^{page 75 note 1} For example, Kromayer, J., Hermes XXXIII, 1898, p. 68Google Scholar, says that more than half his army was Graeco-Asiatic; Holmes, T. Rice, The Architect of the Roman Empire, p. 147Google Scholar, says ‘two-thirds of the legionaries were Orientals.’

^{page 75 note 2} Grueber, H. A., Coins of the Roman Republic in the British Museum II, 1910, p. 526Google Scholar.

^{page 75 note 3} Op. cit., pp. 17 sqq.

^{page 75 note 4} Grueber, ib., p. 528. Mr. H. Mattingly kindly informed me that this is still so.

^{page 75 note 5} Op. cit., p. 29; followed by Gardthausen, V., Augustus, I, i, p. 354Google Scholar.

^{page 75 note 6} The three taken over from Sextus Pompey; App, . B.C. V, 571, 598Google Scholar.

^{page 75 note 7} 36: Parthian war; wintered in Syria. 35: campaign against Sextus Pompey; wintered in Syria. 34: conquest of Armenia; army stayed in Armenia, and was brought back about November, 33.

^{page 76 note 1} Jos, . Ant. XV, 72Google Scholar.

^{page 76 note 2} If Plut, . Ant. 71Google Scholar be correct, it was still there in 30; but this is so unlikely that Plutarch must be taken to mean Herod's own troops.

^{page 76 note 3} Unfortunately Kromayer, op. cit., also overlooked it, which affects some of his conclusions.

^{page 76 note 4} The source-figures are fully discussed by Kromayer, ib., p. 23.

^{page 76 note 5} For these two legions see App. V, 104, 233–4.

^{page 76 note 6} Jos, . Ant. XIV, 468 [16, 1]Google Scholar, after Sosius took over.

^{page 76 note 7} App, . B.C. V. 320Google Scholar.

^{page 76 note 8} Ib. 321.

^{page 76 note 9} Ib. 320.

^{page 76 note 10} Ib. 208.

^{page 76 note 11} Ib. 213, 215.

^{page 76 note 12} Ib. 213, ἂλλῳ δ' Ἀτωνίου στρατῷ.

^{page 77 note 1} App, . B.C. V, 130Google Scholar.

^{page 77 note 2} Ib. 209.

^{page 77 note 3} Ib. 211.

^{page 77 note 4} Ib. 213–15.

^{page 77 note 5} App. V, 14; Dio XLVIII, 2, 3.

^{page 77 note 6} App. V, 247.

^{page 77 note 7} Ib. 223.

^{page 77 note 8} Ib. 396, called 20,000 men. Plutarch, , Ant. 35Google Scholar, says 2 legions, but this cannot stand against Appian.

^{page 77 note 9} App. V, 243.

^{page 77 note 10} Ib. 13, 14.

^{page 77 note 11} Plut, . Ant. 24Google Scholar; C.I.L. 1, 2nd ed., p. 461; see Gardthausen, V., Augustus II, p. 236Google Scholar.

^{page 77 note 12} App. V, 42; Dio XLVIII, 25, 2 sqq.

^{page 77 note 13} Mon. Ancyr. V, 40–3 (138).

^{page 78 note 1} The number, 19, in Plut, . Ant. 68Google Scholar is the campaign total.

^{page 78 note 2} The eighth was one; Grueber II, p. 583.

^{page 78 note 3} Kromayer, who suggests 4 in Alexandria and 3 in Syria (this has usually been followed since), has again forgotten Macedonia.

^{page 78 note 4} On the epigraphic evidence see Cuntz, O., Jahresh. XXV, 1930, p. 70Google Scholar.

^{page 79 note 1} Plut, . Ant. 50Google Scholar, 51; Kromayer, , op. cit., p. 27Google Scholar. The percentage follows from Plutarch's figures, the Armenian contingent being first deducted from Antony's total force.

^{page 79 note 2} Plut, . Ant. 53Google Scholar.

^{page 79 note 3} Ib. 61.

^{page 79 note 4} Crassus in 53 took 4,000 light-armed to 35,000 legionaries, and Octavian knew that Antony would be tolerably strong in light-armed.

^{page 79 note 5} As Sextus enrolled Italians from Caesar colony at Lampsacus, App. V, 570.

^{page 79 note 6} The exact figure depends on the (unknown) strength of the 7 Macedonian legions. Possibly his aim was 60,000, the same figure as against Parthia.

^{page 79 note 7} Orosius VI, 19, 8 (i.e. Livy).

^{page 80 note 1} See for all this Tarn, , The Battle of Actium, J.R.S. XXI, 1931, p. 173Google Scholar.

^{page 80 note 2} For the invasion of Parthia, Antony, besides legionaries, his own cavalry, and the 16,000 Armenian horse, had 14,000 men to cover his own light armed and the cavalry and light-armed of the client kings (Plut, . Ant. 37)Google Scholar. As the clientkings supplied 6,000 horse to the army of Actium (post), it follows that that army's light-armed could not well exceed 10,000–12,000 at the very outside. Perhaps both sides had about 10,000.

^{page 80 note 3} App. V, 14.

^{page 80 note 4} Only mentioned in the raid on Palmyra, Ib. 37, 38.

^{page 80 note 5} Ib. 208.

^{page 80 note 6} Plut, . Ant. 37Google Scholar.

^{page 80 note 7} Ib. 50.

^{page 80 note 8} Ib. 61.

^{page 80 note 9} Horace, , Epode IX, 17–18Google Scholar.

^{page 80 note 10} I omit the supposed Median cavalry, though appears in Plutarch's, list of auxiliaries (Ant. 61)Google Scholar. Dio XLIX, 44 says that in 33 Antony exchanged some troops with the Median king, but subsequently withdrew his own men without returning his cavalry to the Mede; yet he asks us to believe that, notwithstanding this treachery, the Mede remained faithful to Antony and his interests (LI, 5, 5). As Antony had officially accused Octavian of not giving him the legions promised him at Tarentum in exchange for the ships he had given (Plut, . Ant. 55Google Scholar), Dio's story probably represents an attempt of someone to show that Antony had done much the same thing; it is on a level with the two untrue propaganda stories (charge and countercharge) that Octavian had incited Artavasdes' treachery to Antony, and that Antony had then captured Artavasdes by treachery himself. The methods of Graeco-Roman propaganda, though vigorous, were primitive.

^{page 81 note 1} Horace, loc. cit.

^{apge 81 note 3} Plut, . Ant. 71Google Scholar implies that what horse Antony had there in 30 was his own, not Cleopatra's.

^{page 81 note 3} Assuming 12,000 horse, of which 6,000 were Antony's own (Gauls and Spaniards) and 2,000 were Galatians.