complicated entities as intelligence, personality, and character, since we have as yet only a very inadequate knowledge of humanity, a good start has been made, and the present instruments, imperfect as they are, mark a long advance over the old stage of preconceptions, prejudices, and unsupported opinion, through which research formerly groped in judging and comparing the achievements of school systems. Research has also devised techniques whereby educational situations may be brought under control while striving to measure what takes place under different conditions.

Investigators are constantly learning better how to set up control groups and hold the conditions constant while applying the variable. They are also gradually acquiring the necessary skill in analysis to determine the value of various items to be utilized in comparisons of efficiency, methods of administration or finance, and other like problems. A host of research men, emancipated from all administrative and even class duties, are springing up; and, with the growth of a corps of professional assistants, statistical aids, and technical apparatus, prompt and reliable results can be had. Pure research will soon become more common and general laws will be evolved, while, on the other hand, practical experimentation will no longer be regarded as a dangerous proceeding either for schools or pupils.

1

If up to the present research in education has not reached the precision of physical and biological sciences, it has at least passed the stage of ridicule and should meet with hearty encouragement from scientists. Its efforts to develop standards and formulate procedures indicate that it is now making a successful struggle toward scientific standing. Yet it has as yet been granted little or no general recognition. Amongst the thousand most noted men of science in Cattell's inventory, not a single one has been awarded his place because

1 For their achievements, methods, and bibliography, see Monroe, Walter S., and Engelhart, Max D.: "The Techniques of Educational Research" (University of Illinois, 1928).

2 See Cattell, J. McKeen (Editor): "American Men of Science," Fourth Edition (The Science Press, New York, 1927).

of achievements in educational investigation, and, while the American Educational Research Association is permitted to send a representative to the National Research Council, he is admitted only as a member of the section in Psychology or in Anthropology.

This situation must change. There is a crying need for research in education. An interest which involves the expenditure of more than two billion dollars each year, and, what is of far greater importance, concerns the welfare and future of thirty million boys and girls, should not be subject to the hazards of guess-work and waste. Where the demand for school efficiency and educational results is so great, the means for satisfying it must soon be found and due credit be given those who carefully conduct the search. The day cannot be far distant when the work of men like Ayres and Buckingham, Courtis and Charters, Terman and Thorndike, will bear its legitimate fruit. The influence of hundreds of young men who are each year turning their attention for a time at least to research in education is bound to be cumulative and will eventually be felt.

TEXTUAL CRITICISM OF THE GREEK OLD TESTAMENT

1

By MAX L. MARGOLIS

(Read April 19, 1928)

"THE more frequently a text has been copied," says Otto Stählin, "the greater will have been the influence exerted by the copies one upon the other, resulting in cross relationships and the repeated rise of new mixed texts. Hence given a large number of manuscripts it is often impossible to find any one which is a direct transcript of a 'copy' (Vorlage) still extant. Such is the case for instance with the numerous manuscripts of the Septuagint, among which, in the judgment of P. de Lagarde, no single one is entirely without value, but at the same time none is quite free from grave corruptions. Under such conditions it is impossible to establish a stemma (family tree, pedigree); it is quite feasible, however, to assemble manuscripts within groups of common descent by reason of identical corruptions (in this connection additions. and omissions are particularly significant)."

2. A glance at the farrago of readings in the laboriously compiled apparatus, whether in the Oxford (1798-1827, Holmes-Parsons) or in the Larger Cambridge Septuagint (1906-, Brooke-McLean-Thackeray), will reveal, on the one hand, readings which are singular, i.e., found among the witnesses examined by the editors just in one to the exclusion of all others, and, on the other hand, readings which are held in common by two or more witnesses. Singular readings answer to one of two tendencies marking the course of textual transmission. It is the tendency to alteration. In the singular readings are revealed the habits and idiosyncrasies of scribes, their physical and mental failings, and even their moral delinquencies, as when in sheer haste they pass over portions of the text or wilfully indulge in contraction. Gen

1 Editionstechnik, 1909, 17.

erally speaking, singular readings are so much refuse which must needs be swept up and constitute the subtrahend in the operation thrust upon the student who is always far more interested in the antecedent copy, now lost, than in the extant transcript.

3. For the sake of illustration, let us take in the Book of Joshua 2 codex p (Petrograd Gr. 62 and the concluding portion -24. 26 to the end-in the Bodleian Auct. T. inf. 2. 1, designated as a2 in Brooke-McLean). It has 380 singular readings. The scribe3 slipped up in the uncial script which on occasion he reproduced inaccurately. The greater number of graphic errors consists in the confusion of AAA, ECOO, гI, HN. Thus 19.38 AAA is transformed into AAA, μaydadınd into μαγαδαιηλ. 15. 21 καβσενα for καβσεηλ with A for A and N for H. The worst example is 19. 13 ET тоаукаσεμ; аррarently the antecedent copy read επι πολη κασειμ—A for A requires no comment, but (sign of abbreviation for 4) was misread as г.5 When the scribe found in his copy a d originating in @ (mispronunciation !), he replaced it by λ as 15. 58 βελ < βεδ < βεθ 19.39 νεφλαλειμ < νεφδ. < νεφθ. ; the reverse 15.58 αουθ < αουδ < αουλ. e for C: II. 22 αενδωθ for ασηδωθ (at the same time with N for H). o for e: 15. 6 Bol. 。 for 0: 15. 27 βαιο. o for σ: 15. 3 καδηο. γεσσουρ (γεσουρ); conversely 9. 17 15. 37 ιαδ. II for T: 23. 13 Eπl. axep (at the same time x and

0 for σ: 13. 13 yelσoup for βηρως for βηρωθ. I for г: p in the place of B: 19. 29 change places). M for A

ρ

(also liable to confusion in the cursive script): 21. 18 appr for aλuwv; N for A (as above): 17. 3 μaava. M for N (graphically similar in both scripts, but also subject to indistinct pronunciation): 19. 19 αμαρεθ.

Cursive characters confused: u up. u (ß) for u (k): 15. 13

2 A critical edition of the Greek text, with an apparatus containing the divergences in four principal recensions and their attestation within each as well as the variations in the sub-recensions and individual group members, has been prepared by the writer and is now in course of publication.

3 The errors need not all have been his own. Some he may have taken over from copies no longer extant (or at present uncollated). In the end they all represent aberrations by individual scribes.

Thompson, "Handbook of Greek and Latin Palæography," 94. 5 Comp. similarly 15. 3 ακραβογ for ακραβειν.