...elements would no doubt have evolved as they did on the Continent, without the features peculiar for special reasons to Rumanian, but possibly closer to Carnic (Celtic) Friulan than to Italian. [I have looked at DB Gregor's book on Friulan, but found nothing distinctly 'Celtic' to the language. --andrew] Its syntax could hardly have owed anything to the totally different Brythonic. It is from the phonetic system of Brythonic that we can dimly see what might have emerged when Latin words began to be used daily on Brythonic lips.
In the first place, the Brythonic borrowings show that Vulgar Latin or the ordinary speech of the middle classes (halfway between slang and the literary language) had phonetic peculiarities of its own in Britain, due partly to the 'scholastic' nature of the Latin taught to the elite which wished to learn it in the class room; and it was from these that Latin filtered down. It was a more old-fashioned form of Latin that would have evolved, with even the old system of syllabic quantity preserved, and V still pronounced as W long after it had become labio-dental V elsewhere; whence Welsh gw as in Gwener (Friday) from (dies) Veneris; gwir (true) from verus, - a change which was effected by the eighth century (later in Cornish and Breton). Secondly, there were idiosyncrasies such as the insertion of W between U and another vowel (e.g. destruo was pronounced destruwo, whence Welsh distrywio 'to destroy', and posuit, 'he placed' was pronounced posuwit (cf. W. clywed) which produced posiit in some inscriptions). Even more remarkable, the same W was inserted between E and O and between E and U. Leo, for example, was pronounced lewo, whence Welsh llew (lion), oleum became olewum W. olew (oil), deus became dewus, W. duw (god; cf. the Gaulish prefix Devo- seen in the name of the Galatian king whom Cicero defended, Deiotarus). Thus, if Brythonic had borrowed the word for boy, puer, Welsh would have had a word like pewyr.
Again, Welsh plwm, from plumbum (lead) and ffydd from fides (faith) show that short stressed U and I did not become respectively O and E, as on the Continent; cf Italian piombo and fede. (Fedes on a coin of the Belgic usurper Carausius, who set up an Empire in Britain in 286-293, perhaps shows his non-British origin). The palatalization of ti, di, etc. also seems not to have taken place in the Latin of Britain; e.g. ratio and diurnus which evolve into ragione and giorno in Italian, become in Welsh rhaid and diwrnod, with unimpaired dental. (The od represents the Latin syllable at generalized into a collective suffix). Similarly, L and N resisted palatalization, i.e. there is no L mouille as in French and the group li affect the preceding vowel instead (e.g. solea 'sole' through solia became seil in Middle Welsh), just as did the group ni (e.g. cuneus 'wedge', through cunius, became cŷn). Nor at first was there voicing of P, T, C intervocally or before L and R, though later it occured in all Welsh words, whether borrowed or not. It is significant too that L after A did not become U as in French (e.g. palf from palma F. paume, and mn remained unassimilated (e.g. colofn from columna; cf. F. colonne, and, this time, also Italian, colonna).
Apart from these specific peculiarities of British Latin, which Celtic orthography enables us to identify, changes in the pronunciation of Latin common to the whole Empire, such as are reflected in the orthography of British inscriptions, must also be taken into account; e.g. E for I (ella), N for Gn (sinum), S for Ns (libes), D for T (capud). There must also have been sounds with which the Britons could not cope; e.g. the group nct as in sanctus holy, where the N was dropped and the C pronounced as a fricative. (Hence W. Sant, C. Sans must be an importation from Italy).
Thus the interaction of a specifically British form of Latin and a specifically Brythonic form of Celtic could have produced a distinct Romance language, as different from French and Italian as they, thanks to the specific qualities of Gaulish and the Latin of Tuscany, were from each other*.
*If we can imagine an earlier equivalent to the Strassburg Oaths, which marked the birth of French, i.e. a Brython and a Saxon swearing an oath of alliance against a third party, as Charles the Bald and Louis the German did against their brother Lothair in 842, a romanized Cadwallon, using the same chancellery-language, might have sworn as follows on allying himself with King Penda of Mercia in 632:
Per amur dy Dew ed per salwament dy pobol cristiane ed dy nus tuts, dy yst di inawant, in cant Dew saber ed poder mi duna, sig eo salwerai yst mew fradre Penda, sig cum hom dy drieff sew fradre salwar debe, in ho che ill altertan face, ed cun Edwyn nul plegit nunche prendrai, che willy mewe a yst mew fradre Penda damnuse sî.
From Celtic: A Comparative Study, by D.B. Gregor, pp. 37-51.
For comparison, here is the actual Strasbourg Oath:
Pro Deo amur et pro christian poblo et nostro commun salvament, d'ist di in avant, in quant Deus savir et podir me dunat, si salvarai eo cist meun fradre Karlo et in aludha et in cadhana cosa, si cum om per dreit son fradra salvar dift, in o quid il mi altresi fazet. Et ab Ludher nul plaid nanquam prindrai qui, menn vol, cist meun fradre Karle in damno sit.