A carefully quiet receiver still heard the sky
In 1964 and 1965, Arno Penzias and Robert Wilson were working with a large horn-shaped radio antenna at Bell Laboratories in Holmdel, New Jersey. They wanted to account for the faint noise entering a sensitive receiver. Even after estimating the known contributions, an excess remained. It came from different directions and persisted through the seasons. Their 1965 report put it at an antenna temperature of 3.5 kelvin, with an uncertainty of 1.0 kelvin. A leftover in a careful radio measurement had become evidence of radiation filling the universe.
The antenna had a practical advantage: it mostly received radio waves from the direction in which it pointed, with little unwanted response behind it. That helped reduce contamination from the warm ground. The measurement used a frequency of 4.08 gigahertz, corresponding to waves about 7.35 centimetres long. These are radio microwaves, far longer than visible light. The receiver measured the background as radio noise spread across the sky.
Sources: [1]
A temperature here described the strength of radio noise
A receiver hears thermal radiation from many places. Its own components emit; the antenna loses energy and emits; the atmosphere contributes too. Radio astronomers often describe that noise by an equivalent temperature: how warm a thermal source would need to be to supply the same noise power in the observing band. This “antenna temperature” turns several faint contributions into quantities that can be compared on one scale.
Penzias and Wilson compared the antenna with a reference cooled by liquid helium. Pointing the horn at different elevations helped estimate the atmosphere, because a low view passes through more air. They examined electrical losses in the metal structure and tested whether unwanted ground radiation could enter from behind. They even taped seams in the horn with aluminium tape to check for an effect. The measured temperature stayed the same.
The original paper reports a total zenith antenna temperature of 6.7 kelvin, with an estimated 2.3 kelvin from the atmosphere and 0.9 kelvin from antenna losses and backward response. The remaining approximately 3.5 kelvin was the puzzle. Within their observations, it had no preferred direction or polarization and no seasonal variation between July 1964 and April 1965. A local contaminant would have needed to imitate all of those properties.
Sources: [1]
Nearby, another team was asking what a hot universe would leave behind
At Princeton, Robert Dicke, James Peebles, Peter Roll and David Wilkinson were investigating a cosmological possibility. If the universe had once been hot and dense, matter and radiation would have exchanged energy frequently. Radiation from that early state should still fill space. Cosmic expansion would stretch its wavelengths and lower its temperature, leaving a faint microwave glow.
The important connection was between a measured excess and a physical expectation. The Bell Laboratories team had a stubborn signal and a careful description of its instrument. The Princeton team had an explanation for why a widespread background should exist. Their papers appeared together in the July 1965 Astrophysical Journal: the interpretation first, followed immediately by the antenna measurement. The discovery did not depend on naming an exotic object in one direction. Its striking property was that the signal kept appearing wherever the horn looked.
The thermal explanation also predicted something the single-frequency result could not yet determine: a particular spectrum, the way brightness varies across frequencies. That prediction gave later experiments a clear task. They could ask whether measurements at many wavelengths lay on the same thermal curve, with one temperature describing them all.
The excess became a new way to look into the past
Subsequent measurements established a background with a temperature near 2.7 kelvin and an extraordinarily close blackbody spectrum. A blackbody has the thermal spectrum of matter and radiation in equilibrium. It supplies a much more specific signature than a faint microwave level at one frequency. Later, very small temperature differences across the sky opened another question: how an almost uniform early universe developed galaxies and clusters.
The 1978 Nobel Prize recognized Penzias and Wilson for discovering the cosmic microwave background. Their original 3.5-kelvin antenna excess should be read with its uncertainty and observing frequency; it is not a modern high-precision measurement of the thermodynamic temperature. What endured was the experimental pattern: an all-sky excess survived checks of the receiver, atmosphere and antenna, and a hot early universe supplied a testable explanation for it.