A test of time
In October 1971, two scientists set out from the Washington, D.C. area to test Albert Einstein's theories of relativity. Physicist Joseph C. Hafele and astronomer Richard E. Keating of the U.S. Naval Observatory (USNO) didn't use a rocket or a futuristic laboratory. Instead, they used regularly scheduled commercial airline flights and four caesium-beam atomic clocks. The plan was to fly the clocks around the world, once eastward and once westward, and compare their elapsed time against reference clocks left stationary at the USNO.
The experiment began on October 4, 1971, with the eastward trip. Hafele and Keating, along with their four suitcase-sized Hewlett-Packard 5061A atomic clocks, took their seats on a series of commercial flights. The clocks were so large they required their own tickets, booked under the name "Mr. Clock". After 41.2 hours of flight time, they returned to Washington. A few days later, on October 13, they departed for the westward trip, involving 48.6 hours in the air. The entire project was remarkably inexpensive, costing just $8,000, with $7,600 of that spent on airline tickets.
Nanoseconds of difference
The experiment was designed to measure two competing effects predicted by relativity. The first, kinematic time dilation from special relativity, suggests that a moving clock ticks slower than a stationary one. The second, gravitational time dilation from general relativity, predicts that clocks at higher altitudes, where gravity is slightly weaker, tick faster than clocks at the surface.
The direction of travel was essential. On the eastward flight, with the direction of the Earth's rotation, the plane's speed added to the rotational speed of the planet's surface. This high velocity meant the kinematic (slowing) effect was expected to be greater than the gravitational (speeding up) effect. For the westward flight, against the Earth's rotation, the plane's ground speed partially canceled out the planet's rotational velocity. In this case, the gravitational effect was predicted to dominate.
The predictions were that the eastward-flying clocks should lose time, and the westward-flying clocks should gain time relative to the ground-based clocks. The theoretical calculations predicted the eastward clocks would lose 40 ± 23 nanoseconds. The actual measured result was a loss of 59 ± 10 nanoseconds. For the westward trip, the theory predicted a gain of 275 ± 21 nanoseconds. The clocks returned having gained 273 ± 7 nanoseconds. The results aligned with the predictions, providing direct, empirical confirmation of relativistic time dilation.