we have little information, except general statements as to their irregularity, and the evidence of this from fac-similes. There is an exception, however, in the case of Codex F of the Gospels, of which J. Heringa's careful collation has been published by H. E. Vinke: Disputatio de Codice Boreeliano, nunc Rheno-Trajectino, etc., Traj. ad Rhen. 1843, 4to. This MS. reads έστηκεν, John i. 26; ἐστήκατε, Matt. xx. 6; έστηκασιν, Matt. xii. 47; ECTS, John vi. 22, and tørks, John iii. 29; EOTηKÓT, Mark xi. 5; iors, John xii. 29; iaròs, Matt. xxiv. 15, Mark xiii. 14; ¿OTWTES, -Tas, -Twv, Matt. xx. 3, 6, xxvi. 73, xxvii. 47; OTEL, Matt. xiii. 2, John i. 35; and iornкεGav, Matt. xii. 46. Thus, of the twenty-seven examples which it contains of the perfect and pluperfect forms of ion, the collator has expressly noted seventeen in which they have the smooth breathing.

· ουχέστηκεν.

Another uncial MS. of the Gospels, Codex H at Hamburg, has been recently examined with reference to this matter by Dr. C. R. Gregory, of Leipzig, at the request of a member of our Committee, with a result still more striking. Of the thirty-one examples which it contains of the perfect and pluperfect forms of ior, twenty-eight have the smooth breathing; two, the rough; and one is doubtful, having been altered. from one to the other. The reading in the present passage is peculiar, Dr. Gregory has also made notes on the cursive MS. 234 of the Gospels (Acts 57, Paul 72) at Copenhagen. In the perfect and pluperfect forms of iorquit has the smooth breathing ten times, and the original smooth has been altered to the rough six (or perhaps seven) times. besides. It may be worth while to add that in the fac-simile of Codex 33 of the Gospels, given by Scrivener in his Introduction, 2d ed., Plate xii. No. 34 (3d ed., Plate xiii. No. 39), the participle or has the smooth breathing. So in the fac-simile of Evan. 348, dated A.D. 1023, published by the Palæogr. Soc. Part ix. (1879), pl. 130.

As to the reading οὐκ ἔστηκεν [sic] for οὐχ έστηκεν in one place in some editions of Origen and of Didymus (see above), and in Cyril of Alexandria (iv. 563, ed. Aubert), it may be a mere misprint; it is so treated by the later editors of Origen (Lommatzsch, ii. 264), Didymus (Migne, Patr. Gr. xxxix.

1105), and Cyril (Pusey, ii. 100). If, on the other hand, it was derived from a MS., as may be the case, though it is nowhere so stated, we can only infer that the scribe pronounced the perfect ForKev without the aspirate.

In view of the facts which have been presented, it appears that the evidence for an imperfect orKev in this passage, though it may at first seem strong, breaks down at every point; and till some proof of the actual use of an imperfect of or shall be produced, we must regard its very existence as imaginary.

XV.

ON THE READING "CHURCH OF GOD,"

ACTS XX. 28.

[From the Bibliotheca Sacra for April, 1876.*]

Common Version: "Take heed therefore unto yourselves, and to all the flock, over the which the Holy Ghost hath made you overseers, to feed the church of God, which he hath purchased with his own blood." Received Text: Προσέχετε οὖν ἑαυτοῖς καὶ παντὶ τῷ ποιμνίῳ, ἐν ᾧ ὑμᾶς τὸ πνεῦμα τὸ ἅγιον ἔθετο ἐπισκόπους, ποιμαίνειν τὴν ἐκκλησίαν τοῦ θεοῦ, ἣν περιεποιήσατο διὰ τοῦ ἰδίου αἵματος. Various readings: or, "therefore," is bracketed by Lachmann, and omitted by Tischendorf, Tregelles, Green (Twofold New Test.), and Westcott and Hort, but is retained by Alford and Wordsworth. For Tov bεov, "God," Lachmann, Tischendorf, Tregelles, and Green read To Kupiov, "the Lord"; Alford, Wordsworth, and Westcott and Hort retain ɛov. But Tregelles places ɛov in the margin with a mark of interrogation, implying some doubt whether it should not be regarded as an alternative reading; and Alford, on the other hand, puts upon in the margin, in large type, as of nearly equal authority with tɛov. All the editions named above read in the last clause διὰ τοῦ αἵματος τοῦ ἰδίου for διὰ τοῦ ἰδίου αἵματος.

Of those who have written treatises on the textual criticism of the New Testament, Porter, Davidson, and Hammond give the preference to Kupiov; Scrivener and Milligan defend BEO, Among recent commentators and translators, do is preferred by Dr. Gloag; on the other hand, Meyer, Ewald, Lechler (in Lange's Bibelwerk) very confidently, Overbeck, Dr. David Brown (with hesitation), Holtzmann (in Bunsen's Bibelwerk), the new Dutch translation (1868), and Weizsäcker adopt the reading κυρίου,

* [The substance of this article was originally prepared at the request of the New Testament Company of the American Biblical Revision Committee.]

