In Quintilian and Festus we read that slaves were, in some antique age, known simply by a derivation from their masters' names; thus Marcipor from Marcus, Lucipor from Lucius, and so on. The practice is also mentioned by Pliny and the grammarians Probus and Priscian. 1 The-por element has been explained as puer, undergoing syncope and passing from an-o-to a consonantal stem in the third declension. It is attached to the genitive of the master's praenomen, the long final syllable of which has, in some cases at least, been shortened. 2 Many have accepted this reconstruction. 3 It has also been noted that Varro bewailed some sort of change in the form or nature of slaves' names, a change which it is easy to identify as the desuetude of these names in-por. 4 As Mommsen observed, names like this are hardly names at all; more like labels. 5 They certainly have a beguilingly archaic flavour that sits well with the legal conception of slaves as property. But the same historian regretted the absence of a useful explanation of their origin and usage. 6 This absence persists today. It will be argued here that this is because there is really no salvaging the standard account of these (in practice, very rare) names, to which scholars have nonetheless generally, if at times uncertainly, clung.