If you connect a dual-beam 'scope with its common to the neutral, channel 1 to one hot leg and channel 2 to the other hot leg, then you will see two sinewaves of equal amplitude but 180 deg. out of phase (i.e. the zero crossings will coincide but the positive peak of one will coincide with the negative peak of the other).
A similar center-tapped arrangement is common for obtaining DC supplies in much electronic equipment, because full-wave rectification can then be achieved with just two diodes.
Connecting the scope common to one hot leg with channel 1 on the neutral and channel 2 on the other hot changes the reference point. In this case you would see two sinewaves in phase, but ch. 2 with twice the amplitude of ch. 1.
Notes if anyone tries these experiments:
1. Many 'scopes have the chassis and common point grounded via the cord. Obviously, connecting the common of such a unit to a hot leg will result in a flash and a tripped breaker. Use an isolated scope or a separate isolation xfmr, and remember that the case of the scope could end up at 120V to ground, so take appropriate precautions.
2. Modern dual-beam scopes usually have different trigger modes for the horizontal timebase. To get a true comparison of the phase of the two inputs, make sure the triggering is set to one channel only, otherwise it will synchronize to each sinewave individually and you won't get a true phase comparison.