To recount the opinions of the earlier critics, or to give a sketch of the literature of the subject, would carry us too far. But as a mistake made by one scholar often leads many astray, it may be well to say that Matthaei does not read fro, as stated by De Wette, Davidson, and Alford, but kupiov kaì bɛoì, in both of his editions; that Gratz does not reject Kupiov, as is affirmed by Bloomfield (9th ed.), but adopts it; and that although Michaelis defends tov in his Introduction to the New Testament (4th ed., 1788), in a later work (Anmerkungen zu seiner Uebers. d. N. T., 1790, ii. 407 ff.) he gives the preference to Kupiov as the best supported reading.

The passage presents one of the most interesting and important problems in the textual criticism of the New Testament; but no thorough investigation of the evidence for the different readings has been published, so far as I am aware, since the time of Wetstein. The recent accession of the Sinaitic MS. to the authorities for fou may be thought by some to turn the scale in its favor; and the fact that this reading is received into the text by scholars so eminent as Professor Westcott and Dr. Hort might alone justify a new discussion of the question, if any excuse were needed.

In stating the evidence for the different readings, we may begin with

81

95

I. THE AUTHORITIES FOR KVрiov.

13

15

18 36 40 69 73

58

MANUSCRIPTS. —A, C*, D, E, I' X or XI' XIII' XIII' XI' XIII' XI' 130 156 163 180 acr B-C. 7, and Lect. ; in XI' XI or XII' XII' XI' XIV' XII' XII' XIII XII all, four uncials and sixteen cursives.* As to date, two are supposed to be of the fifth century, two of the sixth, one of the tenth or eleventh, five of the eleventh, one of the eleventh or twelfth, four of the twelfth, four of the thirteenth, and one

*I omit Tischendorf's "cat", by which he means not "some catenae," but the text of the MS. (New Coll. Oxford, 58), published by Cramer with its catena, in 1838. Tischendorf sometimes cites this as "cat", sometimes as catox", but does not seem to be aware that it is identical with No. 36. Bloomfield (Crit. Annot., Lond 1860, p. 194) says, "I am now, indeed, enabled to add to the evidence for kupio, 9 Lam. and Scr. MSS." But krop here must be a mistake for κυρίου καὶ θεοῦ, B-C. II. 7 is one of the Burdett-Coutts MSS. recently collated by Scrivener; see his Introduction, 2d ed., pp. 221, 540 [ 3d ed., p. 236, MS. 549; it is now numbered 219 of the Acts; aer in the list above is now known as 182 of the Acts].

of the fourteenth. Here the high character of the cursives which read Kupiov is particularly to be remarked. Eight of them, Nos. 13, 36, 40, 69, 73, 81, 95, and 180, are marked by Tischendorf with an asterisk in the Prolegomena to his seventh critical edition as noticeable for their agreement with the text of the most ancient copies; and there are three others at least, namely, Nos. 15, 18, and aser, which deserve to be so marked. The first in the list, No. 13 (33 Gosp., 17 Pauline Epist.), is said by Eichhorn to be "full of the most excellent and oldest readings." He styles it "the Queen of the cursive manuscripts." No. 40 Tischendorf designates as "codex admodum insignis;" it represents the text of Euthalius. No. 73 is called by Griesbach "praestantissimus;" "optimis adnumerare non dubito," says Birch (Variae Lect. 1798, p. ix.). No. 180 is justly spoken of by Scrivener as "important." Finally, Scrivener's "a" represents, according to him, "a very interesting and valuable text, . . . being found in harmony. . . with the most ancient MSS., and very conspicuously with that most precious document designated . . . as p" (now 61, formerly Tischendorf's "lo"). (Introd. to Cod. Augiensis, p. lvi.) The excellence of most of the cursives that support Kupi, in contrast with the inferior character of those which read eo, is an important point, and will be illustrated hereafter.

ANCIENT VERSIONS. - The Old Latin (second century), as shown by the quotations in all the earlier Latin Fathers (see below), confirmed more or less by the Latin interpreter of Irenaeus, and the Graeco-Latin MSS. D and E;* the Memphitic or Coptic (third century, or perhaps the second), the Thebaic or Sahidic (same date), the Armenian (fifth century), and the Harclean or Philoxenian Syriac (A.D. 616) in the margin, representing an Alexandrian MS. "very accurate and approved," according to Thomas of Harkel, and which. certainly exhibits an early form of the text, though, like D, disfigured by interpolations.

* Domini is also the reading of the Gigas Librorum, published by Belsheim, Christiania, 1879, the only MS. of the Old Latin containing the Acts complete.