The use of the grammatical term, nnesis and diaeresis is also an act of anachroniam. %%lila it is the ordinary law of language to aggluti nate independent words, and to crush adjoining vowels together, no doubt at time* a poet may find it convenient to avail himself of earlier obsolete forms which were in use before such agglutination or crushing had taken place. But to treat hessian' as deduced from A./Tobin and call the ',receipt diaeresis, or to call it heals, when the old poet himself, before the agglutination of a preposition to a verb was yet effected, wrote flair 111r0 A0170, coma, or when Ilerodotua under like circum stance', wrote ere por stare A erepareser—this is simple perversity.
Again a large amount of unsound philology is concealed under the term inetatlaais. Although no doubt there occur instances where a ktter shifts Its position iu a word either through in the person wbo repeats it, or by a sort of involuntary stuttering, as, for example, In the French frangc, Eng. from the Latin "atria, as though it had been frionizia. The Interchange of sk with is is no doubt common In language, but here the real explanation is probably that for the tsars of many people the sounds a, or perhaps abi, and k or g, are no akin as to be readily interchangeable ; and thus we have had In this inland the names fro"-, frox, fresh, and frog, all applied to the same reptile; and this interchange would naturally carry with it the Interchange of sp and ps in those countries where a k is supplanted by a p. Similarly when a Spaniard substitutes 'allegro for the Latin miracido, the sub stitution off for r, which occurs in the first syllable, was for him a change of the faintest character, and this change once made encouraged the converse change in the third syllable. This however is no meta thesis. As the erroneous doctrine here combated has established itself only too firmly among philologera, it is the more important to point out the cause which has led to the error. No doubt abundant instances can be found like (birch, Germ., and our through, Opeuros and Oapeot ; tero, tie/. Here there in an apparent passage of the vowel from one side of the liquid to the other ; but it is only apparent, the truth being that in all such cases the older form was possessed of both vowels. For the first pair of words just quoted, we still have the full form thorough ; and the truth is that the word consists of two elements Mots "a door," and a diminutive suffix ough. Oaperos and Opacrot are to be explained as alike corruptions of a fuller Bap-aa-ot, in which Bap represents the chief element, being the equivalent of our dare, according to the usual law that holds between the languages of a B and d corresponding. Time
second syllable as is that so commonly found in secondary verbs of the Greek language, as rap-wrow, ap.aoow. The case of fern beside triri and tritutn is explained by the Greek 7p113-w, and better still by the Latin ter-eb-ra, where the base is ter, eb a suffix, seen also in frem-c6 wades, gent-eb-undus ; which it is the less necessary to dwell on hero as it will come under consideration again. What has tended to encou rage the leap-frog doctrine here assailed, is the strange coincidence, as it may well appear, that the vowels should so generally be the same. The answer is to be found in the law above-mentioned, that in the composition of words there is ever a strong tendency to bring dia. cordant vowels into harmony. Thus the Greek verbs rap-cum-co, ep-eaa-w, eat-too-w, op-wax-to, have in truth a common suffix, winch takes the several varieties of vowel to please the preceding syllable. Another interesting case of the same principle is seen in the series of words, all one in origin; TaAAa"- " it young man,' Lat. petite- " a young woman," a euphuism for a concubine, Eng. fn/lie " a young mare; " " a foal," Scotch potluck " a young fish or crab,"pid/o- (nom. indite) "a foal" or "a chicken,"—where the first syllable seems to denote " young," and the suffix is of diminutive] power, so that the words in themselves merely signify "a little young one." The other figures, such as apharesis cutting off the head of a word," syncope " the compression of the interior of a word," aporepe " the cutting off the tail of a word," together with the terms crash, synaloephe and syneresis, which deal with the compression of adjoining vowels, tell no untruth, and no may be left to. the use of those who have an affection for hard words. The one general principle under which they fall, and this alone is of moment, is, that words are ever subject to diminution. No doubt false analogies may occasionally lead to a violation of this general principle. Thus our island and sorereiya, as now written, have letters in them to which they are no way entitled, through a mistaken reference of them to the old French isle (fir), and the word reign or reynum. A more justifiable form would have been iland (like the German ei-land), and sorran (Ital. sorra no).