In the last decade of Alfred’s reign, when he was in the agony of that supreme crisis which tested the value of his institutions, a great muster of force was called for, and the extent of the contributing area is sketched by the annalist as matter of amazement. ‘There gathered Æthered aldorman and Æthelm aldorman and Æthelnoth aldorman, and the king’s Thanes who were then at home in the fortifications, from every garrison east of Pedrida (whether west of Selwood or east), likewise also north of Thames and west of Severn:—moreover some part of the Welsh nation[31].’

Here we mark the startling novelty that the Welsh in 894 are seen aiding the Saxon against the Dane; and we can hardly forgo a passing cry of wonder and pleasure at this signal token of the imperial success of Alfred’s policy. But our present concern is with the recognition of Pedrida as the westernmost limit of Wessex proper instead of Selwood, and the implication that the change was recent. We see that Selwoodshire (as the intervening district was popularly called) was by 894 quite assimilated and included in the military administration of Wessex, but that beyond Pedrida some other rule was operative at that time. Such a fact reflects back an illustrative light upon the year 878, and helps us to estimate the situation of Alfred when he was in Somerset beyond Pedrida.

The political division here indicated has left traces which may still be recognized, particularly in the dialect and in folk-lore. Of the dialect we have a remarkable monument in Mr. Elworthy’s works, The Dialect of West Somerset, and his West Somerset Word-Book. Especially to be noted is the ‘u’ of the West Country, which is radically one with the Welsh ‘u’ and with the French ‘u,’ while at the same time it has a very distinct local character of its own. Every Englishman who is conversant with the French language knows how hard it is to acquire the utterance of the French ‘u’ after the age of infancy. A like strangeness is experienced by English people born east of Pedrida, when they attempt to reproduce the western ‘u.’ In fact, this vowel-sound is Keltic; it is a legacy from our British predecessors.

Not that this British ‘u’ is absolutely confined to the western promontory: it may be occasionally heard in other parts of the country by a cultivated and observant ear. Mr. Mayhew once told me that he had heard it in the Corn Market at Oxford. But though not confined to the lands west of Pedrida, it is in a peculiar manner concentrated there. It is chiefly in Devonshire that this peculiar vowel has wakened wider attention, but this is simply because that county has been the most frequented as a place of holiday resort.

The so-called Devonian ‘u’ and its contiguous sounds have been described many times from first to last, but it has been mostly in that perfunctory vein which contents the summer tourist. It is rare to catch such a plain and solid illustration as the following, which is quoted from the preface to Mr. Elworthy’s West Somerset Word-Book:—‘I was a passive listener at Brandon’s while a bonnet was being discussed, and when making the payment ventured to remark to the young lady, “You must have been a long time in London.” “Oh yes, ten years; but why do you ask?” “Only for information,” said I. “And did you come straight from Teignmouth?” With much surprise at my supposing she came from Devonshire, she said at length that she was a native of Newton Abbott. I could not pretend to define the precise quality of her two, but it was only in that one word that I recognized her locality.’

If the vocabulary of this dialect were minutely examined by a competent Welsh scholar, some British words might be detected. Among those which would deserve early attention are plum (soft, as a bed), pilm or pillum (dust), welt (to beat, thrash).

Another local characteristic of the West Welsh promontory is this, that it is the peculiar haunt of a race of whimsical or mischievous sprites called Piskies or Pixies. In South Devon and Cornwall any one whose conduct is strange and unaccountable is said to be pisky-led. This is a branch of the numerous kindred of that versatile Puck, whose memory is kept fresh by the Midsummer Night’s Dream. In an Anglo-Saxon perambulation of land at Weston by Bath, we meet with a Pucan Wyl, Puck’s Well[32]. The English Dialect Dictionary preserves the name of Aw-Puck for Will-o’-the-Wisp or ignis fatuus, a compound which imports that he is the most dangerous of the species. This name was current in Worcestershire, but is now obsolete [33].

These are the more obvious extant traces of the long isolation of the trans-Pedridan world: others there are which have attracted inquiry, such as peculiar customs, implements, songs and song-tunes, which latter have been investigated by Dr. Bussell and the Rev. S. Baring-Gould.


The Somerset to which Alfred retired was widely unlike the Somerset of to-day. In this respect three points may be taken: (1) Differences in the distribution of land and water; (2) differences in the trees and woods and game; (3) differences in the political aspect of the population.