^{page 204 note 1} Cassius Dio lx. cc. 19–21 (dealing with the Plautian invasion); lxii. cc. 1–12 (Boudicca's revolt).

^{page 205 note 1} Cass. Dio lxii. 2.

^{page 205 note 2} Ib. c 6.

^{page 205 note 3} Id. c. 7.

^{page 205 note 4} Tacitus ^Ann. xiv. 32. 2 (he is obviously borrowing from the same source as Dio).

^{page 205 note 5} Suetonius viii. 3. 4 (though this may be from an epigraphical source).

^{page 205 note 6} Cassius Dio Ix. 21 (contrast Suetonius v. 17).

^{page 205 note 7} 1b. .

^{page 205 note 8} Cf. also lxii. 1.

^{page 205 note 9} But this can probably be inferred from Suetonius v. 17.

^{page 206 note 1} Tacitus Annals, vol. ii. pp. 133 sqq. (Mr. B. W. Henderson, in his paper on ‘The Roman legions in Britain,’ has anticipated certain of my arguments: I had not had the pleasure of seeing Mr. Henderson's paper [Eng. Hist. Rev. 1903] when I worked out this question.)

^{page 207 note 1} See Holmes, Rice, Ancient Britain and the Invasions of Julius Caesar, pp. 334–335.Google Scholar

^{page 207 note 2} lx. 21.

^{page 207 note 3} C.I.L. vii. 11. Hübner (loc. cit.) inclines to regard the Cogidubnus of the Goodwood inscription as the son of the client-prince of Tacitus, relying apparently on the Flavian date of the monument. This seems unnecessary. (1) Cogidubnus seems to have lived on into Flavian times: Tacitus says ‘is fidelissimus ad nostrum memoriam remansit’; i.e. he probably lived on into Agricola's governorship. (2) Why bestow extraordinary titles on the son when the need for conciliation was far less than before? (3) The names ‘ Tiberius Claudius’ surely point back to the Claudian period.

^{page 207 note 4} Agricola c. 14.

^{page 208 note 1} Colchester itself seems to have been stormed (Cassius Dio lx. 21 and Suet. v. 21). Some writers place several battles between the crossing of Thames and the taking of Colchester; but Dio (loc. cit.) distinctly puts the salutations as Imperator after the capture of the city.

^{page 208 note 2} Tac. Ann. xiv. 33. 1–4.

^{page 208 note 3} Furneaux vol. ii. p. 145, note 2.