. Accordingly Athanasius complains loudly of their exegesis (Ep. Æg.3-4, cf. Orat. i.8, 52), and insists (id. i.54, cf. already de Decr.14) on the primary necessity of always conscientiously studying the circumstances of time and place, the person addressed, the subject matter, and purpose of the writer, in order not to miss the true sense. This rule is the same as applies (de Sent. Dion.4) to the interpretation of any writings whatever, and carries with it the strict subordination of the allegorical to the historical sense, contended for by the later school of Antioch, and now accepted by all reasonable Christians (see Kihn in Wetzer-Hergenröther's Kirchen-Lex. vol. i. pp.955-959, who calls the Antiochene exegesis certainly a providential phenomenon;' also supra, p. xxviii., note 1).
(b) The Canon of Scripture accepted by Athanasius has long been known from the fragments of the thirty-ninth Festal Letter (Easter, 367). The New Testament Canon comprises all the books received at the present day, but in the older order, viz., Gospels, Acts, Catholic Epistles, Pauline Epistles (Hebrews expressly included as S. Paul's between Thess. and Tim.), Apocalypse. The Old Testament canon is remarkable in several ways. The number of books is 22, corresponding to the Alexandrian Jewish reckoning, not to the (probably) older Jewish or Talmudic reckoning of 24 (the rolls of Ruth and Lam. counted separately, and with the Hagiographa). This at once excludes from the Canon proper the so-called Apocrypha,' with the exception of the additions to Daniel, and of Baruch and the Epistle,' which are counted as one book with Jeremiah. The latter is also the case with Lamentations, while on the other hand the number of 22 is preserved by the reckoning of Ruth as a separate book from Judges to make up for the exclusion of Esther. This last point is archaic, and brings Athanasius into connection with Melito (171 a.d.), who gives (Eus. H. E. iv.26.14, see also vol.1, p.144, note 1, in this series) a Canon which he has obtained by careful enquiry in Palestine. This Canon agrees with that of Athanasius except with regard to the order assigned to Esdras' (i.e. Ezra and Nehemiah, placed by M. at the end), to the twelve in one book' (placed by M. after Jer.), and Daniel (placed by M. before Ezekiel). Now, Esther is nowhere mentioned in the N.T., and the Rabbinical discussions as to whether Esther defiled the hands' (i.e. was canonical') went on to the time of R. Akiba (135), an older, and even of R. Juda the holy' (150-210), a younger, contemporary of Melito (see Wildeboer, Ontstaan van den Kanon, pp.58, sq., 65, &c.). The latter, therefore, may represent the penultimate stage in the history of the Hebrew canon before its close in the second century, (doubted by Bleek, Einl.^5, §242, but not unlikely). Here, then, Ath. represents an earlier stage of opinion than Origen (Eus. H. E. vi.25), who gives the finally fixed Hebrew Canon of his own time, but puts Esther at the end. As to the number of books, Athan. agrees with Josephus, Melito, Origen, and with Jerome, who, however, knows of the other reckoning of 24 (nonnulli' in Prol. Gal.). Athanasius enumerates, as outside the Canon, but appointed by the Fathers to be read by those who newly join us,' Wisdom, Ecclesiasticus, Esther, Judith, and Tobit, as well as what is called the Teaching of the Apostles and the Shepherd. In practice, however, he quotes several of the latter as Scripture' (Wisdom repeatedly so, see index to this vol.); The Shepherd' is most profitable,' and quoted for the Unity of the Creator (and cf. de Decr.4), but not as Scripture;' the Didache' is not used by him unless the Syntagma (vide supra, p. lix.) be his genuine work. He also quotes 1 Esdras for the praise of Truth, and 2 Esdras once, as a prophet.' Daniel' includes Susanna and Bel and the Dragon.
