tained. These families or recensions are three or four in number,

viz.

1. The Spanish manuscripts, which were corrected after the Codex of Hillel. They follow the Masoretic system with great accuracy, and are on this account highly valued by the Jews, though some Hebrew critics hold them in little estimation. The characters are written with great elegance, and are perfectly square: the ink is pale; the pages are seldom divided into three columns; the Psalms are divided into hemistichs; and the Chaldee paraphrases are not interlined, but written in separate columns, or are inserted in the margin in smaller letters. Professor Tychsen speaks in high terms of the calligraphy of the Spanish manuscripts. As the Spanish monks excelled in that art, he thinks the Jews, who abounded in Spain in the twelfth and thirteenth centuries, acquired it from them, and he appeals to manuscripts which he had seen, where the letters are throughout so equal, that the whole has the appearance of print.1

2. The Oriental manuscripts are nearly the same as the Spanish manuscripts, and may be referred to the same class.

3. The German manuscripts are written with less elegance than the Spanish codices: their characters are more rudely formed; the initial letters are generally larger than the rest, and ornamented; the ink is very black. They do not follow the Masoretic notation, and frequently vary from the Masoretic manuscripts, exhibiting important readings that are not to be found in the Spanish manuscripts, but which agree with the Samaritan text of the Pentateuch, and with the antient versions. The Chaldee paraphrases are inserted in alternate verses. This class of manuscripts is little esteemed by the Jews, but most highly valued by biblical critics.

4. The Italian manuscripts hold a middle place between the Spanish and German codices, and sometimes have a nearer affinity to one class than to the other, both in the shape of the Hebrew characters, and also as it respects their adherence to or neglect of the Masoretic system. M. Bruns, the able assistant of Dr. Kennicott in collating Hebrew manuscripts, has given engraved specimens of the Spanish, German, and Italian manuscripts, in his edition of Dr. K.'s Dissertatio Generalis (8vo. Brunswick, 1783); and Professor Tychsen has given fourteen Hebrew alphabets, of various ages and countries, at the end of his Tentamen de variis Codicum Hebræorum Vet. Test. MSS. Generibus. Antient and unpointed Hebrew manuscripts, written for the use of the synagogues, and those Masoretic Spanish exemplars, which have been transcribed by a learned person, and for a learned person, from some famous and correct copy, are preferred by M. De Rossi to the copies written for private use, or even for the synagogue, from Masoretic exemplars, of which last the number is very great. But M. Bauer pronounces those manuscripts to be the best, whose various lections are most frequently confirmed by the an

I Tychsen, Tentamen de variis Cod. Heb. MSS. pp. 302-308.

tient versions, especially by the Alexandrian and Syriac, and also by the Samaritan Pentateuch and version.1

VIII. M. De Rossi has divided Hebrew manuscripts into three classes, viz. 1. More antient, or those written before the twelfth century; -2. Antient, or those written in the thirteenth and fourteenth centuries;-3. More recent, or those written at the end of the fourteenth, or at the beginning of the fifteenth century. The most recent, or those written since the fifteenth century, which are very numerous, and are those found in the synagogues, he pronounces to be of little or no use, unless it can be proved that they have been transcribed from antient apographs. The total number of Hebrew manuscripts collated by Dr. Kennicott for his critical edition of the Hebrew Bible (of which an account is given in a subsequent page), is about six hundred and thirty. The total number collated by M. De Rossi for his Collection of Various Readings (also noticed in a subsequent page), is four hundred and seventy-nine manuscripts, besides two hundred and eighty-eight printed editions. The following are the most antient manuscripts collated by Dr. Kennicott.

1. The CODEX LAUDIANUS A. 172 and 162, and numbered 1. in Dr. Kennicott's list of Hebrew manuscripts. Though now in two folio parts, it is evident that they originally formed only one volume: each part consists of quinquernions, or gatherings of five sheets or ten leaves, and at the bottom of every tenth leaf is a catch-word beginning the next leaf, which is the first of the succeeding gathering of ten leaves. But at the end of the first part or volume, there is pasted on, one leaf of the next quinquernion, completing the book of Deuteronomy; so that this volume concludes with five sheets and one leaf over. And the first gathering in the second volume consists of only four sheets and one leaf, which last is likewise pasted on, for want of its fellow-leaf. This manuscript is written on vellum, according to Dr. Kennicott, in the Spanish character, but in the opinion of Dr. Bruns it is in the Italic character, to which M. de Rossi assents. The letters, which are moderately large, are plain, simple, and elegant, but universally unadorned; and they were originally written without points, as is evident from the different colour of the ink in the letters and in the points. Some of the letters, having become obliterated by the lapse of ages, have been written over a second time; and though such places were re-written in the same strong character, yet many of the words were becoming a second time invisible, when collated by Dr. K. This eminent critic assigns it to the tenth century, but De Rossi refers it to the eleventh. The Laudian manuscript begins with Gen. xxvii. 31.: it contains fourteen thousand variations from Vander Hooght's edition of the Hebrew Bible. More than two thousand are found in the Pentateuch, which confirm the Septuagint Greek version in one hundred and nine various readings; the Syriac, in ninety-eight; the

1 Walton, Prolegom. c. iv. § 1-12. pp. 171-184. cc. vii. viii. pp. 225-331. edit. Dathii. Carpzov. Critica Sacra, pp. 283-387. Dr. Kennicott, diss. i. pp. 313317.; also his Dissertatio Generalis, passim. Jahn, Introd. ad Vet. Fœdus, pp. 153 -170. Bauer, Critica Sacra, pp. 215-226. 343-407. De Rossi. Var. Lect. tom. i. Prolegom. § xi.—xix. pp. x1.—XXII.

