Underived Cognates

It may be useful to collect "underived cognates" between two languages, as I will explain.

For example, the English "Prophet" is known to come from Latin through Old French (Norman). The Latin term, Propheta, is clearly < Greek Prophetes. (I underline the letter E in order to differentiate the used Greek Heta from Epsilon.) The mentioned Latin and Greek words are cognates (nearly the same in meaning and in sound), but, for various historical reasons, we know or assume that the Latin word is historically descends from the original (aboriginal) Greek word. In particular, one can say that PRO[pheta] is < PRO[phetes], but it is more accurate to say that Latin preserves (rather than "derives" from) the Greek PRO. What's the big deal about describing the provenance one way or the other? If Lat. Pro is a descendant, it was learned (borrowed, adopted) by the Latin people; if it is a preserved word, it is the very original word that was spoken by Greeks: Greeks and Latins were ethnically the same. (Indeed, the idea that Greek developed various dialects, like Ionic, Aeolic, etc., is based on the fact that the so-called dialects preserve identical words or parts of words, not "reconstructed" words.) So, to make ethnological studies (if you are interested), start analyzing the words of two languages. Differentiate homophones, cognates, synonyms, and "identicals".