The evidence for geared mills in Roman Britain is sufficiently slight for the addition of a further item to be worthy of record.

Among the hoard of ironwork discovered in a well at Silchester in 1890 and described by Sir John Evans was an iron bar, 36.1 in. long, tentatively identified as an unfinished carriage axle. Though this explanation was clearly unsatisfactory, no better one was forthcoming and the use of the bar remained a mystery. Whilst it is clearly an axle or spindle of some kind its weight and asymmetry show it was not intended for a carriage and, contrary to Evans's statement, it is a finished object showing signs of wear in places. A recent re-examination has suggested that it was the pivot of a large geared mill (fig. 1a), transmitting the power from a toothed wheel, through the central hole in the stationary lower millstone, to the upper stone. The well-known example from Great Chesterford, Cambs., offers a fairly close parallel and that from the German limes fort at Zugmantel possibly an even better one. The absence of the terminal rind from the Silchester spindle is explained by the fact that it was keyed to the axle by means of the rectangular end on the bar and could thus be easily removed when the spindle was taken as scrap. The end of this keying has been battered over to give greater security in fixing.