a1 Cholsey, Oxon
This comes near to satisfying; but even with ipsa the change of subject from tecta to plaustra is awkward, and exsultant is inappropriate to a lumbering plaustrum (cf. Virgil, G. 1.163 tardaque Eleusinae matris uoluentia plaustra). I suggest reading cisia instead of ipsa. The cisium was a fast light two-wheeled vehicle which might well jump up on a rough road; and the first three letters cis could have become the -es of the MS exsultantes. Two further points: lapis uiai is not ‘a stone on the road’ (Bailey and Rouse/Smith [Loeb, 1982]), but rather the stone of the road, i.e. the paving; and utrimque is not ‘on one side or the other’ (Bailey in notes) but ‘on both sides’. There remains Ernout's objection that the suppression of the final s of lapis (which stands for lapids*) is unlikely. One can only say that no one would have ventured to introduce by conjecture pendentibu' structas or manantibu' stillent, but both are found in Lucretius' text (6.195, 943).