and sold' would be ȧréÔOVтO. 6 Conscious of this, Mr. Freeman writes in his note, 'whatever is to be made out of ȧrédoσav or ȧédovтo, it is plain' that the captives were sold. He is quite sure of the fact—for the 'Hykkarian slaves' are mentioned at a later stage of the story, and if so, he is logically bound to alter ȧédoσav to ȧédoνTO, as many editors do. Grote, keeping the text, interpreted, they gave back the slaves,' namely for a ransom ; but the mention of the Hykkarian slaves again confutes this, and, asks Mr. Freeman, who was there to ransom them?' We must therefore either accept the conjecture άédovтo-on Mr. Freeman's own showing-although here we have only the 'meaning,' and not both meaning and palaeography' playing into one another's hands; or we must find some interpretation of åπédoσav consistent with the Hykkarians becoming slaves. As we find these Hykkarians afterwards as slaves of Athenian officers and soldiers, might the word possibly be used here in its regular sense of duly assigned,' 'gave as a due?' The slaves were distributed according to some method among the officers and soldiers, all glad, no doubt, to have the opportunity of buying them cheap. It would be natural to suppose that the preemption was given to the officers, the next choice to the soldiers; and the captives who still remained were then sold together. It is just conceivable that a proceeding of this kind, familiar to the contemporaries of Thucydides, might have been expressed by

ἀπέδοσαν.

I must refer to one more difficulty on which Mr. Freeman has a good deal to say in Appendix xiii. Thucydides (vii., 2, 4) is describing the state of the Athenian works at the time of the coming of Gylippos. After speaking of the wall on the southern side of the round fort, which he throughout calls a Kúkλos, and stating that it had been finished as far as the great harbour, except a small bit on the seaside (πλὴν παρὰ βραχύ τι τὸ πρὸς τὴν θάλασσαν τοῦτο δ ̓ ἔτι ᾠκοδόμουν) he goes on : τῳ δέ ἄλλῳ τοῦ κύκλου πρὸς τὸν Τρώγιλον ἐπὶ τὴν ἑτέραν θαλασσαν λίθοι τε παραβεβλημένοι τῷ πλέονι ἤδη ἦσαν κ.τ.λ. What Thucydides means to say is perfectly plain. Stones had been already deposited in sufficiency to nearly complete the wall which, extending northward from the round fort, was to reach the sea at the bay of Trogilos. The Athenian wall north of the

kyklos, as well as the wall south of the kyklos, was progressing, though it had not progressed quite so quickly. But though this is evidently what Thucydides means to say, he does not succeed in saying it, if the text is right. Indeed if the text is right, it is very hard to know what he says. If TOû KÚKλou could signify the whole Athenian wall, т äλ TOû KÚKλou could mean the other part of the wall,' namely the northern half, taking the round part as the centre. But the wall was not a circuit or even the segment of a circle, and therefore it could not have been called Kúkλos. If Thucydides used TOû kúkλov in this connexion, he could only mean the round fort, to which he has already applied the word repeatedly. But if TOû Kúkλou means the fort, what is the meaning of τῷ ἄλλῳ τοῦ κύκλου, ‘the other part of the round fort? Commentators have attempted to twist the words to mean the part of the wall on the other (namely the northern) side of the round fort;' Grote takes it to mean this; and Mr. Freeman asks 'why should it not?' But this is to play fast and loose with the Greek language. Such a usage is quite unparalleled. This is a clear case in which the text cannot stand, and one is obliged to consider what error may have crept in. Mr. Freeman has a hit at those who strike out the 'the very important words, TOû KUKλov πρὸς τὸν Τρώγιλον.” It seems to me that it is indeed quite gratuitous to strike out the last three words, but I am very much disposed to think that TOû kúkλou may be due to interpolation. Strike out these two words, which might easily have been inserted by some one who did not understand what the kúkλos of the Athenians was; and the sentence becomes intelligible. As the words stand, I submit that TOû KUKλOU is not 'important,' because it is not intelligible. If it is to be kept and become intelligible, some change of another kind must be made. Insert årò before it and read τῷ δὲ ἄλλῳ ἀπὸ τοῦ κύκλου πρὸς τὸν Τρώγιλον,—then we have a sentence which is really lucid. But something must be done; some 'konjektur' must be resorted to in order to make sense. And it seems to me that Mr. Freeman is a little inconsistent in being so coy towards emendations in a case where the presence of error is so obvious, while in some other cases where it is possible to construe he entertains conjectures. Interpolations in classical

texts are so frequent that in a case like this it would not be very bold to suspect one.

Mr. Freeman's general spirit of caution cannot be praised too much, and if he goes too far, it is on the right side. But it seems to me that he has been betrayed into certain inconsistencies by his theoretical dislike of 'konjektur.' His doctrine in the preface is not strictly consistent with his treatment of all the particular cases in the volume; and his treatment of some particular cases is not consistent with his treatment of others. Conjectural emendation often goes to such outrageous lengths, that sober-minded people who have a weakness for evidence are tempted to denounce it altogether. Yet they have to admit, when special instances are put before them, that conjectural emendation is permissible and desirable within certain limits. The real problem is to define these limits; and I only wish that Mr. Freeman, with his unrivalled power of distinguishing, had not been hindered by his horror of emendation from determining its legitimate range according to general principles of evidence.