(c) On the sufficiency of Scripture for the establishment of all necessary doctrine Athanasius insists repeatedly and emphatically (c. Gent.1, de Incarn.5, de Decr.32, Vit. Ant.16, &c., &c.); and he follows up precept by example. His works are a continuous appeal to Scripture.' There is no passage in his writings which recognises tradition as supplementing Scripture, i.e., as sanctioning articles of faith not contained in Scripture. Tradition is recognised as authoritative in two ways: (1) Negatively, in the sense that doctrines which are novel are prima facie condemned by the very fact (de Decr.7, note 2, ib.18, Orat. i.8, 10, ii.34, 40, de Syn.3, 6, 7, and Letter 59, §3); and (2) positively, as furnishing a guide to the sense of Scripture (see references in note on Orat. iii.58, end of ch. xxix.). In other words, tradition with Athanasius is a formal, not a material, source of doctrine. His language exemplifies the necessity of distinguishing, in the case of strong patristic utterances on the authority of tradition, between different senses of the word. Often it means simply truth conveyed in Scripture, and in that sense handed down' from the first, as for example c. Apol. i.22, the Gospel tradition,' and Letter 60.6 (cf. Cypr. Ep.74.10, where Scripture is divinæ traditionis caput et origo.'). Moreover, tradition as distinct from Scripture is with Athanasius not a secret unwritten body of teaching handed down orally  , but is to be found in the documents of antiquity and the writings of the Fathers, such as those to whom he appeals in de Decr., &c. That the appeal of Athanasius was to Scripture, that of the Arians to tradition' (Gwatkin) is an overstatement, in part supported by the pre-Nicene history of the word homoousion (supra, p. xxxi. sq.). The rejection of this word by the Antiochene Council (in 268-9) is met by Athanasius, de Synod.43, sqq., partly by an appeal to still older witnesses in its favour, partly by the observation (§45) that writing in simplicity [the Fathers] arrived not at accuracy concerning the homoousion, but spoke of the word as they understood it,' an argument strangely like that of the Homoeans (Creed of Niké, ib. §30) that the Fathers [of Nicæa] adopted the word in simplicity.'
(d) Connected with the function and authority of tradition is that of the Church. On the essential idea of the Church there is little or nothing of definite statement. The term Catholic Church' is of course commonly used, both of the Church as a whole, and of the orthodox body in this or that place. The unity of the Church is emphatically dwelt on in the opening of the encyclical written in the name of Alexander (infr., p.69 and supr., p. xvi.) as the reason for communicating the deposition of Arius at Alexandria to the Church at large. The joyful mother of children' (Exp. in Ps. cxiii.9) is interpreted of the Gentile Church, made to keep house,' hate ton Kurion enoikon echousa, joyful because her children are saved through faith in Christ,' whereas those of the synagogue' are apolei& 139; paradedomena: the strong city' polis perioches and Edom' of Ps. lx.11 are likewise interpreted of the Church as gathered from all nations; similarly the Ethiopians of Ps. lxxxvii.4 (where the de Tit. pss. gives a quite different and more allegorical sense, referring the verse to baptism). The full perfection of the Church is referred by Athanasius not to the (even ideal) Church on earth but to the Church in heaven. The kingdom of God' (Matt. vi.33) is explained as the enjoyment of the good things of the future, namely the contemplation and knowledge of God so far as man's soul is capable of it,' while the city of Ps. lxxxvii.1-3 is he ano Ierousalem in the de Titulis, but in the Expositio the Church glorified by the indwelling of the Only-begotten.' In all this we miss any decisive utterance as to the doctrinal authority of the Church except in so far as the recognition of such authority is involved in what has been cited above in favour of tradition. It may be said that the conditions which lead the mind to throw upon the Church the weight of responsibility for what is believed were absent in the case of Athanasius as indeed in the earlier Greek Church generally.
But Athanasius was far from undervaluing the evidence of the Church's tradition. The organ by which the tradition of the Church does its work is the teaching function of her officers, especially of the Episcopate (de Syn.3, &c.). But to provide against erroneous teaching on the part of bishops, as well as to provide for the due administration of matters affecting the Church generally, and for ecclesiastical legislation, some authority beyond that of the individual bishop is necessary. This necessity is met, in the Church as conceived by Athanasius, in two ways, firstly by Councils, secondly in the pre-eminent authority of certain sees which exercise some sort of jurisdiction over their neighbours. Neither of these resources of Church organisation meets us, in Athanasius, in a completely organised shape. A word must be said about each separately, then about their correlation.
