aid — de Vaan
aid 'to say' < *hig-, due to the identity of some pr forms (*agis, agif) at an earlier stage» However, LIV reconstructs *h2eg- for the root of aid, in which case this theory does not work. The alternative is to regard egi as analogical after e.g.facio -feci (2) agmen < *ag-men and ex-amen probably < *ex~ag-(s)men. Ammentum / amentum is problematic next to agmen; EM explain it from *ap-mentum 'attachment'. But maybe … — [de Vaan, s.v. aid, p. 45]