The next “a-hah!” moment for me in understanding geophysics came in 1965 with a paper by J. Tuzo Wilson, “A new class of faults and their bearing on continental drift.” At the time it seemed a light had gone off in my head, even though by then I was deeply enmeshed in ice fog studies.

The red lines are spreading centers; the blue line is a transform fault which has its origin in the offset between portions of the spreading center. The part of the transform fault between the offset segments is active; the plates are sliding past each other. The parts to either side are "fossil"--although the offsets are still visible, there is no movement.

Up to that point, it was assumed that if a feature—such as a stream, a mountain range, or a magnetic stripe—was displaced by a fault, it was evidence that they had once been continuous, and movement along the fault had separated them. This paper suggested the existence of a new kind of fault, now called a transform fault, where the apparent displacement was due to an offset in a spreading center. Spreading centers were themselves new, though grabens (places where the crust on either side was moving apart, resulting in a downdropped block in the middle) had previously been recognized.

Suddenly the apparent faults in the Pacific Ocean floor, faults that stopped at the coast, made sense. The rocks on either side of the transform fault did not move relative to each other. The only motion was between the offsets on the spreading center.

We now know that plates are bounded by three kinds of faults:

Spreading centers, where plates move apart,

Converging boundaries, where plates collide,

Transform boundaries, which often connect the two, and allow plates to slide past each other.

The familiar faults such as the San Andreas are the active portions of transform faults, but the offset magnetic stripes in the Pacific are generally the “fossil” parts of transform faults. Looked at this way, the “end” of an active fault makes sense: it is where the transform fault joins another type of fault to make part of a plate boundary.

The full plate tectonics hypothesis had not yet been fully formulated, but the essential pieces had been found.