substituted Samech for the Sin that was originally written. That SK was the original value of Samech is confirmed by the value of the Greek letter, which corresponds in figure and position to the latter; which is KS, and which was originally SK. It is also confirmed by the etymological relation of D, "a bowl," with σkáón, and, a “sword," with gipos. What the Ephraimites in Judges xii. 6 said, when desired to say Shibboleth (which they, like the Assyrians, could not pronounce), must have been Skibboleth. It is SK, and not S, that a person would naturally utter, who was making an unsuccessful attempt to pronounce SH.

Prior, however, to the twelfth century B.C., the date of the earliest existing Assyrian inscriptions, Samech had acquired the secondary value of ST. In those forms of Assyrian verbs where a T was introduced after the first radical, when the first radical was Shin, Samech was used for the double letter. Thus we have from, representing astakan. Also when the affixes of the third person, all of which begin with S, were attached to nouns ending in T, we have Samech for the ST, which, according to a well-known Semitic analogy, would be substituted for TS. That K is particularly liable to be replaced by T, which is easier to be pronounced, appears from the languages of many Polynesian tribes, and from the first attempts at speaking made by children among ourselves. The passage of K into T appears in the verbal pedigree of the pronoun of the second person (fig. 11). It is evident that the T was derived from an original K in the three languages independently of one another. In the Egyptian, T was substituted in the feminine gender, the original harsher sound being retained for the masculine. In the Semitic languages, T was substituted for K in the independent pronoun, and in the preformatives and afformatives of verbs, K being retained for the possessive and objective affixes. In the Japhetic languages K has disappeared altogether, being everywhere replaced by a dental. The values of Samech were, therefore, originally SK only; from about the 13th century B.C., SK and ST, the latter gradually supplanting the other; and from the 6th century B.C., S.

Zayin expressed the sonant or softened sound corresponding to the second value of Samech, that is ZD; and from the 6th century B.C., Z.

Tsaddi expressed the strengthened sound corresponding to the above. After the 6th century B.C., it represented the Arabic Sad, and before that date, the combination of that letter with Teth.

Teth represented a strengthened TH. With the vowel a after it Teth is generally confounded, in Assyrian writing, with Daleth; and Tsaddi always with Zayin; but they are distinguished before the other vowels. Perhaps it may be fairly inferred from this, that tha was pronounced as in that, but thi and thu as in thin and thumb; always, however, with the peculiar Semitic strengthening of the consonant.

Qoph is another strengthened letter, bearing the same relation to Gimel and Kaph, that Teth bears to Daleth and Tau.

Heth must also have been a strengthened letter, being frequently substituted for Teth. It may be conjectured, but is scarcely capable of proof, that the letters to which it was related were Ghayin, pronounced as in the name of Gaza, and He.

Aleph is usually considered to denote the simple commencement of utterance, or the absence of any breathing or semivowel; and this certainly seems to have been its value in all the Assyrian inscriptions which are extant. If, however, we go back to the infancy of language, we shall see reason to think that Aleph had once a positive value. It expresses the preformative of the first person singular; sometimes alone, and at other times with a vowel, which is evidently one of connexion. As T represents the pronoun of the second person singular, and N that of the first person plural, so Aleph represents that of the first person singular; and this must have been originally something substantial, not a mere negation as at present.

The character which represents the preformative with the vowel a as a link of connexion, is that which as an ideograph denotes "water" (fig. 3). It represents other Hebrew letters as well as Aleph. In fig. 12 it has this value, that is; but in fig. 13 it represents ; in figs. 14 and 15; and in fig. 15 it also represents 1. It is probable that the Hebrew words in which it represents initial ya were originally

T

T

pronounced with wa. This seems evident in the case of "wine," and of verbs like. It would appear then that the letter Aleph passed into He and Vaw; and it is therefore natural to suppose that its original value was HW, which would easily do this. Indeed there is no sound which would be likely to be dropped that Aleph could represent except this, H, W and Y having other letters to represent them. In confirmation of this, it is to be observed that mu is the pronominal root of the first person in Accadian; and the Indo-European objective me implies the existence of a nominative mu, from which it would be derived as te from tu. The transition of hw into m is exceedingly common: see the annexed Table. The Japhetic form would then be hwu, the Semitic and Egyptian or, the vowel being unsettled. By adding ku," here," to the primeval hwu, we have kwu-ku, changed for euphony into hwáku or hwiku, "I here." This became the Indo-European iku, and with the verb of presence hwan-hwáku, properly " adsum,' ," and not used till a late period for "" ego.

T

The character for "water" was in Accadian hwa; generally used in the plural as hwa-hwa. These were Japhetic forms; and from them we have the Indo-European ahwa, and from the singular form ma, wa, wi, u, all of which are in use. The Semitic mi and mami and the Egyptian mu are clearly connected with them.

