Quote Originally Posted by blitzkrieg80
How can you be certain of this? Has any scholar supported such an assertion? Is this an educated guess based on examples only? How would you apply this to the plentitude of -i endings on Germanic tribes and all others? the same o-class IE nominative plural? Not a single tribe with the typical Latin -i ending has any other root than o-class?
Probably. Certainly not an i-stem, because that gives a plural -es which is attested in Gaulish and Latin. The Romans wouldn't bother to change a plural termination which was already grammatically correct for their own language. Don't know why so many -i and -ii Germans compared to relatively few Britons with the same ending. From a statistician's point of view, they are both rather small samples.