¹ The sherd has been examined by Mr C. J. Young, who recognises it as of a normal coarse Romano-British ware.

² Catalogue of the Roman pottery in the British Museum (1908), p. 436, item M 2838.

³ EE ix 1003a. He considered that the broken letters here cited as line I were faint and unconnected with the present text.

⁴ At Amerden in, or about, 1891 a Late Celtic sword and scabbard were dredged up from the River Thames, and in 1893 presented to the British Museum (V.CH Bucks i, 185–6). At Taplow, one mile north upstream, some Bronze Age weapons ‘of an unusually interesting character’ were recovered (ibid., 183) and in, or shortly before, 1903, in a creek near Taplow a Bronze Age spearhead was found (Read, , PSA² xix (1903), 287–9Google Scholar, VCH Bucks i, 184). When clearance work was undertaken in 1853 near the river-crossing beside the churchyard of the old church of Taplow, demolished in 1827, large quantities of pottery, early British, Roman and Saxon were discovered. In 1883, when the large barrow at the west end of that churchyard was excavated, a rich deposit of Anglo-Saxon date was recovered, and near ground-level ‘at the bottom of the excavation occurred a fragment of ‘Samian’ ware and part of a brick, both undoubtedly of Roman origin’. (PSA² x (1883), 19Google Scholar; VCH Bucks i, 199–200, drawn partly from the excavator's unpublished report.)

⁵ In seeking to interpret the name or word µμλοøμμικοσ one saw that Liddell-Scott-Jones (1940) s.v. øνμικοσ gave ‘army-surgeon’ as a conjectured meaning for øγεικοε on an inscription (Inscr. Graecae ed. minor i, 950, 1. 153) which records Athenian naval casualties from the battle, probably of Cynossema, in 411 B.C; this interpretation is accepted without question by D. Bradeen, W. (Classical Quarterly n.s. xix (1969), 147Google Scholar, n. 10). In this text trierarch occurs twice and phylarch once, as ranks after a name, but øμσικοσ comes at the beginning of the line followed by a personal name. Dr P. J. Rhodes has been unable to match Øυσικοσ as a personal name except for this sole example in Kirchner, Prosopographia Attica ii, but considers that it is more probably an unmatched personal name than an occupational description for ‘surgeon'. In the usage of Aristotle, half a century later, it refers to ‘nature’ or ‘natural philosophy’. Dr J. D. Thomas made a similar search in papyrological lexica and indices and found no example of Physikos as a personal name or as the second part of a compound.

⁶ In Liddell-Scott-Jones (1940), and Lampe, A Patristic Greek Lexicon (1961), or in Dornseiff-Hansen, Ruckläufiges Wörterbuch der griechischen Eigennamen (1957).

⁷ For occupations in compound words see from Pompeii before A.D. 79 (CIL iv, 3483) lupinipolus ‘lupinseed-seller’ (expressed as lupinarius in item 3423 by the same slogan-writer), and ibid. 5182 sortilogus ‘soothsayer’ (for sortilegus). For i instead of y see Nimfa (for Nympha), (RIB 1526, Carrawburgh Brocolitia; ILS 4726).

⁸ Edict. Dioclet. 7, 20. Literary evidence for the use of mules in the transport service seems to be scanty until the fourth century, but it can be assumed that they were used from the start of the cursus publicus. Pairs of mules are displayed on monuments, often tombstones. In 370 Valentinian I and Valens ruled that the muleteers, coachmen and mulomedici assigned to the cursus publicus should receive no reward beyond their maintenance and clothing which were officially provided (Cod. Theod. VIII, 5, 31).

⁹ e.g. RIB 808 (Maryport, , Uxellodunum), JRS lix (1969), 235Google Scholar, No. 3 (Chester, Deva).

¹⁰ Grateful acknowledgement is made to Dr P. J. Rhodes and Dr J. D. Thomas for their specialist advice and to Mr K. S. Painter for making the sherd available in the British Museum, and to Mr R. Wilkins for his photograph which is here reproduced by courtesy of the Trustees of the British Museum.