For over fifty years the standard source for the form and meaning of ancient Celtic names has been Holder's Altceltischer Sprachschatz. As usual with such encyclopaedic works, the result has been to endow it in the eyes of non-specialists with an authority sometimes greater than it deserves, fine piece of scholarship though it was in its day; for there is a good deal in Holder which Celticists now know to be mistaken. He made much use of the opinions of Stokes, Rhys, and Arbois de Jubainville; and these are by now often out of date. It is important that historians should realize this. The study of British place-names has made remarkable progress of late years, but the Celtic side of it has not kept pace with the rest; for place-name scholars very rarely have any first-hand knowledge of Celtic historical grammar. Hence even in the best recent works on our place-names the Celtic etymologies are sometimes unsound, because the authors put too much reliance in obsolete or untrustworthy authorities. The present article is a short list of addenda and corrigenda to Holder and other such works, in respect of a few Romano-British names whose Celtic etymologies seem particularly to need revision. The forms taken as the starting points are those given in the Ordnance Survey Map of Roman Britain; or, failing that, in Holder. References to Ptolemy are from C. Müller's edition, Paris, 1883; to the Antonine Itinerary from Otto Cuntz's, Leipzig, 1929 (also to K. Miller, Itineraria Romana, Stuttgart, 1916); to the Peutinger map from Miller's edition ibid.; to the Notitia Dignitatum from O. Seeck's, Berlin, 1876; and to Ravennas from J. Schnetz's, Leipzig, 1930.