An atom is gone before the question is finished
An atom in a beam crosses the apparatus and is gone. A stronger magnet or a brighter detector cannot give back that lost time. Norman Ramsey found a way to make the time between two brief encounters do the measuring; Hans Dehmelt and Wolfgang Paul developed ways to keep a charged particle available much longer. The 1989 Nobel Prize brought these two approaches together.
The shared idea is easy to picture. To decide whether two clocks run at the same speed, compare them, wait, then compare them again. A tiny difference that was hard to see at the start has had time to accumulate. An atom supplies a remarkably reproducible rhythm, set by the energy difference between two of its internal states.
Read the evidence: NIST, A Brief History of Atomic TimeNobel Prize in Physics 1989 — award record
Ask twice, and let the atom keep time in between
Ramsey used two short regions of oscillating field, with an unperturbed flight between them. The first encounter prepares a combination of the two internal states. Their relative phase, meaning where one rhythm is in its cycle compared with the other, then advances during the flight. The second encounter compares that phase with the driving field and turns the comparison into a measurable fraction of excited atoms.
Turn the driving frequency slowly through the atomic frequency. The detected fraction rises and falls, tracing Ramsey fringes. A slightly faster drive gains phase on the atom during the gap; a slightly slower one falls behind. When the difference amounts to a whole extra cycle, the pattern repeats.
For ideal, instantaneous quarter-turn pulses of the same phase, the excited fraction follows the expression below. The frequency difference is measured in radians per second and the free-flight time in seconds. Doubling that time halves the spacing between neighbouring peaks. A coarse frequency reference identifies which of the repeating peaks belongs to the atomic transition.
Read the evidence: NIST, Sullivan et al., atomic frequency standards: Ramsey interrogation
Keep the particle instead of letting it fly away
A trap answers the time problem in another way. Give an atom an electric charge and electric or magnetic fields can bend its path. Paul made an alternating electric field repeatedly gather an ion back toward the centre. Dehmelt used a Penning trap, combining a steady magnetic field with shaped static electric fields, to hold a single electron. These are different trapping mechanisms, each allowing the same particle to be examined repeatedly.
Keeping a particle nearby also makes careful preparation possible. Motion can be cooled and the response to a probe tested again. The particle still moves, and its surroundings still matter: small field errors, collisions or the probing light can shift the transition being measured. Precision comes from learning those effects well enough to correct them.
Read the evidence: Leibfried et al. (2003), Quantum dynamics of single trapped ions, Rev. Mod. Phys. 75, 281
A sharper question still needs a careful answer
Longer waiting makes a frequency difference easier to notice, provided the atomic phase remains usable. If random disturbances wash that phase out, the bright and dark fringes fade together. If fewer atoms reach the detector, each fringe becomes noisier. The useful waiting time balances phase accumulation against those losses.
This is why an atomic clock needs both a narrow resonance and checks on what moves its centre. Ramsey supplied a powerful way to compare rhythms. Traps supplied prolonged access to particles. Together they let experiments ask finer questions, then return to the same physical system to check the answers.
Read the evidence: NIST, Sullivan et al., atomic frequency standards: Ramsey interrogationLeibfried et al. (2003), Quantum dynamics of single trapped ions, Rev. Mod. Phys. 75, 281Nobel Prize in Physics 1989 — award record