6

Arabic, in eighty-two; the Vulgate or Latin Version, in eighty-eight; and the Chaldee Paraphrase, in forty-two: it also agrees with the Samaritan Pentateuch, against the printed Hebrew, in seven hundred instances. What renders this manuscript the more valuable is, that it preserves a word of great importance for understanding 2 Sam. xxiii. 3-7., which word is confirmed by the Greek Version, and thus recovers to us a prophecy of the Messiah.1

2. The CODEX CARLSRUHENSIS 1, (No. 154 of Dr. Kennicott's list of manuscripts,) formerly belonged to the celebrated and learned Reuchlin, whose efforts contributed so much towards the revival of literature in the fifteenth century. This manuscript is now preserved in the public library at Carlsruhe, and is the oldest that has a certain date. It is in square folio, and was written in the year of the world 4866, corresponding with 1106 of our æra. It contains the Prophets with the Targum.

3. The CODEX VIENNE (No. 590 of Kennicott) contains the Prophet's and Hagiographa. It is written on vellum in folio, and if the date in its subscription be correct, (A. D. 1018 or 1019) it is more antient than the preceding. Bruns collected two hundred important various readings from this manuscript. The points have been added by a later hand. According to Adler's enumeration, it consists of four hundred and seventy-one leaves, and two columns, each column containing twenty-one lines.

4. The CODEX CESENE, in the Malatesta Library at Bologna, (No. 536 of Kennicott,) is a folio manuscript written on vellum, in the German character, towards the end of the eleventh century. It contains the Pentateuch, the Haphtaroth or sections of the Prophetical Books, and the Megilloth or five Books of Canticles, or the Song of Solomon, Ruth, the Lamentations of Jeremiah, Ecclesiastes, and Esther. De Rossi pronounces it to be a most antient and valuable manuscript, and states that in its margin are inserted some various readings of still more antient manuscripts.2

5. The CODEX Florentinus 2, (No. 162 of Kennicott,) is written on vellum, in quarto, in a square Spanish character, with points, towards the end of the eleventh, or at the latest, in the beginning of the twelfth century. It contains the books of Joshua, Judges, and Samuel. Very many of the letters, which were obliterated by time, have been renewed by a later hand.

6. The CODEX MEDIOLANENSIS 9, (193 of Kennicott,) is written on vellum, in octavo, in the German character, towards the close of the twelfth century. It has neither the points nor the Masora. This manuscript comprises the Pentateuch; the beginning of the book of Genesis, and the end of Leviticus and Deuteronomy, have been written by a later hand. Both erasures and alterations occur in this manuscript; and sometimes a worse reading is substituted in place of one that is preferable. Nevertheless it contains many good various readings.

7. The CODEX NORIMBERGENSIS 4, (201 of Kennicott,) is a folio manuscript, written on thin vellum, in the German character, and

1Kennicott, Dissert. I. pp. 315–319. Dissert. II. pp.533, 534. Biblia Hebraica, tom. ii. Dissert. Generalis, pp. 70, 71. De Rossi, Varie Lectiones, tom. i. Proleg. p. LIX. 2 De Rossi, tom. i. Proleg. p. LXXXVII.

containing the Prophets and Hagiographa. It is mutilated in various parts. It is of great antiquity, and from the similarity of its character to that of the Codex Carlsruhensis, both Dr. Kennicott and M. de Rossi assign it to the beginning of the twelfth century.

8. The CODEX PARISIENSIS 27, (Regius 29, 210 of Kennicott,) is a quarto manuscript of the entire Bible, written on vellum, in an elegant Italic character. The initial words are, with few exceptions, of the same size as the rest. The Masora and Keri are both wanting; and the Megilloth precede the books of Chronicles. It is highly valued by Kennicott and De Rossi, who refer it also to the beginning of the twelfth century.

9. Coeval with the preceding is the CODEX REGIOMONTANUS 2, (224 of Kennicott,) written in the Italic character, in small folio. This manuscript contains the Prophets and the Hagiographa, but it is mutilated in various places. The initial letters are larger than the others, and three of the poetical books are written in hemistichs.

