The results from the 1910s—the perihelion advancement of Mercury and the behavior of starlight near the Sun—were important early gauges of relativity’s overall viability—akin to making sure a patient has a reasonable heart rate and blood pressure. Another critical test, the existence of gravitational redshifts, similarly checked out fine. But by the 1950s many researchers expressed dismay that no more tests were available. For example, at a 1957 conference, physicist Bryce DeWitt threw a piece of chalk up in the air, caught it, and then remarked (slightly exaggerating): “We know almost nothing about gravitation. There is only one experiment which we do over and over again, and that is what I have just done.” Fortunately, a bevy of new probes now offer Einstein’s body of work an even more extensive physical examination. In assorted experiments, precise equipment has been scanning it from head to toe, seeking signs of even the slightest flaw.
Providing the very legs on which relativity stands, the equivalence principle must remain solid enough to support the theory. Accurate measurements of the equality of inertial and gravitational mass offer vitally important data. If they were to indicate even the slightest discrepancy, the implications would be monumental. Modifying Einstein’s theory would become a necessity, not just speculation.
One device for testing equivalence, called a torsion balance, dates further back than general relativity itself yet continues to be updated and refined. Torsion means twisting or turning. Through a balance device, such actions can reveal how forces affect materials. At the turn of the 20th century, Baron Roland von Eötvös of Hungary used such a sensitive instrument—a weight hanging from a rotating rod delicately balanced on a pivot—to measure minute differences in the accelerations of various substances. He devised it to record any subtle effects produced by small discrepancies between inertial mass and gravitational mass. Thanks to its meticulous design, the equipment was precise enough to rule out such a difference down to one part per hundred million.