(a) Synods. Synods as a part of the machinery of the Church grew up spontaneously. The meeting of the Apostles and Elders' at Jerusalem (Acts xv.) exemplifies the only way in which a practical resolution on a matter affecting a number of persons with independent rights can possibly be arrived at, viz., by mutual discussion and agreement. Long before the age of Athanasius it had been recognised in the Church that the bishops were the persons exclusively entitled to represent their flocks for such a purpose; in other words, Councils of bishops had come to constitute the legislative and judicial body in the Church (Eus. V. C. i.51). Both of these functions, and especially the latter, involved the further prerogative of judging of doctrine, as in the case of Paul of Samosata. But the whole system had grown up out of occasional emergencies, and no recognised laws existed to define the extent of conciliar authority, or the relations between one Council and another should their decisions conflict. Not even the area covered by the jurisdiction of a given Council was defined (Can. Nic.5). We see a Synod at Arles deciding a case affecting Africa, and reviewing the decision of a previous Synod at Rome; a Council at Tyre trying the case of a bishop of Alexandria; a Council at Sardica in the West deposing bishops in the East, and restoring those whom Eastern Synods had deposed; we find Acacius and his fellows deposed at Seleucia, then in a few weeks deposing their deposers at Constantinople; Meletius appointed and deposed by the same Synod at Antioch in 361, and in the following year resuming his see without question. All is chaos. The extent to which a Synod succeeds in enforcing its decisions depends on the extent to which it obtains de facto recognition. The canons of the Council of Antioch (341) are accepted as Church law, while its creeds are condemned as Arian (de Syn.22-25).
We look in vain for any statement of principle on the part of Athanasius to reduce this confusion to order. The classical passage in his writings is the letter he has preserved from Julius of Rome to the Eastern bishops (Apol. c. Ar.20-35). The Easterns insist strongly on the authority of Councils, in the interests of their deposition of Athanasius, &c., at Tyre. Julius can only reply by invoking an old-established custom of the Church, ratified, he says, at Nicæa (Can.5?), that the decisions of one Council may be revised by another; a process which leads to no finality. The Sardican canons of three years later drew up, for judicial purposes only, a system of procedure, devolving on Julius (or possibly on the Roman bishop for the time being) the duty of deciding, upon the initiative of the parties concerned, whether in the case of a deposed bishop a new trial of the case was desirable, and permitting him to take part in such new trial by his deputies. But Athanasius never alludes to any such procedure, nor to the canons in question. (Compare above, pp. xlii., xlvi.).
The absence of any a priori law relating to the authority of Synods applies to general as well as to local Councils. The conception of a general Council did not give rise to Nicæa, but vice versa (see above, p. xvii.). The precedent for great Councils had already been set at Antioch (268-9) and Arles (314); the latter in fact seems to be indirectly called by S. Augustine plenarium universæ ecclesiæ concilium; but the widely representative character of the Nicene Council, and the impressive circumstances under which it met, stamped upon it from the first a recognised character of its own. Again and again (de Decr.4, 27, Orat. i.7, Ep. Æg.5, &c., &c.) Athanasius presses the Arians with their rejection of the decision of a world-wide' Council, contrasting it (e.g. de Syn.21) with the numerous and indecisive Councils held by them. He protests (Ep. Æg.5, Tom. ad Ant., &c.) against the idea that any new creed is necessary or to be desired in addition to the Nicene. But in doing so, he does not suggest by a syllable that the Council was formally and a priori infallible, independently of the character of its decision as faithfully corresponding to the tradition of the Apostles. Its authority is secondary to that of Scripture (de Syn.6, sub. fin.), and its scriptural character is its justification (ib.). In short, Mr. Gwatkin speaks within the mark when he disclaims for Athan. any mechanical theory  of conciliar infallibility. To admit this candidly is not to depreciate, but to acknowledge, the value of the great Synod of Nicæa; and to acknowledge it, not on the technical grounds of later ecclesiastical law, but on grounds which are those of Athanasius himself. (On the general subject see D.C.A.475-484, and Hatch, B.L. vii.)
