wood with clenched roots overshrouds the water.” The word to be noted here is the word rimy, i.e. covered with rime or hoar-frost. The original Anglo-Saxon text has the form hrinde, the meaning of which was long doubtful. Grein, the great German scholar, writing in 1864, acknowledged that he did not know what was intended, and it was not till 1880 that light was first thrown upon the passage. In that year Dr Morris edited, for the first time, some Anglo-Saxon homilies (commonly known as the Blickling Homilies, because the MS. is in the library of Blickling Hall, Norfolk); and he called attention to a passage (at p. 209) where the homilist was obviously referring to the lonely mere of the old poem, in which its overhanging groves were described as being hrimige, which is nothing but the true old spelling of rimy. He naturally concluded that the word hrinde (in the MS. of Beowulf) was miswritten, and that the scribe had inadvertently put down hrinde instead of hrimge, which is a legitimate contraction of hrimige. Many scholars accepted this solution; but a further light was yet to come, viz. in 1904. In that year, Dr Joseph Wright printed the fifth volume of the English Dialect Dictionary, showing that in the dialects of Scotland, Northumberland, Durham, and Yorkshire, the word for “hoarfrost” is not rime, but rind, with a derived adjective rindy, which has the same sense as rimy. At the same time, he called attention yet once more to the passage in Beowulf. It is established, accordingly, that the suspected mistake in the MS. is no mistake at all; that the form hrinde is correct, being a contraction of hrindge or hrindige, plural of the adjective hrindig, which is preserved in our dialects, in the form rindy, to this very day. In direct contradiction of a common popular error that regards our dialectal forms as being, for the most part, “corrupt,” it will be found by experience that they are remarkably conservative and antique.

CHAPTER II

DIALECTS IN EARLY TIMES

The history of our dialects in the earliest periods of which we have any record is necessarily somewhat obscure, owing to the scarcity of the documents that have come down to us. The earliest of these have been carefully collected and printed in one volume by Dr Sweet, entitled The Oldest English Texts, edited for the Early English Text Society in 1885. Here we already find the existence of no less than four dialects, which have been called by the names of Northumbrian, Mercian, Wessex (or Anglo-Saxon), and Kentish. These correspond, respectively, though not quite exactly, to what we may roughly call Northern, Midland, Southern, and Kentish. Whether the limits of these dialects were always the same from the earliest times, we cannot tell; probably not, when the unsettled state of the country is considered, in the days when repeated invasions of the Danes and Norsemen necessitated constant efforts to repel them. It is therefore sufficient to define the areas covered by these dialects in quite a rough way. We may regard the Northumbrian or Northern as the dialect or group of dialects spoken to the north of the river Humber, as the name implies; the Wessex or Southern, as the dialect or group of dialects spoken to the south of the river Thames; the Kentish as being peculiar to Kent; and the Mercian as in use in the Midland districts, chiefly to the south of the Humber and to the north of the Thames. The modern limits are somewhat different, but the above division of the three chief dialects (excluding Kentish) into Northern, Midland, and Southern is sufficient for taking a broad general view of the language in the days before the Norman Conquest.

The investigation of the differences of dialect in our early documents only dates from 1885, owing to the previous impossibility of obtaining access to these oldest texts. Before that date, it so happened that nearly all the manuscripts that had been printed or examined were in one and the same dialect, viz. the Southern (or Wessex). The language employed in these was (somewhat unhappily) named “Anglo-Saxon”; and the very natural mistake was made of supposing that this “Anglo-Saxon” was the sole language (or dialect) which served for all the “Angles” and “Saxons” to be found in the “land of the Angles” or England. This is the reason why it is desirable to give the more general name of “Old English” to the oldest forms of our language, because this term can be employed collectively, so as to include Northumbrian, Mercian, “Anglo-Saxon” and Kentish under one designation. The name “Anglo-Saxon” was certainly rather inappropriate, as the speakers of it were mostly Saxons and not Angles at all; which leads up to the paradox that they did not speak “English”; for that, in the extreme literal sense, was the language of the Angles only! But now that the true relationship of the old dialects is known, it is not uncommon for scholars to speak of the Wessex dialect as “Saxon,” and of the Northumbrian and Mercian dialects as “Anglian”; for the latter are found to have some features in