The following is a list of the verbal pedigrees, several of them imperfect, which have been obtained by me. Of the numerals with which it begins, the Accadian words for "three" and "four" are less to be depended on than the rest. It is supposed that the second character in fig. 1 represents "a set of three," three bars or other objects connected by a string; that the Accadian name of this was the phonetic value of the character ri; while the Semitic word corresponding to this, tal (whence talit in the feminine), was a second phonetic value, as in fig. 10. The analogy of the numeral for "two" favours this view. It is supposed again that a rhombus or foursided figure, of which the phonetic value was ut, represented the Accadian word for four," as four small horizontal wedges represented the corresponding Semitic word. The equivalence in value of these characters seems probable, though it has not been demonstrated. Here again the analogies of Nos. 11 and 12 corroborate the pedigree.

No. 1. ana "one,” Ac. hwana; I.E. hwan ; J. hwan.

No. 2. (See above), Ac. mi; I.E. thwi; J. thmi; E. thui; S. thni; Pr. thmi. No. 3. (See above), Ac. ri (?); I.E. tri; J. tri; S. tal; Pr. tar.

No. 4. (See above), Ac. hut (?); I.E. hwat-war "four complete" and ahwt-au "two fours, i. e. eight;" J. ahwat, hwat; E. afat; Pr. ahwat.

No. 5. iya "five;” Ac. hwihwa; I.F. hwinhwi; J. hwinhwa; E. thihw; S.hinth;

Pr. hwinhw.

No. 6. as; Ac. has; I.E. sihs; J. sahs; S. sas; Pr. sahs.

No. 7. (See above), Ac. mu; I.E. mu, mi; J. hwu; E. and S. &; Pr. hwu; I.E. iku "I here;" J. hwiku; S. hwan-hwaku It is I here;" E. JN-IN ; Pr. hwa-ku and hui-ku, by euphonic change for hwu-ku "I here."

No. 8. Ac. zdu; I.E. stu, sti; J. stu; E. and S. and Pr. sku.

No. 9. a, ha,wa "water;" Ac. hwa and hwa-hwa; I.E. ma, wa, wi, u and ahwa; J. and Pr. hwa, and in plural, hwa-hwa; E. mu; S. mi and mami. No. 10. igh Y a house;" Ac. hwigh; I.E. hwik; J. and Pr. hwih; E. pi. No. 11. us a male;" Ac. hus; I.E. mas; J. hwas. No. 12. un 66 a man;" Ac. hun; I.E. man; J. hwan; cf. No. 1; character for "one" is used as determinative of names of men.

observing that the The connexion of

the last two roots with and N is by no means impossible, though certainly not to be relied on.

No. 13. "A king," Ac. man; I.E. and J. hwan.

No. 14. "A lion," Ac. lig; I.E. and J. lihw; S. lith, lavi; E. lavo.

No. 15. "A jaw." Ac. ka; I.E. kahw.

To these might be added sku, the Accadian word for "a skin," and perhaps others. Now it must strike a person at first sight, that the above words are much better represented in the Greek, Latin, and Teutonic languages than in the Sanskrit. The

first on the list, and all those after the ninth have no representatives at all in Sanskrit; and of the eight which have Sanskrit representatives, it is only in the case of the numerals for "two" and "three" that they come anything like as near to the primitive Indo-European forms as the Classical or Teutonic representatives of these words. It thus appears that, contrary to what has been generally thought of late years, so far as the roots are concerned, the Greek, Latin and Teutonic forms approach the nearest to the primitive forms; while the Sanskrit deviates from them

most.

Even as respects the verbal terminations, where the Sanskrit is more to be depended on, it appears that Bopp has given more weight to its testimony than it deserves. He considers the ST of the second person, preserved in English and German to this day, and found also in the Latin and Armoric preterperfects, to contain an unorganic addition; whereas it appears from what has been said that m and st are the commencements of the primitive Indo-European pronouns of the first and second persons singular. It is now manifest that the original declension of the Latin present was reg-u-m, reg-i-st...reg-u-mus, reg-i-stis. The st was reduced to s in the singular and to t in the plural; just as in the separate pronoun it was reduced to s in Greek and to t in Latin.