(b) Jurisdiction of bishops over bishops. The fully-developed and organised patriarchal' system does not meet us in the Nicene age. The bishops of important towns, however, exercise a very real, though not definable authority over their neighbours. This is especially true of Imperial residences. The migration of Eusebius to Nicomedia and afterwards to Constantinople broke through the time-honoured rule of the Church, but set the precedent commonly followed ever afterwards. In Egypt, although the name patriarch' was as yet unheard, the authority of the Bishop of Alexandria was almost absolute. The name archbishop' is here used for the first time. It is first applied apparently to Meletius (Apol. Ar.71) in his list of clergy, but at a later date (about 358) to Athanasius in a contemporary inscription (see p.564^a, note 1). At the beginning of his episcopate (supra, p. xxxvii.) we find him requested to ordain in a diocese of Upper Egypt by its bishop. He sends bishops on deputations (Fest. Ind. xxv., &c.), and exercises ordinary jurisdiction over bishops and people of Libya and Pentapolis (cf. reference to Synesius, supr., p. lxii.). This was a condition of things dating at least from the time of Dionysius (p.178, note 2). In particular he had practically the appointment of bishops for all Egypt, so that in the course of his long episcopate all the Egyptian sees were manned by his faithful adherents (cf. p.493). The mention of Dionysius suggests the question of the relation of the see of Alexandria to that of Rome, and of the latter to the Church generally. On the former point, what is necessary will be said in the Introd. to the de Sent. Dion. With regard to the wider question, Athanasius expresses reverence for that bishopric because it is an Apostolic throne,' and for Rome, because it is the metropolis of Romania' (p.282). That is his only utterance on the subject. Such reverence ought, he says, to have secured Liberius from the treatment to which he had been subjected. The language cited excludes the idea of any divinely-given headship of the Church vested in the Roman bishop, for his object is to magnify the outrageous conduct of Constantius and the Arians. Still less can anything be elicited from the account given by Ath. of the case of the Dionysii, or of his own relations to successive Roman bishops. He speaks of them as his beloved brothers and fellow-ministers (e.g., p.489) and cordially. welcomes their sympathy and powerful support, without any thought of jurisdiction. But he furnishes us with materials, in the letter of Julius, for estimating not his own view of the Roman see, but that held by its occupant. The origin of the proceedings was the endeavour of the Easterns to procure recognition at Rome and in the West for their own nominee to the bishopric of Alexandria. They had requested Julius to hold a Council, and to be himself the judge if he so pleased' (Apol. c. Ar.20). This was intended to frighten Athanasius, but not in the least, as the sequel shews, to submit the decisions of a Council to revision by a single bishop. Julius summoned a Council as described above (p. xliii.), and at the end of a long period of delay and controversy sent a letter expressing his view of the case to the Orientals. This document has been already discussed (p. xliv.). It forms an important landmark in the history of papal claims, standing at least as significantly in contrast with those of the successors of Julius, as with those of his predecessors.
(g) Bishops and Councils. The superiority of councils to single bishops (including those of Rome, Alexandria, and Antioch) was questioned by no one in this age. Julius claims the support, not of authority inherent in his see, but of canons, and on the basis of them claims a voice in matters affecting the Church at large, not in his own name, but in that of us all, that so a just sentence might proceed from all' (Apol. c. Ar.35). Again, just as the judgment of his predecessor Melchiades and his council was revised at Arles in 314 (Augustin. Ep.105.8), so the case of Athanasius and Marcellus was reheard at the Council of Sardica three years after the decision of Julius and his council. The council was the supreme organ of the Church for legislative, judicial, and doctrinal purposes; had any other of superior or even equal rank been recognised, or had the authority of councils themselves been defined a priori by a system of Church law, the confusion of the fourth century would not have arisen. Whether or no the age would have gained, we at least should have been the losers.
