I certainly concur in the opinion of Mr Grote that this account cannot possibly be reconciled with the passage in Andocides. Professor Rawlinson, on the other hand, believes that it can. Herodotus, he thinks, 'has not told us all the facts'; and he considers that this is almost certain,' both (a) from anterior probability, and (b) from Herodotus' own words.

(a) The 'anterior probability' is thus explained by Professor Rawlinson. The Alcmæonidæ, with a large party of exiles, had shortly before taken up a position in Attica on the outskirts of Parnes, at a place called Leipsydrium, had fortified the post, and from it kept up a long though unsuccessful struggle against the tyrants. It is probable that they were still at this place when the Spartans, at the instance of the Delphic oracle, determined to expel Hippias'; and the invading army, marching thither to join them, might thus (it is argued) have come into the neighbourhood of Pallene. Now it appears to me more probable that, according to Herodotus, the Alcmæonidæ were not any longer at Leipsydrium when Cleomenes invaded Attica. Her. V. 62:—Αλκμαιωνίδαι, γένος ἐόντες ̓Αθηναῖοι καὶ φεύγοντες Πεισιστρατίδας, ἐπεί τε σφι ἅμα τοῖσι ἄλλοισι Αθηναίων φυγάσι πειρωμένοισι κατὰ τὸ ἰσχυρὸν οὐ προεχώρεε κάτοδος, ἀλλὰ προσέπταιον μεγάλως, πειρώμενοι κατιέναι τε καὶ ἐλευθεροῦν τὰς ̓Αθήνας, Λειψύδριον τὸ ὑπὲρ Παιονίης τειχίσαντες· ἐνθαῦτα οἱ ̓Αλκμαιωνίδαι πᾶν ἐπὶ τοῖσι Πεισιστρατίδῃσι μηχανώμενοι παρ' ̓Αμφικτυόνων τὸν νηὸν μισθοῦνται τὸν ἐν Δελφοῖσι, κ.τ.λ. The occupation of Leipsydrium is here spoken of as an experiment which had been tried, and had failed. The next move (ἐνθαῦτα of the Alcmæonidæ was to establish themselves at Delphi. In the following chapter (63) they are spoken of as 'settled' there : ἐν Δελφοῖσι κατήμενοι. If Leipsydrium and its garrison had figured prominently in the triumphant conclusion of the war against the tyrants, would the name of the place have been ultimately associated in a popular σkоλóv with memories of utter failure and disaster? Athen. XV. 15:—αἶαι, Λειψύδριον προδωσέταιρον, | οἵους ἄνδρας ἀπώλεσας, μάχεσθαι [ ἀγαθούς τε καὶ εὐπατρίδας, ] οἳ τότ ̓ ἔδειξαν οἵων πατέρων ἔσαν.

(b) 'Further,' Professor Rawlinson says, 'the words of Herodotus show that he has not intended to give a full account

of the campaign. For he tells us that, when Cleomenes entered Attica near Eleusis, "his first engagement was with the Thessalian cavalry” (ἡ τῶν Θεσσαλῶν ἵππος πρώτη προσέμιξε, ν. 64)— an expression which implies, at the least, one more engagement, of which he has made no express mention. In fact the engagement with the Thessalians was a mere cavalry fight, of little moment, except that it left the Spartan infantry free to act. The main fight must have been one between this infantry and that of Hippias...' That is to say, Herodotus has given us a particular account of a cavalry skirmish 'of little moment'; but has 'intentionally' omitted to say anything about the 'main fight' by which the war was decided! The necessity of supposing him to have made so singular a demand on the penetration of his readers will, I think, be obviated by attention to the sense of the words πρwτη πρoσéμice. They mean that, in this battle, the Thessalian cavalry was the part of the tyrant's force which first came into action; pooμîçaι having its usual sense of 'coming to close quarters.' They do not mean that this was a first battle, as distinguished from a second fought at another time and place.

If, then, Andocides is alluding to the expulsion of the Peisistratidæ, I conclude with Mr Grote that his account cannot be reconciled with that of Herodotus. Could it be shown that they were reconcileable, then one of Mr Grote's grounds for suspecting the authority of Andocides would have been removed. I should have welcomed such unexpected aid; but I am compelled to doubt, for the reasons I have given, whether Professor Rawlinson has established this point.

