On this page
Most great discoveries are planned. This one began as a nuisance: a faint, steady hiss in a radio receiver that would not go away. Arno Penzias and Robert Wilson spent about a year trying to get rid of it before they learned it was a signal from the early universe. This post tells how they worked, does the arithmetic behind the discovery, and draws out the habits that made it possible.
Who were Penzias and Wilson?
The Nobel facts page for Arno Penzias gives his dates as 26 April 1933, Munich, to 22 January 2024, San Francisco. His Nobel autobiography recalls his family being rounded up for deportation when he was a small child, then escaping: at six he and his younger brother were put on a train to England, and the family reached New York in January 1940.
Robert Wilson was born on 10 January 1936 in Houston. In his Nobel autobiography he says he fixed radios and television sets for fun and spending money as a boy, then studied physics at Rice University and Caltech.
Both worked at Bell Laboratories in Holmdel, New Jersey. They shared half of the 1978 Nobel Prize in Physics "for their discovery of cosmic microwave background radiation".
What were they actually trying to measure?
In his Nobel lecture, Wilson explains that Bell Labs had a 20-foot horn-reflector antenna, built in 1960 to receive signals bounced off the Echo satellite. Its horn shape shielded the receiver from the ground, which would add less than 0.05 K, which made it unusually good at measuring faint, smooth backgrounds. In 1963 it was fitted with a low-noise maser receiver working at 7.35 cm, and Penzias and Wilson planned to use it for radio astronomy.
To measure carefully, they compared the antenna with a "cold load", a reference cooled with liquid helium. The antenna should have read only the sum of the atmosphere (2.3 K) and its own walls and the ground (1 K). Instead it read about 7.5 K, well above the expected 3.3 K.
How did they rule out every ordinary cause?
Wilson's lecture lists their checks:
The atmosphere: their own sky measurements ruled out a larger value.
Man-made noise: pointing at New York City made no difference.
Our galaxy and radio sources: they predicted only about 0.02 K in the coldest part of the sky.
The antenna itself: they calculated losses, and found "a pair of pigeons was roosting up in the small part of the horn". They evicted the pigeons and cleaned up "a white material familiar to all city dwellers", but the excess fell only a little. Taping over the riveted joints barely helped.
Changes over time: after almost a year the excess had not changed, ruling out anything in the solar system or left over from a 1962 high-altitude nuclear test.
Then Penzias mentioned the noise to Bernard Burke of MIT, who had heard of work by P. J. E. Peebles in Robert Dicke's group at Princeton predicting a leftover glow from a hot early universe. The two groups published side-by-side letters. Penzias and Wilson's was titled, with careful modesty, A Measurement of Excess Antenna Temperature at 4080 Mc/s. Wilson says they left out the cosmology "because we had not been involved in any of that work" and thought "our measurement was independent of the theory and might outlive it."
Worked example 1: frequency and wavelength
Their receiver worked at 4080 Mc/s, which is 4.080 × 10⁹ Hz.
λ = c / f = 2.998 × 10⁸ / 4.080 × 10⁹ = 0.0735 m = 7.35 cm.
This matches the 7.35 cm maser in Wilson's lecture and the "about 7 centimeters" on the Nobel facts page.
Worked example 2: where does this radiation peak?
The excess they found was 3.4 ± 1 K. Later measurements pinned it down: a 2009 review by D. J. Fixsen gives 2.72548 ± 0.00057 K.
Use Wien's displacement law, λₘT = b. The NCERT Class 11 chapter on thermal properties of matter gives b ≈ 2.9 × 10⁻³ m K; NIST gives 2.897771955 × 10⁻³ m K.
λₘ = 2.8978 × 10⁻³ / 2.72548 = 1.063 × 10⁻³ m, about 1.06 mm.
So the background peaks at millimetre wavelengths. At 7.35 cm, Penzias and Wilson were measuring its long-wavelength tail, roughly 70 times longer than the peak.
Which mistakes does this help you avoid?
Confusing the peak with the measurement. The background peaks near 1 mm; the discovery was made at 7.35 cm.
Dropping the powers of ten in c/f. Write 4080 MHz as 4.080 × 10⁹ Hz before dividing.
Assuming noise is always error. It was a real signal; their job was to prove it was not the instrument.
The JEE (Main) 2026 syllabus covers electromagnetic waves and their spectrum, and Wien's law appears in the NCERT Class 11 book. Check the syllabus for your exam year. Our post on George Gamow and the Big Bang gives the theory side of this story.
When a measurement and a formula refuse to agree in your own practice, Fermi's AI tutor can notice where you are stuck and check your understanding as it goes, instead of handing over an answer.
What can a student take from Penzias and Wilson?
List every possible cause, then test each one. Their checklist is a model for debugging any experiment.
Do not hide what you cannot explain. They lived with the "antenna problem" for a year rather than adjusting it away.
Report what you measured, not what you hope. Their title said "excess antenna temperature", nothing more.
Next step: use Wien's law to find the peak wavelength for a body at 300 K, and compare it with the cosmic background.
