• allometry;
  • evolution;
  • flying fishes;
  • ground effect;
  • stability

Flying fish wing area and wing-loading both rise in strongly negative allometric fashion with increasing body length and mass. Evidence is presented to show that this occurs because: (1) the leading edge of the pectoral fin ‘wing’ is fixed at 24% of standard length ( LS) from the snout, (2) the wing length cannot exceed 76% of LS or the tips will interfere with propulsive tail beat and (3) increased mass demands faster flying and wings with better lift : drag ratios; this selects for tapered, higher aspect ratio wing shapes. A consequence of this situation is that larger flying fishes have centres of mass increasingly further behind the centre of wing pressure. Resultant longitudinal instability restricts the maximum size of the two-winged design and the pelvic fins of four-wingers act as a stabilizing tailplane. These data indicate that the accepted model of evolution of flight in flying fishes (by extension of ballistic leaps) is flawed; it is proposed that evolution of lift-supported surface taxiing in half-beaks with enlarged pectoral fins (enhanced by ground effect) was an essential preliminary; subsequent forward migration of the centre of mass to within the wing chord permitted effective gliding.