The increased value of the Greek and Latin forms given to them by this discovery renders it more than ever important to ascertain what they were in the earliest periods. It is known that while the Greek which has come down to us is written with an alphabet of twenty-four letters, the Greeks had at first only sixteen. It is therefore a question of great interest what these sixteen letters were; and in connexion with this it has to be considered how the Greeks originally wrote words which were subsequently written with the newly introduced eight letters. That the Greeks derived their letters from the Phoenicians is generally admitted; and both the names that are given to most of them and their forms prove that this is the case. The vowels of the original alphabet were the three Semitic breathings, Aleph, He, and Ayin. Vaw and Yod retained their original values, and represented the semivowels W and Y. Of the remaining seventeen Semitic letters there were six which represented either combinations of consonants, or those strengthened consonants which were peculiar to the Semitic languages. These six were rejected by the Greeks. The remaining eleven with the three vowels and the two semivowels constituted the original alphabet. These were the first twenty letters of the full alphabet, with the exception of Z, H, O and E. An old Greek Scholiast expressly affirms that these were the sixteen original or Cadmean letters; and the fanciful theories of modern writers in England and on the Continent, in opposition to this statement, are unworthy of the slightest attention. Two different sets of four letters were added to the original sixteen in different parts of Greece. In certain districts the four excepted letters given above were added; but H was used as an aspirate, not as a long vowel. In other districts H was added as an aspirate, and along with it 8, ♣, and X as compound aspirates. Prior to the introduction of H, its place was supplied by Σ, which was originally pronounced as a sibilant in the masculine article 20 and many other words, and which afterwards retained its place in the word when an aspirate only was sounded there. This was the Ɛàv kíßdaλov of the fragment of Pindar. About the time when H was introduced as an aspirate, the semivowels lost their original values, except in the diphthongs; and were used to express modifications of the original vowels E and O. The Æolians retained the semivowel for W, as well as the new vowel which the other Greeks substituted for it; but neither this additional letter nor the Koppa ever formed part of the regular Greek alphabet. At the time when Z and were first added to the alphabet, which must have been before 1000 B.C., their values were σd and σk respectively. By combining together the letters added in different parts of Greece, an alphabet of twenty-two letters was formed. The values of Z and having been changed from those just stated to do and kʊ, Simonides added a twenty-third letter, wo; and he then completed the alphabet by adding the long at the end; changing at the same time the value of H from an aspirate to a long vowel.

It appears from what has been said that in the earliest Greek writing the vowels E and O were used to express what were afterwards divided into two vowel sounds; the former comprehending E and I, the latter O and Y. In like manner, the conso

nants K, II and T were used to express not only the sounds that were afterwards expressed by those letters, but also those that were afterwards expressed by X, and . It is important to attend to this; as the mode of spelling which prevailed before this distinction was made, was for the most part retained in roots and even in suffixes of derivation. Thus Téνте, TÉμжTоs might be intended to express not only the words which those letters would represent when the alphabet was completed, but also divde, Piμpdos, and so in other instances. Analogy leads us to think that if the twenty-four letters had been all in use when the language was first written down, these words would have been written in the last-mentioned mode.

In these instances the ancient spelling has been preserved, but the pronunciation which it suggests has been changed. In other instances the spelling has been altered; and there is particular need to attend to that change of spelling which was occasioned by the dropping of the semivowels and the change of & to H, and this again to a sign of aspiration which was dropped in the middle of a word, and not always expressed at the beginning of one. The examples of d,, originally 20, EIA, then HO, HIA, or HA, and of ȧvdávw, originally ƐYANAANO, will sufficiently illustrate what is meant.

On the supposed Biblical Names of Baalbec, and on the position of Baalgad. By JOHN HOGG, M.A., F.R.S., L.S., R.G.S. &c., Foreign Secretary of the Royal Society of Literature.

The author began this communication with some remarks on the vast interest which has recently become attached to the geography, both historical and physical, of the Bible; especially since the important geographical and topographical excavations and discoveries in ancient Assyria, and other parts of the East, have thrown fresh light upon many portions of Scripture. Also, as regards the cartography of Syria and Palestine, much had been done of late years by Russegger, Robinson, Porter, Van de Velde, and the officers of the American Survey. The author said, that notwithstanding these great additions to science, he felt satisfied that much still remained uncertain respecting the natural positions of many of the valleys and rivercourses in both Lebanons, a more correct knowledge of the heights of the chief mountains, the magnetic influences, and certain meteorological phenomena.

Considering that Baalbec had been so long known to travellers and artists, and its magnificent ruins so often visited and described, it was a remarkable fact that its Biblical name remained at this day undetermined. And it was further remarkable, that history has not preserved the name of its founder, or recorded in what year of the world it was begun to be built. As a Roman colony at the commencement of the Christian era, under the name of "Heliopolis," it bore the title of " Julia Augusta," derived from its benefactors, Julius and Augustus Cæsars.