If Mr. Freeman does not spare the guessers, who undertake to restore texts, neither does he spare those other guessers who undertake to search out Quellen or start new Fragen. Much of the First Appendix, which deals with his authorities, is taken up with ridicule of the Quellen-litteratur, and, if one may coin the word, the Fragen-litteratur which have arisen in Germany. 'It calls up strange feelings,' he writes (p. 589) when one turns from reading his [Thucydides'] pages by the shore of the Great Harbour, from testing the perfection of his picture on the height of Epipolai or by the banks of Assinaros, and finds that the restless ingenuity of German scholars has developed a Thukydideische Frage. Everything else has been cavilled at and guessed at; so those who cannot live without cavilling and guessing have come at last to cavil and guess at those things which cannot be spoken against. Things have indeed changed, since it was thought a heinous sin in Grote himself to hint, not that Thucydides had misrepresented a single fact, but that personal feelings had once led him to pronounce a judgment which the facts of his own narrative did not bear out. On such grounds, in those days a

clever writer of imitative verses ventured to match himself with the great master, and to rejoice that such an one as he was no member of either English University. The position taken by Grote, which then was deemed impiety against Thucydides, would now pass for a superstitious worship of him. For the tone of the new school is often that of religious reformers attacking some form of idolatry.' 'Sometimes we are forbidden to believe what Thucydides tells us; sometimes it seems that we are almost forbidden to believe that there was any Thucydides at all. Even in our own land we have been ordered, with all the irresistible authority of a "headmaster," to cast away half the text that was good enough for Thirlwall, Arnold and Grote. And a German scholar, with a double allowance of Scharfsinn, knows exactly how much was thrust into the text by a "bloodthirsty forger ("ein blutdürstiger Verleumder") a being more terrible, one is driven to suppose, than the author of the false Phalaris or the false Ingulf.' Mr. Freeman's attitude to the products of German ingenuity, which he enumerates, is characterized by English common sense; and in his own views on the way in which Thucydides wrote his story, he shows-to use the words which he himself applies to Grote- the true Scharfsinn of a man who knows practically what he is about.'

In the course of this Appendix he criticizes those strange words of Mr. Jowett, on which, in the January number of this Review, I also ventured to protest. He defends Diodôros and Plutarch, as sources of history, against the unqualified sentence of condemnation which Mr. Jowett does not hesitate to pass against them. A great principle is involved, and Mr. Freeman's words are clear and unanswerable.

'It is truly wonderful how a Professor of Greek, who must be familiar with every word of so important a part of Greek literature as the writings of Diodôros and Plutarch, can have mistaken their useful compilations for 'the fictions of later writers." Mr. Jowett surely does not suppose that Diodôros and Plutarch deliberately invented everything which they record but which is not recorded by Thucydides. Plutarch and Diodôros used such materials as they had, Thucydides himself among them. "Fiction is a hard word even for Timaios; it is utterly out of place as applied to the part of the history of Philistos with which we are now concerned. From his narrative, the narrative of a contemporary and actor, Diodôros and

Plutarch have preserved to us endless little local and personal details which it was natural that a Syracusan eyewitness should record, but which had little interest for an Athenian visitor, even a few months later. Precious scraps like these, fresh from the scene and the actor, have much less of the character of "transparent gauze" than the grossly partisan writings of Xenophon, whom Mr. Jowett counts among the vendors of "good cloth." It would be the most curious question of all to see what kind of history of Pelopidas and Epameinondas could be woven out of that cloth only. The writers of "modern histories of ancient Greece "-Thirlwall and Grote for instance-have 'simply done their duty by "patching together," in Mr. Jowett's scornful phrase, every means of knowledge which they found open to them. In attempting to carry out the same process somewhat further than they did, I feel sure that I should have had their good word. In short, if Mr. Jowett's rule were to be accepted, there would be an end to all historical criticism. There would be an end to all writing of history, almost to all reading of it. We are solemnly called on to shut our ears to a large part of our evidence. Because one writer undoubtedly stands high above all others, we are bidden to pass by the statements- fragmentary, indeed, but still the statements-of another writer, doubtless his inferior in many points, but whose means of knowledge were, from one side of the story, greater than his own.'

Mr. Jowett's view is the view of a man of letters, who judges history altogether from a literary standpoint, and who does not care to hear what happened for its own sake, but only when it is told with literary effect. Nor is it the case, as he seems to imply, that literary merit and truth are always united. Tacitus is a notable instance of a historian in whose pages we find what we can hardly hesitate to call 'fictions' related with consummate art. How could we tell the history of the early Empire trusting to Tacitus alone, and rejecting the aid of Suetonius and Dion Cassius? It is perfectly true that Plutarch and Diodôros are, from a literary aspect, nowhere when compared with Herodotus, Thucydides, and Xenophon. But their literary inferiority does not abolish their historical value. If Diodoros was uncritical, and no master of style, he could nevertheless use his eyes, and copy into his own pages what he read in Philistos and other sources which were available to him, and are no longer available to us. Speaking quite seriously, one might wish that Mr. Jowett, or some one else, would construct a history of Greece, using no sources except Herodotus, Thucydides, and Xenophon. It would be