10. To the beginning of the twelfth century likewise is to be referred the CODEX PARISIENSIS 84, (San-Germanensis 2, No. 366 of Kennicott): it is written on vellum, in large quarto. It is imperfect from Jer. xxix. 19. to xxxviii. 2.; and from Hosea iv. 4. to Amos vi. 12. Isaiah follows Ezekiel according to the Talmudical Canon.1 The following are among the most antient of the manuscripts in the possession of the late M. De Rossi, and collated by him, viz.

1. The Codex, by him numbered 634, which is in quarto. It contains a fragment of the books of Leviticus and Numbers, from Levit. xxi. 19. to Numb. i. 50.; and exhibits every mark of the remotest antiquity. The vellum on which it is written is decayed by age; the character is intermediate, or Italic, - approaching to that of the German manuscripts. The letters are all of an uniform size; there is no trace of the Masora, or of any Masoretic notes, nor is any space left before the larger sections; though sometimes, as in other very antient manuscripts, a few points are inserted between the words. M. De Rossi assigns this manuscript to the eighth century.

2. A manuscript of the Pentateuch (No. 503), in quarto and on vellum, containing from Gen. xii. 41. to Deut. xv. 12. It is com posed of leaves of various ages, the most antient of which are the ninth or tenth century. The character is semi-rabbinical, rude, and confessedly very antient. Points occur, in some of the more antient leaves, in the writing of the original copyist, but sometimes they are wanting. There are no traces of the Masora or of the Masoretic notes, and sometimes no space at all before the larger sections. It frequently agrees with the Samaritan text and antient versions.

3. A manuscript of the Pentateuch (No. 10), with the Targum and Megilloth. It is written in the German character, on vellum and in quarto, towards the end of the eleventh or in the beginning of the twelfth century. The Masora is absent. The character, which is defaced by time, is rudely formed, and the initial letters are larger than the rest. Coeval with this manuscript is,

4. A manuscript of the book of Job, in quarto, also on vellum, and in the German character. It is one of the most valuable ma

1 Kennicott, Dissortatio Generalis, pp. 85, 87, 88, 89.99. 104.

nuscripts of that book. The pages are divided into two columns, the lines being of unequal length.

5. A manuscript of the Hagiographa (No. 379), the size, character, and date of which correspond with the preceding. It begins with Psal. xlix. 15. and ends with Neh. xl. 4. The Masora and Keri are absent; and the poetical books are divided into hemistichs.

6. A manuscript of the Pentateuch, (No. 611), on vellum, in octavo, and written in the German character, approaching somewhat to the Spanish, towards the close of the eleventh or in the commencement of the twelfth century. The ink is frequently faded by age; there are no traces of the Masora; the Keri are very rarely to be seen, and the initial letters are larger than the others. There are frequent omissions in the text, which are supplied in the margin.1 Dr. Kennicott states that almost all the Hebrew manuscripts of the Old Testament, at present known to be extant, were written between the years 1000 and 1457, whence he infers that all the manuscripts written before the years 700 or 800 were destroyed by some decree of the Jewish senate, on account of their many differences from the copies then declared genuine. This circumstance is also alleged by Bishop Walton as the reason why we have so few examplers of the age of 600 years, and why even the copies of 700 or 800 years are

very rare.

IX. It was long a desideratum with biblical scholars to obtain the Hebrew Scriptures from the Jews who are settled in India and other parts of the East. It was reasonably supposed, that, as these Jews had been for so many ages separated from their brethren in the west, their manuscripts might contain a text derived from the autographs of the sacred writers, by a channel independent of that through which the texts of our printed Bibles has been transmitted to us. Dr. Kennicott was very anxious to obtain a copy, or at least a collation of a manuscript from India or China, for his edition of the Hebrew Bible, in the expectation that it would exhibit important variations from the Masoretic editions; but he was unsuccessful in his endeavours to procure it, and the honour of first bringing an Indian manuscript of the Hebrew Scriptures into Europe was reserved for the late Rev. Dr. Buchanan.

Among the biblical manuscripts brought from India by this learned and pious divine, and which are now deposited in the public library at Cambridge, there is a roll of the Pentateuch, which he procured

1 De Rossi, Var. Lect. tom. i. Proleg. pp. cxvI. CXII. XCVIII. CVII. CVIII.

2 According to the information collected from various sources, by Professor Bauer, it does not appear that the manuscripts of the Chinese Jews are of any remote antiquity, or are calculated to afford any assistance to biblical critics. Although Jews have resided in China for many centuries, yet they have no antient manuscripts, those now in use being subsequent to the fifteenth century. Critica Sacra, pp. 405-407. See an account of the Hebræo-Chinese manuscripts in Koegler's Notitia S. S. Bibliorum Judæorum in Imperio Sinensi. Edit. 2. 8vo. Hale ad Salam, 1805. Brotier, in his edition of Tacitus, (vol. iii. pp. 567, et seq.) has given the best account that is extant of the Jews in China, a colony of whom settled in that country in the first century of the Christian æra. The reader will find an abridgment of it in Mr. Townley's Illustrations of Biblical Literature, vol. i.

89.

pp. 83