Replace one argumentative reply with a repeatable test and a written result.

“I once purpos'd answering the abbé, and actually began the answer; but, on consideration that my writings contained a description of experiments which anyone might repeat and verify, and if not to be verifi'd, could not be defended; or of observations offer'd as conjectures, and not delivered dogmatically, therefore not laying me under any obligation to defend them; and reflecting that a dispute between two persons, writing in different languages, might be lengthened greatly by mistranslations, and thence misconceptions of one another's meaning, much of one of the abbé's letters being founded on an error in the translation, I concluded to let my papers shift for themselves, believing it was better to spend what time I could spare from public business in making new experiments, than in disputing about those already made.”