Page:Notes and Queries - Series 10 - Volume 1.djvu/385

 s. i. APRIL 16, low.] NOTES AND QUERIES.

317

letter representing a real sound], excepting, of course, d after n [not ], and similar well- known insertions due to phonetic causes, is quiteanother matter," &c. The " excrescent d " after n, as in Craigend for craigean, is due to confusion with words like sound (from French son), and is the very thing I meant. It is extremely common, and is explained in my 'Principles of Eng. Etym ,' first series, p. 370, with many examples. Of course, in this case it is entirely wrong, because (as I suppose) the accent doesnot fallen the suffix -ean ; but it became possible by confusion with other cases. Precisely parallel to the excrescent d in sound is the excrescent 7) after m, as in Hampstead : I explain this in the same work, p. 373, and cite as examples em-p-ty, glim-p-se, whim-p-er, sem-p-ster : to which add Dem-p-ster. I also show (p. 370) that d occurs, similarly, after I, as in al-d-er (the tree), &c. ; so also Tinwald, where the d shows that some people, at some time, turned the II into Id, whether it is done now or not.

The second w in Wigtown is purely phonetic ; it shows that (it may be long ago) the suffix in this word was once pronounced as in the Scottish toon, rhyming with boon. For, after all, town is merely a variant of tmtn, the Anglo-French form of A.-S. tun (pronounced toon, as above) ; so that Wiy-toivn was once correct. But, of course, the second syllable has long since been reduced to tun by lack of emphasis, and it pleased the Anglo-French scribes to write ton for tun, monk for micnk, honey for huney, and the like, because un (in MSS.) looked indistinct. It is the fact that Wigton, but not Wigtoion, has lost a written w. The difference of spelling indicates that Wig- town is a name of later date than the other, and that is all. Both are now sounded alike.

In words like Carlisle there is no inserted " letter " in the sense I intended : for the s is not sounded. I was referring to words like Tideslow, in which it is sounded. There is, however, an inserted "symbol"; which is a very different thing, and due, of course, to ignorance. The beginning of it was the Lat. insula ; this gave O.F. isle, with s sounded. But in Norman and later French s was dropped before I, m, and n, and the word became really He; yet s was still written, and found its way into island and Carlisle, by mere mistake. Strictly, there is no gain of s, but a loss not only of s, but of n : for we started from the form insula.

Bardroch-wpod is an excellent example ; the ignorant insertion of a written w arose from the fact that theE. wood was frequently pronounced 'ood, as it is still. It was there-

fore inferred (through ignorance) that what sounded something like Bardrpchood really meant Bard rpch- wood. If this belief were to become universal, the sound of w, and not merely the symbol, would at last be estab- lished ; but I seem to gather that this has not yet happened. Still, it may yet do so ; for the force of " popular etymology " is often considerable. The result, even then, would be due to the fact that ivood became 'ood in other cases.

After all, all changes in the spoken names must be of phonetic origin ; for even when due to popular etymology, they must have been suggested by analogy with some change that had such an origin. The case of Tides- well is quite different ; for if the name could be supposed to refer to tide, the name would be tide- well. We can here only explain the actual presence of an s that is really pro- nounced by the supposition that it has always been pronounced.

I conclude, as before, that it is impossible to discuss pronunciations within reasonable limits. If I am obscure, it is owing to the necessity of being brief. I do not believe, any more than I did before, that the in- troduction of letters that represent real sounds into words or names that did not once possess them is at all a common phenomenon ; that is, when we make due allowances for such well-known instances as are found in em-p-ty, thun-d-er, al-d-er, slum-b-er, amongs-t, most of which are due to what has been so happily called " dissimilated gemination," as explained in my 'Principles of Eng. Etym.,' p. 366.

In cases where place-names have been wilfully perverted, it has generally been done by force of a popular etymology that tries to give a new meaning to a word. The worst instances of this character are not those due to unlearned people, but to the shameless and unpardonable meddlesomeness of those who ought to know better, and who imagine they know what is correct when they are all the while in the blindest ignorance. Place- names are best preserved when they are left in the keeping of the illiterate, who speak naturally and are not ambitious to be always inventing theories. WALTER W. SKEAT.

COBWEB PILLS (10 th S. i. 205, 273). In the spring of 1871 1 was staying at Wakefield, in the house of the Rev. Thomas Pearson, an old West Indian missionary. I was making merry over Wesley's ' Primitive Physic,' and particularly over cobweb pills as a remedy for ague, or for anything. Mrs. Pearson quietly observed, "You may laugh, but I