It does not follow that their mathematical knowledge is not a priori unless Mill can rule out the possibility of epistemic overdetermination—i. e., the possibility that their mathematical beliefs are justified both experientially and nonexperientially.
Source: wiktionary