A glowing beam behind the cathode
In a low-pressure gas discharge, positively charged particles accelerate towards the cathode. Small holes in that electrode let some particles pass through and form a beam behind it: the canal rays. Stark studied the light from this moving gas. In hydrogen in 1905, he observed a spectral displacement that changed with the viewing direction.
The 1919 Nobel presentation speech describes a broad component beside an ordinary hydrogen line. It moved towards shorter wavelengths when the emitters approached the observer and towards longer wavelengths when they receded. The reversal gave the motion a recognizable signature in the spectrum.
Read the evidence: J. Stark (1919), Structural and Spectral Changes of Chemical Atoms — Nobel lecture and 1919 presentation speech, Columbia-hosted volume, PDF pp. 160–172
Think of a passing train whistle
When a train approaches, successive sound waves reach your ear more quickly and its whistle sounds higher. As it recedes, the pitch falls. Light has a Doppler effect too. An approaching atom sends light with a higher observed frequency and a shorter wavelength; a receding atom gives a longer wavelength.
The equation applies when the speed is much smaller than the speed of light. The reference wavelength is for the same transition in the emitter’s rest frame. The line-of-sight velocity is positive for motion away from the observer, and c is the speed of light. Looking from opposite ends of a beam reverses the first-order shift. Looking from the side removes that first-order line-of-sight contribution.
Read the evidence: J. Stark (1919), Structural and Spectral Changes of Chemical Atoms — Nobel lecture and 1919 presentation speech, Columbia-hosted volume, PDF pp. 160–172NIST Atomic Spectroscopy, Spectral Line Shapes, Widths, and Shifts — Doppler and pressure broadening
Which particles actually emitted the light?
The result left a second question. A particle accelerated as an ion could collide with gas or exchange an electron before it emitted light. A canal-ray beam could contain ions alongside fast neutral atoms. The wavelength shift identified moving emitters; identifying their charge required additional evidence.
Stark discussed the complications from neutral rays in his Nobel lecture. In the modern assignment, the visible hydrogen Balmer lines come from neutral hydrogen, which still has one electron. Transitions of that electron produce the colours. A bare proton has none of these electronic levels. That distinction matters when reading the historical argument about charged emitters.
Read the evidence: J. Stark (1919), Structural and Spectral Changes of Chemical Atoms — Nobel lecture and 1919 presentation speech, Columbia-hosted volume, PDF pp. 160–172NIST Handbook, Strong Lines of Hydrogen — H I visible Balmer wavelengths
Why a beam makes a broad line
Atoms in a beam have a range of speeds. Their different Doppler shifts add up to a broad line. Directed motion shifts the centre; random thermal motion adds width. Collisions and the spectrometer add their own broadening. Extracting a temperature requires a description of those contributions.
This explains the value of changing the viewing direction, keeping an unshifted reference and calibrating the instrument. A side view removes the first-order line-of-sight shift, but other broadening and higher-order motion effects remain.
Stark’s later electric-field splitting concerned a change inside the atom. The Doppler shift concerned motion between emitter and observer. Both could occur in a discharge tube. Changing the field and the viewing direction separately offered ways to tell their effects apart.
Read the evidence: NIST Atomic Spectroscopy, Spectral Line Shapes, Widths, and Shifts — Doppler and pressure broadeningJ. Stark (1919), Structural and Spectral Changes of Chemical Atoms — Nobel lecture and 1919 presentation speech, Columbia-hosted volume, PDF pp. 160–172