For nearly all of the 1930'ies, the fastest aeroplanes were all of the seaplane category (as opposed to landplanes, to use the terminology of FAI).
Presumably, the designers of the two categories had access to more or less the same technology (engines, materials, aerodynamics, etc.) given the large interest in the subject and funding available.
If we for a second rely on Wikipedia for the record listing, we see that the speed record for a seaplane set in 1931 is 655.8 km/h (Supermarine S.6B), while the record for a landplane set in 1935 (4 years later) is "only" 567.1 km/h (Hughes H-1 Racer).
Seeing how fast aviation technology developed in that period, it is remarkable that four years after the Supermarine record, the fastest landplane was still significantly slower than the fastest seaplane.
I would expect the large pontoons on the seaplanes to provide more drag compared to a simple landing gear, and hence by logic the seaplane should be penalised and achieve lower maximum speeds. All things being equal.
However, since the speed differences are so pronounced, all things obviously are not equal. So what are the differences between the two categories (or their designers and manufacturers) that can explain the gap in maximum speed?