As regards another inference drawn by Valckenär from the passage under discussion,-viz. that the burning of Athens in the second Persian invasion is placed by Andocides in the first,I endeavoured to show that it is not only unwarrantable, but opposed to the natural meaning of the words. I regret that Professor Rawlinson did not find space to discuss this point. The case, then, against Andocides, on the most unfavourable supposition, amounts to this: he has given an account of the expulsion of the Peisistratidæ different from that given by Herodotus; and has spoken of the amnesty granted to exiles

and to the disenfranchised in 480 B. C. as a general amnesty; whereas it is probable (though not certain) that the adherents of the family of Hippias would have been specially excepted from it. This is hardly sufficient, as I conceive, to warrant the statement that he is a witness of no value' for the earlier history of his own city. My object was to inquire whether the severity of this judgment would not fairly admit of some abatement. Enough has been said, I think, to show that, though it may sometimes be difficult or impossible to decide between Andocides and other authorities, the charges against him as 'loose,' 'confused,' 'unscrupulous,' require proof; and that, though his evidence may in some points be perplexing without fuller information than we possess, a modern historian is scarcely justified in putting him out of court.

R. C. JEBB.

ON A LOST DIALOGUE OF ARISTOTLE.

THE Hortensius of Cicero is said by Trebellius Pollio (Vita Salonini Gallieni, c. 2) to have been written on the model of the Protrepticus, i. e. the Aristotelian Dialogue entitled the 'Exhortation to Philosophy.' This statement, to be understood doubtless with many limitations, is confirmed in the main by the scattered fragments of Cicero's work which have come down to us from the wreck of Classical literature. The analogy of the Hortensius, therefore, puts us even now in a position to form some idea, however inadequate, of the original Protrepticus: in the first place Aristotle may be supposed to have shewn that the very opponents of philosophy establish their point by philosophizing, in other words, by refuting themselves in the second place the exceeding blessedness of a speculative life was maintained by a line of argument not unlike that in the Tenth Book of the Ethics. Guided by such hints as these Prof. Bernays of Bonn has reconstructed the Aristotelian Dialogue (Dialoge des Arist. pp. 116—122) with the critical tact and poetical insight into the mind of antiquity by virtue of which he stands so completely alone among living scholars.

'Exhortations to philosophy' were a favourite theme with the ancients. A contemporary of Constantine the Great, the Syrian Iamblichus was the author of a Protrepticus which we still possess in a complete form. It would perhaps be difficult to imagine a book more singularly devoid of any literary or philosophic merit of its own; it is the most shameless of centos, about one-third of it being a plagiarism from Plato, while for another third the compiler is manifestly indebted to some Peripatetic archetype. I hope in the following pages to suggest some reasons for the belief that this Peripatetic arche

type was a writing by Aristotle himself, and indeed no other than the long-lost Protrepticus which Cicero is said to have taken as the model for his own Hortensius. For the present I assume that the original Dialogue was a genuine work of its reputed author.

Iamblichus makes no secret of the composite origin of his book (comp. p. 12, ed. Kiessling). The more recent elements in it are easily distinguished from the older by their Neoplatonism, the character of the quotations introduced (from pseudoPythagorean literature) and a certain want of style, a defect in Iamblichus which is acknowledged even by his admiring biographer Eunapius. The first four chapters are of this Neoplatonic kind. At the beginning of the fifth there is a sudden change of manner which the writer is at no pains to conceal: he proclaims his intention henceforth to adopt a more scientific and consecutive tone, an Aristotelian one, as we should term it, interpreting his words by the light of the actual performance which ensues. Here then he starts de novo with the statement of a sort of axiom: 'We men all of us desire happiness'language which at once reminds us of the broad and pregnant assumptions which serve as introduction to more than one wellknown treatise of Aristotle ::

πάντες ἄνθρωποι τοῦ εἰδέναι ὀρέγονται φύσει (Metaph. I. 1). πᾶσα τέχνη καὶ πᾶσα μέθοδος ὁμοίως δὲ πρᾶξίς τε καὶ προαίρεσις ἀγαθοῦ τινὸς ἐφίεσθαι δοκεῖ (Eth. Nic. Ι. 1).

πᾶσα διδασκαλία καὶ πᾶσα μάθησις διανοητικὴ ἐκ προϋπαρχούσης γίνεται γνώσεως (Anal. Post. I. 1).

Can it have been a mere accident that in Cicero's Hortensius also, the Defence of Philosophy commenced with what would naturally seem to be a literal translation of the above words in Iamblichus ?

IAMBLICHUS.

CICERO'S Hortensius. Beati certe omnes esse vo

πάντες ἄνθρωποι βουλόμεθα ev πρáттew (p. 64, ed. Kiessl.). lumus1.

1 Cicero cum vellet in Hortensio dialogo ab aliqua re certa de qua nullus

ambigeret sumere suae disputationis exordium, 'Beati certe,' inquit, ‘om