Some notes on the discrepancies between the lateral motions of oscillation tests and full scale manoeuvres