from the closer to the more open vowel-if both are fully pronounced, there is a natural tendency to evolve the sound of y; for example, eat-eë-ate, or yeät. On the other hand, in words beginning with a vowel lower down in the series, and passing up-as in such combinations as oa, oi, for instance -the tendency will necessarily be to produce the sound of w. Bearing in mind the characteristic tendency of the dialect to lengthen and open the vowel-sounds, the process will be something like the following:

There is no tendency towards any such pronunciation in current English, simply because, as I have said, no attempt is made to sound both vowels, the two being crushed or commuted into one-ea and oa into e and o respectively. The initials y and w are prefixed in order to prevent this crushing of two vowels into one, which is contrary to the whole spirit of the Somersetshire pronunciation; and you will thus see how systematic the whole process is, and how completely it accomplishes its end.

It is, moreover, thoroughly Anglo-Saxon. Looking first at the initial y, it is a universal law of Anglo-Saxon pronunciation that the initial e before a or o is sounded as y. From what has just been said, you will now understand the reason of this, and see how it must be so, as we see in the few English words where both vowels are still sounded,

as in ewe, a sheep, pronounced yoo, and ewer, a watervessel, yooer, both of Saxon origin. And in cases where there is an apparent inconsistency, the dialect will be found faithful to the older tongue. In the class of words already referred to, beginning with a singly, where the a is treated as ea, the Somersetshire pronunciation represents the older form of the word—e. g., ale and arm, sounded in Somersetshire yeäle and yeärm, are in Anglo-Saxon eale and earm.

The initial w, too, dates back to the older tongue. The Anglo-Saxon, indeed, prefixed w and aspirated w to many words where it is now altogether lost in English, and sounded it in others where it remains only to the eye, being silent, or all but silent, to the ear. In the words whole, wholesome, wholesale, for instance, the w is not sounded at all, and in who, whose, whom, it is only indirectly heard in the modification of the vowelsound it determines. The initial w is, indeed, quite archaic, the universal tendency of languages being to shorten, condense, and cut off both initial and final vowels as much as possible; and this full 00 sound of the Somersetshire pronunciation dates. back not only to Anglo-Saxon, but, in all likeli hood, through it to the old Gothic, of which it is a characteristic feature. The word hoop, for instance, in hooping-cough, appears in Gothic as hoo-opan; in Anglo-Saxon as hwòpan, or (the accented o being equal to oo) hwoopan; in English as hoop; for though whoop does still exist in spelling and pronunciation, it is rarely used, hoop being the common form both to the eye and the ear.

As there is a class of words in which the initial a is treated as ea, so there is another in which the initial o is dealt with as oa; but the vowel in these cases generally stands for the Anglo-Saxon accented a, which had a broad diphthongal sound, represented in English pronunciation by o (as bán, bone; stán, stone, &c.); often, however, in spelling by oa (as fám, foam; ár, oar, &c.); sometimes by oe (as fá, foe; wá, woe, &c.); but which in Somersetshire is really pronounced oa, as in home (A.S. hám) — whoäme; oak (sounded oke in current English, the a being silent), (A.S. ák), woäk, &c. Thus, even in its apparent irregularities, the dialect, as compared with the literary language, is still the more faithful and consistent representative of Anglo-Saxon sounds.

Only one other point remains to be glanced at in relation to y and w, and that relates to another apparent inconsistency. While, on the one hand, the dialect prefixes y and w to a number of words where no such letters exist in current English, on the other, it throws them away in many cases where they really do. In the mouth of a Somersetshire man, for example, yes and yesterday become eeze and eezeterday—will and would become 'ool and 'ood, so that really a spirit of contradiction seems to be at work. A little examination, however, will show that this is not so. It must be observed that in these cases there are not two initial vowels, only one, so that the semi-consonants are not prefixed in order to bring out a double vowel-sound. On the contrary, they are produced by crushing an original long vowel into a short

one, and thus represent the result of that condensation. The Somersetshire man, however, delighting in vowels, will not endorse any such process. He accordingly resolves the semi-consonant and short vowel into the original long vowel, and yes accordingly becomes eeze; will, 'ool, &c. "Vowels, vowels," is his cry-"the more and longer the better:" accordingly, as he prefixes the semi-consonant in order to make one vowel two, so he throws it away in order to make a short vowel long. The same principle is at work in both

cases.

What has already been said about the letter w will fully explain the triphthongs to which I referred at the last meeting, and which are so marked a feature of the dialect. These occur in words having the vowel combinations oi, oa, or oe; and, as you will now understand, in order to bring out fully the sound of both, it was necessary to prefix the sound of u or w to the first. The following are examples :

This, too, is thoroughly Anglo-Saxon, as such forms as sweord, a sword, and cweorne, a mill, would sufficiently prove. Anglo-Saxon is, indeed, most probably the only language in which such combinations ever existed, as the Somersetshire is the only

living dialect in which they are fully and familiarly pronounced.

The first part of the inquiry closes here, and I think, as the result of it, that the Somersetshire pronunciation—in many features of its vowel system, at all events-may fairly claim to be a tolerably good representative of classic Anglo-Saxon.

I will now look for a little at the consonants; but the evidence in this case being much less minute and conclusive than in that of the vowels, they can be dealt with in a more summary manner. I will follow the arrangement already laid down, and take up the four classes of consonants in order.

The first to be considered are D, T, and TH. There was a tendency in Anglo-Saxon to change T and TH into D-or rather this substitution was certainly in many cases made, though, at the same time, it must be confessed, not in the regular, characteristic way which marks the process in the Somersetshire dialect. That the substitution existed, and was even not uncommon, is shown by such examples as these:

In the first canto of Beowulf, it is said of his