Chester G. Starr, The Roman Imperial Navy 31 B.C.- A.D. 324. Second edition. Cambridge, W. Heffer, 1960. 1 vol. 15,5 χ 23,5 cm, 232 pp., 1 carte. Prix : 28 s.

W. Heffer and sons, who have been performing a very useful service in recent years by resuscitating valuable books that have gone out of print, can be congratulated for bringing back Starr's Roman Imperial Navy, first published in 1941. It is the only complete work of its kind in any language, it treats an important subject, and it is a thorough, scholarly piece of work.

This printing is called a « second edition ». It is not really that ; like Heffer's other republications, it is a photographically reproduced copy of the original plus several pages of addenda. Since ample notices appeared when the work first came out (e.g., my review in Classical Weekly 35 [1942] 136-7, Sherwin-White's in Journal of Roman Studies 33 [1943] 110-12, Tarn's in Classical Review 56 [1942] 125-6), I will deal only with the new material.

The addenda are limited strictly to the listing of literature and evidence published since 1941 ; nowhere does Starr suggest any changes for his text. This is a pity. His book, though good, is not perfect ; reviewers had revealed half a dozen places where he had, with more of less serious consequences, forced the evidence to conclusions it could not legitemately bear. As a matter of fact, Starr treats his original text as if it were sacrosanct. I had, for example, pointed out that the name Sol, which he suggested (p. 113) for the liburnian in P. Grenf. II. 180, was paleographically ; his addendum on the point (to p. 121) merely contains a vague reference to my review (the ship is actually named the Fides ; see Mitteis, Chrest. 339). An addendum to p. 53 mentions J. S. two articles on the oarage of a trireme, brilliantly reasoned and argued pieces which, in my judgements, effectively refute Tarn's views and Starr's as well, since his are a variation of Tarn's ; the addendum reads as if Tarn had had the final word, though in actual fact it was the other way around.

Starr has overlooked some significant new material. He had missed S. Panciera's exhaustive study of the liburnian (« Liburna », Epi-