Attested: AI iter 2 Blatobulgio 12 Roman miles from Castra Exploratorum (probably Netherby), which was 12 miles from Luguvallo (Carlisle).
Where: The confident assertion that *Blatobulgium was the Roman fort at Birrens at NY21907518, near Middlebie, Dumfriesshire, appears to be based on Kenneth Jackson's linguistic analysis shown below. Unfortunately: (1) that is not the only (or necessarily the best) possible etymology; (2) the name Stodoion has a stronger claim to fit the prominent granaries at Birrens; and (3) there is no sign of a direct Roman road between Netherby and Birrens, which one might expect for an AI route, although they are the right distance apart. Actually, identifying Blatobulgium with the Roman fort at Broomholm beside the river Esk at NY37868145 in Dumfriesshire is as good linguistically, and helps to locate other names in the region.
Name Origin: Jackson's Celtic parallels, Welsh blawd ‘flour’ and Irish bolg ‘bag, satchel’, are not the only, or even the best, available etymology. For the Blato- part several PIE roots are in the running, including precursors of blade (to fit the shape of a river confluence) or blather (to fit a demonstrative military outpost) but perhaps best is Latin blatteus used for the colour of dried blood or a cockroach. Obviously -bulgium resembles Latin Latin bulga ‘purse’ (also womb and other body cavities in common speech) but the notion that this word originated in Celtic and passed from there into Latin rests on a single throwaway line by one Roman author, referring to Galli. It is more likely to have passed from PIE *bhelgh- ‘to swell’ into multiple language families. Several Germanic languages (including Gothic and Old Frisian) have that ‘purse’ sense, but most of them (plus Gaelic) have a word related to OE balca, from which comes modern balk, meaning ‘ridge’, which seems to be why Smith (1956:18) perceived *balg- as ‘rounded hill’ in later English place names.
Notes: Heather-covered moorland is emblematic of the Scottish borders, and the photo below of Breckeny Knowe, kindly provided by Walter Baxter, shows a particularly fine example. It was taken in October, when the heather blooms have faded, looking north from the Broomholm fort.
To fit Broomholm to AI's mileage figure, either xii must be emended to vii, or the Roman route (route 868 of Margary, 1973) must have taken a wide dogleg to the east.
Last Edited: 5 September 2017