It is important to observe how the consonants of different nations differ. For instance, the German p and b are intermediate to the English p and b, so as to be difficult to our ears to distinguish, and the Armenians have two different p's. So the English h is intermediate in strictness to Ti and n, if at least we assume that these Hebrew letters had the sound of the Arabic 3 and Now this is a general phenomenon, in comparing the Indo European with the Syro-Arabian sounds. Our k is between the two Hebrew or Arab k's: our t is between their two t's; and so on. To explain this, observe that we may execute a t in various ways ; first, by slapping the tongue flat against the teeth, as an Irishman or a man of Cumberland does when he says water; secondly (what is rather less broad), by slightly touching the root of the teeth, as a Frenchman or Italian does ; thirdly, by touching only the gums, which is the English method ; fourthly, by touching the palate, or by pressing on the gums with a muscular jerk. One or other of the last is the Hebrew 0, the Arab ; hence some call it a palatal, others a strong t. In touching the palate, the throat is involuntarily opened, and a guttural sound is imparted to the letter and to the following vowel; for which reason it has been also called a guttural t. The other method, of pressing the tongue firmly, but not on the palate, is an Armenian t, but perhaps not the true Syro-Arabian.
What we have here to insist on is, that differ ences which with us are provincialisms, with them constitute differences of elementary sounds. To a Hebrew, n differs from D, or 2 from p, as decidedly as with us p from b. On the other hand, t and th (thin), as d and th (full), which with us have an elementary distinction, are but euphonic variations in Hebrew.
After this, we have to explain that 2 was ori ginally sounded forwarder on the palate than English k, as p was far backwarder, at the root of the tongue. So D was probably forwarder, and V certainly backwarder than our s, each of them being nevertheless, a kind of s. That V was not is is seen by riv, fov, r.nr), etc. etc., which are written Ze;\Xci, Meo-patv, etc. etc. in the Sept., as well as from the analogy of the Arabic The is pronunciation is a late invention, as is the ng sound, which has been arbitrarily assigned to 31. Nevertheless, out of my the Greeks made Thpos, which is contrary to the analogy of ;Stacsov for : yet the adjective Sarrrznus, instead of Tyrius, used by Virgil, may prove that Sarr or Sour was in ancient, as in modem days, the right pronunciation of Tyre. In English we have the double sound s and sh, which is illustrative of n and D, D and p, etc., to which modification it is closely analogous. For sh is only a modified s, being formed with the broad or central part of the tongue, instead of the tip. In this action the fore part of the tongue forms itself into a sort of cup, the whole rim of which comes near to the palate while the breath rushes between. On the contrary, in sounding V, only a single transverse section of the tongue approaches the palate ; but this section is far back, and the lips are protruded and smacked, so as to constitute a mouthing s. Farther, the alliance of r to s, so strongly marked in the Greek and Latin languages, justifies our arranging them in one row. The r is formed by a vibration along the tongue, which bears some analogy to the rush of the breath along its surface, on which the s and sh depend. The Armenians have a twofold r, of which one, if we mistake not, is related to the other, as our sh to s.
The Hebrews were commonly stated to have given two sounds to each of the letters 1 n so as to produce the twelve sounds, p f, by, t 0, d 5, k x, g 7; but it is now generally admitted that it was not so originally. The Greeks (at least provincially), even in early days, pronounced Mira, Vita, as they now also say Gharnma, Dhelta ; and the Italians for Latin b sometimes have v, some times b. The Hebrew corruption was however so early as constantly to shew itself in the Sept.; indeed, as a general rule, we must regard the thin consonants n n z as having assumed the continu ous, instead of the explosive, pronunciation ; i. e.
they were become f, 0, x. Thus 11/)2 are written 4'w-(.2,v, op€X, Xavaciv, in spite of the dagesh lease by which the later Masorites directed the initial letters to be sounded P, T, K. Yet
there is no immovable rule. Thus the tom is in the same book variously rendered XETTELE44 and Ktngmu (i Mace. i. r, and viii. 5). It will be observed that a decidedly dental t is very near to th, and a k, very mincing and forward in the mouth, easily melts into ky, as in the Turkish language, and thence into soft x. In this way, 0 and x having been adopted for n and 2, T and K were left as the general representatives of D and p. It is well known that the Ephraimites at an early period said s, at least in some words, for sh, as in the celebrated tale of Shibboleth ; but this corrup tion went on increasing after the orthography had been fixed, so that it became requisite to denote by a dot many a CI sh, the sound of which had dege nerated into D s. It is rather perplexing to find D occupy the same place in the Hebrew alphabet as in the Greek, a fact which perhaps still needs elucidation.
But we must turn to an important subject—the tendency of aspirates to degenerate into vowels. The muscular language of barbarians seems to love aspirates ; in fact, a vowel energetically sounded is itself an aspirate, as an aspirate softened is a vowel. Let it be noticed in passing that an over-vocalised language is by no means soft. Such a word as Inte has of necessity strong hiatuses between the vowels, which hiatuses, although not written in Western languages, are virtually consonantal aspi rates ; in which respect an English representation of some barbarous languages is very misleading. The Hebrew spelling of Greek names often illustrates this ; for example, Antiochus is D:12iN+DM, where the central ti indicates the hiatus between i and o. That the letters 71 (final), +, from the earliest times were used for the long vowels A, I, U, seems to be beyond doubt. At a later period, perhaps, L.: was used for another A : the Greeks adopted 3/ for 0, and finally 17 for a long E. It is probable that a corruption in the Hebrew pronunciation of n and n had already come in when the Sept. adopted the spelling of proper names which we find. As for 71, it is the more remarkable that the Greek aspirate should not have been used for it ; for both in Greece and in Italy the it sound must have been very soft, and ultimately has been lost.
So we find in the Sept. 'Apex for 5nn Hebel, '12o* for Veijn Hoshe'a ; and even the rougher and stronger aspirate n often vanishes. Thus 'Eviox for Hhenok ; 'Po4360 for Rehho bot, etc. Sometimes, however, the n becomes x, as in Xiiih for an, XaXax for n55 ; which may possibly indicate that n, at least in proper names, occasionally retained the two sounds of Arabic and .nd hh and • Liz The I/ was of necessity omitted in Greek, since, at least when it was be tween two vowels, no nearer representation could be made than by leaving a hiatus. Where it has been denoted by Greek 7, as in P6aoiVa, PatSciS, Itryibp, there is no doubt that it had the force of the Arabic a, (ghain), whether or not this sound ever occurred in Hebrew except in proper names.
Respecting the vowels, we may add that it is now historically established, alike in the Syro Arabian and in the Indo-European languages, that the sounds a and 5 (pronounced as in maid and boat) are later in time than those of d, fi, and are in fact corruptions of the diphthongs ai, au. Hence, originally, three long vowels, fi, with three vowel-points for the same when short, ap peared to suffice. On the four very short vowels of Hebrew a needless obscurity is left in our gram mars by its not being observed that we have the same number in the English language, really dis tinct ; as in smicren (or castle"), contrary, noMdy, Mneatit; although it is probable that with 11 the vowel was clearer and sharper than in any short English a. We have even the furtive vowel of which the Hebrew grammars speak ; namely, when a word ends in r, preceded by a long accented vowel or diphthong. In this case, a very short a is heard in true English speech, but not in Irish, before the r, as in beer, shore, flour (whence the orthography flower, bower, etc.), which corresponds to the Hebrew nn, In. The Arabs have it also when the final letter is p.—F. W. N.