http://www.historyoftheancientworld.com/2016/01/roman-weaponry-in-the-province-of-britain-from-the-second-century-to-the-fifth-century-ad/
A sensible chap who points out that "The problem with using statistical methods ... is one of applicability." He notes that the varying sizes of many of the weapon finds suggest non-standard production, an interesting observation given the fabricae of the later Empire. He does not conjecture whether these relate to different units, e.g. auxilia from varying backgrounds, but he does present the data as he finds it, which is most useful for the likes of us if and when we try to think of explanations.
very true Patrick, most scholarly! ;)
Quote from: Holly on September 29, 2016, 08:13:15 PM
very true Patrick, most scholarly! ;)
The author, that is, rather than me. ;)