Just a guess: Consider the magnetic field induced by the RF signal. I presume it is strongest when along the length of rod. Thus the two ends are the poles of a magnet flipping direction at the RF frequency. Will the field be changing more rapidly at the ends or in the middle?
I remember from the 60s, adjusting a the frequency tracking on an AM transistor radio. Both the local ascillator and the antenna were adjusted to fit the printed frequency dial. High frequency adjustments were made with the trimmer capacitors built into the two gang tunng capacitor. The low frequency was adjusted with the slug in the local oscillator coil, and by moving the antenna tuned winding along the ferrite rod. From that, I would assume that the inductance of the antenna coil depends on its position on the ferrite.