There was a statement in an old Aeromodeller Annual (1949?) to the effect that a spinner 1/5 the diameter of the prop was aerodynamically most efficient. So, 2 in spinner for 10 in prop.