^{page 18 note 1} Similarly, the completion of H. M. S. Dreadnought rendered all ‘pre-Dreadnought’ battleships obsolete and reduced to the second rank those naval powers which were unable or unwilling to build battleships of the new type in sufficient numbers. Ensor, R. C. K., England 1870–1914, 1936, 364–5Google Scholar discusses some of the political implications of the original Cawdor–Fisher plan, of which Dreadnought was the first result.

^{page 18 note 2} Latest in Gomme, A. W., Historical Commentary on Thucydides, i, 1945,122Google Scholar, and Cary, M., review of Gomme, , C.R. lx, 1946, 28Google Scholar. Torr, , Ancient Ships, 1894, 4Google Scholar, does not ascribe the building of triremes to Aminocles, but his short summary of Th. 1.13.2–4 is very misleading in other respects.

^{page 18 note 3} The analogy of νη⋯ς ⋯εικ⋯σορος for a merchant ship in Od. 9.322 suggests that τριακ⋯ντορος, πεντηκ⋯ντορος (and the alternative forms in -ερος) were early names for warships of 30 and 50 oars respectively. τρι⋯ρης results from a completely different method of classification, which gives the number of oars in a unit but does not state the number of units in the ship. The analogous μον⋯ρης, δι⋯ρης are not attested before Pollux but may have come into use at least as early as τ⍴ɩ⋯⍴ης; δɩ⋯⍴ης may have been the earliest of the three. X. HG. 2. 1. 28 (Athenians surprised at Aegospotami) uses μον⋯κροτος, δ⋯κροτος of triremes with only one or two oars in each unit manned; δ⋯κροτον = ‘bireme’ is first attested in Polybius; the analogous τ⍴⋯κροτος is not attested before the rhetor Aristides. The -κροτος group have so obviously a poetical flavour that they may well be at least as early as the -⋯⍴ης group, For details, see LS.⁹ s.vv., which (s.v. τ⍴⋯κροτος) does not quote Clem. Al. Strom. 1.16. 76. 7. See p. 31, n. 1, below.

^{page 20 note 1} How much earlier the introduction of the penteconter should be dated we do not know, Marm. Par. 15 ascribes the first use of the penteconter to Danaus about 1510 B.C. (Jacoby, , Das Marmor Parium, 1904, 5)Google Scholar, but the legends implying the existence of 50-oared ships appear to be relatively late. For example, Danaus' fifty daughters are not attested before Aeschylus. The hundred-handed Briareos, (Il. 1. 403)Google Scholar has been interpreted as referring to a ship of 100 oars, mainly on the strength of his human name, Aegaeon; the shakiness of this as a foundation for serious argument is obvious.

^{page 19 note 2} With the exception of the passages already quoted, Il. 2. 719–20 and Od. 10. 203–8 (Odysseus has 44 men besides Eurylochus in his crew on his arrival at Aeaea), the standard Homeric ship has 20 rowers: cf. II. 1.309 (since this carried a φορτ⋯ς εὐ⍴εῖα as well as the crew, it may have been a φορτ⋯ς εὐ⍴εῖα like the ship in Od. 9. 322–3), Od. 1. 280 (the ship to take Telemachus to Pylos), 4. 679 (the ship with which Antinous plans to waylay Telemachus). No one who studies Antinous can believe that he would have allowed any scruples about ‘fair play’ to deter him from taking a bigger ship if one had been available. Did Homer not notice this, or can it be that he believed that Odysseus had taken all the Ithacan penteconters to Troy?

^{page 19 note 3} Compare the old East Indiaman, which was built on the same general lines as the contemporary battleship; it was intended to be strong enough to fight any hostile ship which it could not outsail and fast enough to outsail any ship which it could not fight.

