It turns out that, if floating point numbers are always stored in a normalized form (AKA scientific notation), then the leading bit (the one on the left of the binary point) is always a 1. So, why store it at all? It's considered a hidden bit since it's never stored, just implied.
