The Roman Map of Britain Saravus and Sarviodunum Old Sarum, Wiltshire

Coloneas Aramis var. Aranus (R&C 34) next
Sorbiodoni (AI 486₁₃ Iter xv)
Sorvioduni (AI 483₄ Iter xii)

    The Cosmography's form Saramis is not compounded with dunum 'fort'. The -mi- and variant -nu- are simply four minim strokes and are easily -uu-, Sarauus. This form is found as Pons Saravi Sarrebourge, Moselle, France on the River Sarre; Sarâuus 4th is found in Ausonii Mosella, and Sâra 6th in Venantius Fortunatus (see further Ekwall under Soar ERN 375) . This meshes with the earliest  OE form of Salisbury Searobyrg 552, byrg 'fort'. Also in Staffordshire, compare the early forms of Great and Little Saredon, Sarebrook, and Shareshill which undoubtedly share a common element. In this context Sorvio- should probably be emended to Sarvio- , a combining form.
    We are almost certainly dealing with the old name of the East Avon above Salisbury. Under Severn p.360, Ekwall considers the forest name Savernake SU2266 as potentially derived from the old name for the Bedwyn or the eastern arm of the East Avon. Savernake Forest is recorded as Safernoc 934, Savernac 1156, and Savernak 1275. Ekwall compares this with Severn. If this is so, metathesis  took place under the influence of the name Severn Sabrina.
    Thus, Sorviodunum should be emended to Sarviodunum.

Old Sarum SU1332
Old Sarum National Monument Number 26715 SU13783266
ROMANO-BRITISH SETTLEMENT AT STRATFORD SUB CASTLE 
NATIONAL MONUMENT NO: 26715 (413569, 131946)

    Again, I feel like a dullard. I've omitted mention of R&C 239 Sarna, though I read it as Sarua, the etymology is the same as presented by R&C.

    Coates CVEP p 108-9, while acknowledging Sarva, and apparently unaware of Schnetz's comment on the entry Coloneas, mistakenly gives the AI's Sorvioduni  (Iters XII and XV) more credence. PNRB, perhaps if not so intent on confounding so many names in the Cosmography, might have made the same mention of Schnetz's comment under their Armis (?)  p. 258 as they did under Colonia² p. 313. The simple mention of Schnetz's indicated sAramis would surely have elicited an earlier flash of recognition.
    A river Sarua between Southampton Water and the Tamar, and a Saravus that meets all needs is decisive. 
    I was once convinced that Sarua was an error for Sabrina. I found a credible paeleographic path from Sabrina to Sarua. I faltered in my faith that the cosmographer was orderly.