L. J. D. Richardson'S article ΥΠΗΡΕΤΗΣ, CQ xxxvii (1943) 55–61, asserted the manifest truth that the word hyperetes, for which ‘an unattested original meaning “under-rower” has been universally assumed, had ceased to be used literally by the time our records, literary and epigraphic, begin’. ‘When ὑπηρέτης and its derivative ὑπηρετέω first appear (in Herodotus saepe), both the particular transferred sense “servant” and the generally transferred sense “subordinate” were already well established to the exclusion of the original meaning (whatever it was), and, what is more, the metaphor from the seafaring usage seems to be already “dead”. The essential note, however, in the group of words (ὑπηρὲτης, ὑπηρετέω, ὑπηρεσία, ὑπηρετικός) is not hard to assess: it is implicit, unquestioning service in response to another's authoritative bidding.’ The truth of this assessment for the fourth as well as the fifth century is plain from an examination of the very frequent use of the word in, e.g., Plato. Richardson proceeded to point out ‘a source of confusion in the dictionaries’ (which still prevails), viz., that of the four chief members of the group of words mentioned, hyperesia alone, having become ‘figurative (as service) in all areas, was then reapplied in the naval domain’, as a name for part of the ship's company of the fifth- and fourth-century trireme. This pattern (of reapplication in a naval context) is in fact also true of hyperesion and hyperetikos. The former is the name given by Thucydides and Isocrates to the oarsman's cushion, called proskephalaion by Hermippus and Theophrastus, while the latter is used of a naval dispatch boat.