the American scholar in reading loci with some MSS. for “ioci” in 22. 4 (the subjects were serious and not jokes, and iocus could hardly mean "conviviality "); and in 46.1, in maintaining the MSS. reading sordes omnium et torpor procerum, "all are dirty and the chiefs lazy"; for omnes opposed to proceres, Prof. Gudeman compares 11. 1. But he seems to have rightly disregarded Lipsius's omission of non (approved of by Prof. Gudeman) before multum in 15. 1. The succeeding lines show that the nation had become less active since Cæsar's time. We venture to think that victus inter hospites comis (21. fin.) is sound. It was not only that the Germans kept open house, but they did so willingly: they were always quite courteous towards their uninvited guests, a course of conduct a sophisticated people like the Romans of Tacitus's day would marvel at. In 33. 1, penitus excisis is rightly retained, though inaccurate historically. Tacitus was not infallible. Perhaps in 38. 3 we should read retrosum comuntur for "retro sequuntur": in 40. 4 fota for "nota" (fovere being a word much affected by Virgil, and in harmony with the general poetical colouring of the passage, and with "amata" immediately following); and in 45. 7 qua, for “ quae," and translate "as there are groves in the East where balsam exudes, so there are [groves] in the West where substances, forced out by the heat of the sun which is close to them, and becoming liquid, flow into the sea," &c. Of course sucina for "vicini" is very attractive, but we fear delusive.

In 28. 2

In Agricola 10. 4, perhaps we should read sed transgressis et: and in 11. 2, we think that Mr. Furneaux should not have adopted vi of Rhenanus for the MSS. usu which gives a very fair sense "the practices of their original state still continuing." Also we cannot approve of his deserting the manuscripts in 19. 4, which virtually give tributorum auctionem (or -es) inaequalitatem onerum. remigrante of Puteanus (MSS. " remigante") seems very probable; t simply means 'returning': cf. Cic. Fam. ix. 18. 4. The simplest remedy of 34. 3 is to suppose a small lacuna extremo metu corpora defix<a fix>ere aciem in his vestigiis. In 36. 2, there is no reason to reject the MSS. foedare (cp. Plaut. Amph. 246) for fodere: in 37. 4, perhaps 'ntem" conceals temere rather than idem: and in 38. 1, Schoene's Britanni <ubi>que palantes, for "Britannique palantes" at least deserved mention. It is deeply to be deplored that the Toledo Ms. is not yet accessible; as it will probably throw some light on the many serious corruptions of this most interesting treatise.

66

We think that in Dialogus 10. 4 a mention should have been made of Andresen's altiorum for "aliarum": but we think Mr Furneaux is certainly right in ejecting "ex his" and "expressis " in §§ 7 and 8 of that chapter. Possibly in § 6 "aut" should be altered to "autem," either = 'on the other hand' (cf. Brix on Plaut. Menaech. 1090, and perhaps 32. 1 below), or = 'now,' like the Greek dǹ, a usage common in Plautus and Terence: cf. for example

Adelph. 185. We could have wished that the difficult Ms. reading in 18. 5, attritum (which can be defended: cf. Plin. Ep. v. 10. 3: ix. 35. 2 Quintil. x. 4. 4. quoted by Peterson) had not been postponed in favour of the easy aridum, however suitable that word may be as far as the sense goes. Mention should have been made of Peterson's quique alii omnes in 21. 1. Possibly in 22. 5 we should read ol<eum redolentia: cf. what Pytheas said of Demosthenes (Plut. Dem. 8. 3) that his arguments Avxvíov olev and in 26. 3 the desperate corruption "sicut his ela," possibly conceals saeculi huius: cf. Cic. Phil. ix. 13.

There are excellent indices of proper names. As the work issues from the Clarendon Press, it is superfluous to praise the printing and general external appearance.

A History of Spain, by U. R. BURKE, M.A., Second Edition, with Additional Notes and an Introduction by MARTIN A. S. HUME, in 2 vols. 1900. Longmans, Green & Co.

THE second edition of Mr. Ulick Burke's valuable History of Spain" differs in no important respect from the first. The format is smaller, and the chapters on the Bullfight, Architecture, the Monetary System, and Music, have been translated to the end of the second volume. We think that Mr. Burke was right in his instinct to attempt to work these subjects into his text, but it must be owned that he did not wholly succeed; they bore the air of an interruption, and it is better that they should wear the air of an appendix. Mr. Hume, who hopes that the work "may be regarded as a classic," has corrected mistakes which the author would have corrected himself, if he had lived, and has added a few additional footnotes of his own.

ment.

In his preface Mr. Hume insists upon the fact that "the social impress that the Romans left upon the people has never been obliterated or greatly diminished." It is true; and it brings us to one great defect in Mr. Burke's book, namely, his inadequate treatHe had not investigated Roman civilisation in Spain, in its full length and breadth and depth; Hispania Romana is dismissed in about ten pages. If the Roman period has exerted such a permanent influence over the subsequent history, then surely the historian is bound to set forth fully what is known of the condition of the country and the municipalities under Roman government. Mr. Burke was aware of the important documents of Julia Genetiva, Salpensa, and Malaca; but he dismissed them in a note. The

early chapters are the weakest part of the work, and an editor would have been justified in dealing more freely with them than Mr. Hume has thought it right to do. On p. 10 the reader is left to suppose that Saguntum may have something to do with Zacynthus. On p. 8 the note on Tarshish is not up to date. The author was right in identifying it with Tartessus; but it is perfectly useless to refer to persons like Depping, Marina, or Dean Stanley, whose authority is of no account in such a matter; while the reference, which is at once necessary and sufficient, to a paper by M. Th. Reinach, is omitted.

We cannot think that the editor's note on the Iberian question is illuminating. It would have been well if he had withheld his hand here; for a writer who talks of the Celtiberians as "speaking some form of Sanskrit, more or less closely allied to what we know as Celtic," shows that he is absolutely ignorant of Indo-European philology, and totally incompetent to speak on the problems of prehistoric Europe.

The new Priscillian documents were unknown to Mr. Burke (p. 60), they seem to be likewise unknown to Mr. Hume.

But the early periods, in which Mr. Hume is not at home, form but a small fraction of the work, which on the whole could not have found a fitter or more capable editor.

CONTENTS OF No. XXVI.

I. Blaydes' Choephoroi.

II. Two Notes on Eusebius.

III. Some Notes on Cicero's Epistles from 57 to 54 B.C.

IV. On the Fixed Alexandrine Year.

V. Some Observations on the Peace of Aristophanes.

VI. Observations on Dr. Merry's Odyssey.

VII. The Quasi-Caesura in Vergil.

VIII. Attachment against Sir James Carroll, dated 1st March, 1631, Mayor of Dublin, 1612, 1617, and 1627.

IX. The Identity of Ajax.

X. The Thorician Stone.

XI. A New Theory of the Ekkyklema; and Two Short Notes.

XII. Notes on Aristotle, Parva Naturalia.

XIII. Dr. Blaydes's Edition of the Agamemnon.

XIV. An Unpublished Essay by Berkeley.

XV. Considerations on the Cause and Regulating Principle of Variable and Common Quantity in Latin.

XVI. REVIews.

Orders to be addressed to the Publishers.