^{page 20 note 1} On Polycrates, see now Smith, H. R. W., CV, U.S.A. Fasc. 10 (San Francisco Fasc. I), 1943, 25Google Scholar and plates V–VII. Smith seems to go beyond the evidence in taking this passage to prove that Polycrates introduced ‘marine archers’, if by that he means that Polycrates made archers a permanent element in his ships' companies. Archers were an important component of Polycrates' land forces (Hdt. 3. 45. 3) and must have played a considerable part in his ‘combined operations’ (Hdt. 3.39.4, Th. 1.13.6), but Herodotus does not mention them in either of his references to naval battles fought by Polycrates—against the Lesbians and Milesians (3. 39.4) and against the rebels (3. 45. 2).

^{page 20 note 2} The scale of this contribution is shown by the report that in 480 the Corcyraeans, for the relative size of whose fleet see Th. 1.14.2 (quoted below), are said to have sent 60 ships to take part in operations in the Aegean (Hdt. 7.168. 2).

^{page 21 note 1} Clem, . Al. Strom. I. 16. 76. 7Google Scholar reports a tradition that the Sidonians were the first to build τ⍴⋯κροτον να⋯ν. This would have to be before the destruction of Sidon by the Assyrians in 672; there is other evidence that the Phoenicians were far ahead of the Greeks in naval architecture.

^{page 21 note 2} Compare Peter the Great's importation of British and Dutch shipwrights into Russia in the latter part of the seventeenth century, Th.'s use of κ α ⋯ ∑αμ⋯οɩς expresses his, or his source's, surprise that a citizen of one Greek State should build warships for another.

^{page 21 note 3} This is not corroborated. Hdt. 2. 159. 1 reports tha t Necho was said to have built triremes before his invasion of Syria (608); in view of the close relations between Corinth and Egypt in the time of Periander, the Corinthians may have learned about triremes then, but there is no evidence that Periander (c. 625–585) owed any of his strength to the possession of triremes.

^{page 22 note 1} If Μασσαƛ⋯αν is not a blunder for ՚Αƛαƛ⋯αν (Gomme, , op. cit., 124Google Scholar), this battle is otherwise unknown. The terminus a quo is the founding of Massalia (c. 600), but the terminus ante quem is less certain; Th.'s arrangement suggests that the battle took place after the rise of Polycrates and perhaps even after his fall.

^{page 23 note 1} Jacoby, , FGrH. 5 F 6 (i. 155)Google Scholar; Schwarz, , RE 4. 2050–1Google Scholar. We do not know what Greek word Pliny represented by ‘biremem’. See p. 18, n. 3 above for the earliest attested uses δ⋯κροτος and δɩ⋯⍴ης.

^{page 24 note 1} Mingazzini, , Vast delta Collezione Castellani, 1930, 212–15Google Scholar; illustrated in Hoppin, , Handbook of Attic Black-Figured Vases, 1924, 105Google Scholar, and (partially) in Pfuhl, , Malerei und Zeichnung der Griechen, iii, 1933Google Scholar, Fig. 233. The artist is not identifiable (Beazley, , BSA. xxxii, 1934, 3Google Scholar).

^{page 24 note 2} The most recent, and most complete, defence of the ‘three-decker’ view is that by Morrison, J. S., The Mariner's Mirror, xxvii, 1941, 14–44CrossRefGoogle Scholar; W. W. Tarn is the champion of the ‘one-decker’ view (latest in his review of Morrison, C.R. lv, 1941,89–90Google Scholar; earlier references in Gomme, , op. cit. 19Google Scholar and in Morrison).

^{page 24 note 3} Morrison, op. cit., illustrates the warship from BM. B 436 (pi. 6a) and dates the cup ‘early seventh century’ (p. 40). Professor Webster tells me that the shape of BM. B 436 makes it impossible that the cup can be older than 540. BM. B 60 is ascribed by Dohrn, , Die schwarzfigurigen etruskischen Vasen, 1937, 96Google Scholar, to his ‘Siren painter’, whom he dates 525–485.