The author then showed that it could not answer in geographical details to Baalath, or Baal-Hamon, or Balamo, or Baal Hermon, or Baalah, or Baalgad, of the Scriptures. Dr. Robinson of America imagines with much probability, that the prophet Amos, in chap. i. v. 5, alluded to Baalbec (Heliopolis) and its idol-worship, where he mentions "the plain of Aven," or as it is given in the margin of the Bible, "Bikath Aven." This would seem to refer to the plain of Colesyria, or the Bukaa, meaning a "valley," which lies between the Lebanons, or the range of Libanus, and that of Anti-Libanus. Mr. Hogg observed that this allusion' had long ago occurred to the traveller Maundrell, and to Dr. Wm. Lowth. The Septuagint translators having rendered the Hebrew word 'Aven' by On-the Heliopolis of Egypt-thought the same. As this word 'ON is supposed to mean in Egyptian the sun, so the Septuagint rightly translated it Heliopolis; consequently Bikath Aven would signify the Bukaa On, or plain of On;' that is to say, the plain or valley of Heliopolis.

The author, in this supposition, agreed with those writers, and he further remarked that the word Baalbec has the same meaning in Aramaic or ancient Syriac, viz. the "Sun-city."

In the next place, as to the geographical position of Baalgad, Mr. Hogg, having demonstrated from the descriptions contained in Scripture, that it could not coincide with the site of Baalbec, as has been strongly urged by many authors, considered

that Dr. Robinson's supposition of Baalgad having been "no other than the secluded grotto," Panium near Paneas, now called Banias, was quite untenable. He, on the contrary, conceived that the ancient and long-lost Baalgad must have stood either in Wadi E' Teim, or in a neighbouring smaller valley of Lebanon, on the west of, and under Mount Hermon, named at this day the Mountain of the Chief,' or Gebel E' Sheikh. Indeed, he said that a hope might be entertained that among the numerous ancient ruins and temples still remaining in the Wadis near the Lebanon, some vestiges of Baalgad might yet be discovered, when more careful explorations shall have been made, in the secluded western gorges adjacent to the noble Hermon.

Mr. Hogg pointed out the supposed position of Baalgad, as it accorded with the accounts of Holy Writ, on a sketch map which he had coloured and drawn on a scale eight-times enlarged, from a part of Kiepert's "New Map of Palestine and Colesyria," 1856, extending from nearly 33° 14' to 34° north latitude, and from about 35° 35' to 36° 11′ east longitude; and he also explained from it the sites of other Biblical localities, as well as the present courses of the principal rivers in that portion of Syria.

On the Cause of the Mild Winter Temperature of the British Islands.

By T. HOPKINS.

In this paper the writer combated the opinion that the mild winter temperature of the north-western part of Europe is attributable to the warm ocean stream that flows from the Gulf of Mexico into the Atlantic Ocean. He denied that the mere contiguity of a warm current makes the adjoining country warm. In support of this, it was shown that the parts of America close to the Gulf and the warm stream, are not warmed by them. On the contrary, Louisiana and Florida, notwithstanding their vicinity to the warm water, were shown to have cold winter climates; and this character attaches to the sea-board of America that extends parallel with the Gulfstream.

It was shown that Dove's Isothermal line for the month of January, of 50° Fahr., passed over the 45° of latitude in the Pacific Ocean, but that it descended to the 32nd degree near the Gulf of Mexico, and when crossing the warm Florida stream the line rose but little; yet as it passed to the middle of the Atlantic it ascended to the same latitude that it passed over in the Pacific, thus showing that it was as cold over the warm stream in a southern as in the middle of the Atlantic in a northern latitude. Extracts were given from Dr. Scoresby's paper on the temperature of the northern Atlantic, in which it was shown that the cold current which flows from Baffin's Bay, and passes the island of Newfoundland, intervenes between the Gulf-stream and the ocean to the north of 40° of latitude, stopping its progress east in that latitude. These facts, it was contended, showed that the warm water of the Gulfstream flowed, as described by Humboldt, towards the Azores, and therefore could not warm the north-west parts of Europe.

It was stated also that an ocean current runs from 7° of south latitude along the Brazilian and eastern Patagonian coasts, but that it no more warms those countries than the Gulf-stream does the adjacent parts of America, all these countries being dry and cool in the winter season. It was contended that the mild winter climate of the British Islands was due to copious condensation of vapour brought from the surface of the Atlantic Ocean. This vapour is condensed freely in the winter and warms the atmosphere, the vapour giving out its heat of elasticity to the air, this warmth extending to other parts in proportion to the amount of vapour condensed. The Isle of Skye in Scotland is warm in the winter, and it is recorded that more rain had fallen in that island in the single month of January than falls in Paris or London in a year.

Instances were given of sudden changes of winter weather from cold to warm, in Manchester and Paris, under circumstances which proved that much heat was brought to those parts in vapour. Other parts of the world, distant from the tropics, were also pointed out, which were rendered warm in their winters by condensation of vapour, and not through the contiguity of a warm ocean. The north-west coast of America, up to the latitude of 60°, has as warm a winter climate as the western coast of Europe in the same latitudes. And Western Patagonia and Cape