(2) Secondary witnesses: containing textual evi- dence of a fragmentary or uncertain nature, like that of the Sibylline Oracles. The parallels in these, as in most of the writings here named, will be found in J. Rendel Harris’ Teaching of the Twelve Apostles.—
Ignatius, Smyr. xiil. 1, Magn. iv.—Did. xvi. 2. Preaching and Apoc. of Peter (ed. M. R. James, 82). Hermas (Mand. ii. etc.)—our Didache. 2 Clement—apparently our Didache (xiii. 4, xvi. 4, Did-si.23) Dye xvil, of) Didsexvigt) eeAipol moja Aristides, ch. xv. Justin—our Didache (Taylor, Hapositor, i. vi. 359 ff.). Theophilus, ad Autol. ii, 34 fin. Ireneus—Did. i. 1, 5, xiv. 3; see below. Tertullian, adv. Marc. iv. 1—Did. xiv. 3, iv. 14. Clement of Alexandria—explicitly iii. 5 (calling it ‘Scripture’), implicitly i. 5, ii. 2, ix. 2. Hip- polytus, Philosoph. ix. 23—implicitly ii. 7; ef. i. 3. Origen +— implicitly ix. 2. Pseudo - Cyprian, de Aleatoribus (ch. iv.; Did. xiv. 2, xv. 3)—ex- plicitly. Clementine Homilies. Pseudo-Clement, de Virginibus. Didascalia (Syriac and Latin). Lactantius and Commodian—echoes of Two Ways only. Athanasius, Syntagma Doctrine —clear traces of i.—vi., less clear of xii., xiii.; de Virgini- tate—quotes or paraphrases, ix., x.; Hragment rept Yevdorpodyrév, cf. xi., xii. Pseudo-Athanasius, Fides Nicena and Didascalia ccexviit. patrum, two recensions of the Syntagma, in which the Did. is freely used. Serapion (of Thmuis, in the Delta), in his Prayer-Book, ¢. 350 A.D., quotes from ix. 4. Optatus, de Schismate Donatist. i. 21, quotes iv. 3%. Augustine cites Did. i. 5 fin. (see below, iii. (d)). Canons of Basil (Egyptian, 5th cent.) uses the 7wo Ways. Severinus, Doctrina de Sapientia—explicitly (Lwo Ways, and perhaps more). Benedictr Regula, iv. (Two Ways). John Climacus (¢. 580)—implicitly i. 4, 5. Dorotheus of Palestine (¢. 590)—implicitly iii. 1, 10. Boniface of Mainz, Admonitio (S. Petri) sive predicatio 8. Boni-
fatu, appears to know more than the Zwo Ways.
(6) Historical testimonia—
[Ireneeus. The authenticity of the Pfaffian Fragment is too dubious to warrant citation of its devrépar r&v droorddwy diardéecs].
Pseudo-Cyprian, de Aleatoribus, iv., ‘Et in Doc- trinis apostolorwum: Si quis frater delinquit in ecclesia et non paret legi, hie nec colligatur donee poenitentiam agat, et non recipiatur ne in- quinetur et impediatur oratio vestra’ (Did. xiv. 2, Xv. 3).
Eusebius, Hist. Eccl. iii. 25, "Ey rots vé0os (here =non-canonical books) cararerdxyOw kat r&v Mavrov mpdtewy 1 ypadh & Te Neyduevos Iouwhy cal h’Amroxd- Auyis ILérpou Kal mpos rovros % hepouévy BaprydBa erioToAH Kal TOV GarogTd\wv ai Aeyopevar Ardayat where Rufinus has Doctrina que dicitur aposto- lorum (so the Syriac).
Athanasius, Festal Epistles, 39, éort kat érepa BiBrla rovTwv ewer, od Kavovigdueva mev TeTuTMméeva de mapa Tov marépwr dvaywioKkerbar Tols Apr mpooepxo-
* The ‘Teaching’ as known to Barn. xxi. 1 seems to have had some such closing exhortations.
t Origen’s quotation of what occurs in Did. iii. 10 is probably from Barnabas, which he has just cited. So the echo of the same passage in Dionysius of Alexandria may also be indirect,
DIDACHE
DIDACHE 443
a ea)
pévows Kal Bovdopevors KarnxeicAar Tov Tis evoeBelas Néyov' Dogpia Toroudvros kat Vopia Tipax Kal’ UoOnp kal *Iovdl@ cal TwBias cat Avdaxy kadoupevy Tov atroatédov Kal 6 ILlouujy.
Optatus, /.c., “Et in capitibus mandatorum, Non facies scisma’ (Did. iv. 3%).
Rufinus, Comm. in Symb. Apost. 38, ‘In Novo vero Testamento libellus qui dicitur Pastoris sive Hermee, [et is] qui appellatur Ducw Vie vel Jud- icium secundum Petrum. Quee omnia legi quidem in ecclesiis voluerunt, non tamen proferri ad auctoritatem ex his fidei confirmandam.’ These are ‘ Ecclesiastical,’ not ‘Canonical’ books. But they are not ‘Apocryphal,’ or such as were not to be read in church. This distinction should be borne in mind in considering the following, and especially the silence of Western lists of canonical and other books, like that of Codex Claromontanus and the Decretum Gelasii.*
Nicephorus, Stichometry (using list of sec. v.—-vi.).
Kal doa THs véas dvTidéyovTac* Arroxdduyrs "Iwdvvov ortx. 1400 a Ilérpov OU BapvaBa émtorony ;, 1360
boa THs véas OtabjKns amdxpupa* Tlepiodos ILavAou
x» Hérpov . Iwdavyvov 29 Gwpua
Biayyéuoy kara Owpar Asay} amoorédov orlxy. 200 Catalogue of the Sixty Books. Kal doa amoxpupa* (OT apocryphal books, mostly not in LXX) *TaxwBou toropla Tlérpou daroxaduyes MeptoSor cat Ardaxal tov arooréhwv BapvdBa émicroH Tlavdov rpagis
The importance of this entry, which recurs in a list in Cod. Baroce. 206, is that the analogy of the meptodor of Apostles (seen from Nicephorus’ list to include separate works) points to more than one work known as Acéax} TGv droorddwy ; and this in turn easts back light on Eusebius’ phrase. On the other hand, the Syriac and Latin versions of Eusebius suggest that only one work of the name was generally known in Syria and Italy respect- ively, towards the end of the 4th century.
(c) Conclusions.—The impressions conveyed by this body of evidence may now be stated. (1) Knowledge of the Zwo Ways (=Did. i.—vi.) is far more general than that of the full Didache. (2) But in this matter a broad distinction long existed between Palestine, or Syria, and other centres of Christianity. In the former, the two seem to stand on much the same level down to Eusebius, though after his day one of them tended to fall out of use. This was probably the fuller form, now superseded
by the Didascalia and Apost. Const. (3) Else- where the bulk of the rules in the full Didache seem never to have suited existing usage, or at least speedily fell out of touch therewith. Hence it is mainly the Eucharistic parts t which have left traces on the literature of the 2nd and 3rd cents., both in Egypt and in the West. Yet several Egyptian witnesses of the 4th and 5th