Full Idea
The traditional characterisation (since Kant) goes: a priori truths are those which can be known independently of any experience - ..but that doesn't mean they MUST be known a priori.
Gist of Idea
A priori truths can be known independently of experience - but they don't have to be
Source
Saul A. Kripke (Naming and Necessity lectures [1970], Lecture 1)
Book Reference
Kripke,Saul: 'Naming and Necessity' [Blackwell 1980], p.34
A Reaction
You may discover through experience that nine matches can't be divided into two equal piles, but Leibniz (and others) say you will only see the necessity of this a priori. No necessity is visible in the matches.