 Athanasius is not always innocent of the method of which he complains; e.g. when he uses Isaiah 1:11, pleres eimi, as a proof of the Divine Perfection.  The idea of a mysterious unwritten tradition is a legacy of Gnosticism to the Church. Irenæus, in order to meet the Gnostic appeal to a supposed unwritten Apostolic tradition, confronts it with the consistency of the public and normal teaching of the Churches everywhere, of which the Roman Church is a convenient microcosm or compendium. The idea of a paradosis agraphos is adopted by Clement and Origen, and passes from the latter to Eusebius, and to the Cappadocian Fathers (Basil de Sp. S. 27, applies it only to practical details), Epiphanius, and later writers. Details in Harnack ii. 90, note, cf. Salmon, Infallibility, Lect. ix. On the somewhat different subject of the Disciplina Arcani,' see Herzog-Plitt. s.v. Arkan-Disciplia'  What is conspicuously true of the Second General Council is in reality not less true of the First. Its high authority to later ages is due not to its formal character as a council, but to the character of its work; the consent of the Church, and that not readily given, but as the result of a long process of searching and sifting, has given to it its irreformable' authority. Its authority is expressly put on a par with that of the Antiochene Synod of c. 269, by Ath. de Syn. 43 (consult the whole discussion, pp. 473, 475, &c.). Short of a council which should include every bishop of the entire Church, in unanimous agreement,--an impossible contingency,--the claims of any given council to be truly ecumenical are relative, not absolute; and no consistent theory is possible of the conditions under which a council could by virtue of its constitution claim infallibility for its decisions. The supposed infallibility of general councils lies in reality outside them, in the authority which sanctions and consecrates their decisions. According to the precedent of Nicæa this is the Church diffusive' (cf. p. 489, and Pusey, Councils, p. 225, sq.), and such consent, again, must necessarily be partial and relative. If a more tangible and expeditious theory is wanted, we have it in the Roman system, according to which a council is infallible if ratified by the Pope. This at once puts all such councils, whether local or general, on one level, and affords a ready criterion. In other words, the only consistent (mechanical) theory of the infallibility of councils is one which makes councils superfluous. If such a theory had been known to the Church in the age of councils, the councils would not have been held.
 The idea of a mysterious unwritten tradition is a legacy of Gnosticism to the Church. Irenæus, in order to meet the Gnostic appeal to a supposed unwritten Apostolic tradition, confronts it with the consistency of the public and normal teaching of the Churches everywhere, of which the Roman Church is a convenient microcosm or compendium. The idea of a paradosis agraphos is adopted by Clement and Origen, and passes from the latter to Eusebius, and to the Cappadocian Fathers (Basil de Sp. S. 27, applies it only to practical details), Epiphanius, and later writers. Details in Harnack ii. 90, note, cf. Salmon, Infallibility, Lect. ix. On the somewhat different subject of the Disciplina Arcani,' see Herzog-Plitt. s.v. Arkan-Disciplia'
 What is conspicuously true of the Second General Council is in reality not less true of the First. Its high authority to later ages is due not to its formal character as a council, but to the character of its work; the consent of the Church, and that not readily given, but as the result of a long process of searching and sifting, has given to it its irreformable' authority. Its authority is expressly put on a par with that of the Antiochene Synod of c. 269, by Ath. de Syn. 43 (consult the whole discussion, pp. 473, 475, &c.). Short of a council which should include every bishop of the entire Church, in unanimous agreement,--an impossible contingency,--the claims of any given council to be truly ecumenical are relative, not absolute; and no consistent theory is possible of the conditions under which a council could by virtue of its constitution claim infallibility for its decisions. The supposed infallibility of general councils lies in reality outside them, in the authority which sanctions and consecrates their decisions. According to the precedent of Nicæa this is the Church diffusive' (cf. p. 489, and Pusey, Councils, p. 225, sq.), and such consent, again, must necessarily be partial and relative. If a more tangible and expeditious theory is wanted, we have it in the Roman system, according to which a council is infallible if ratified by the Pope. This at once puts all such councils, whether local or general, on one level, and affords a ready criterion. In other words, the only consistent (mechanical) theory of the infallibility of councils is one which makes councils superfluous. If such a theory had been known to the Church in the age of councils, the councils would not have been held.