First record a current, then infer an endpoint
Light illuminates an emitting surface while an opposing electrode collects electrons. A retarding voltage progressively suppresses the photocurrent. Repeating the scan at different light frequencies produces a family of curves. The stopping potential is inferred near the current endpoint; an instrument does not display the kinetic energy of each emitted electron.
An endpoint can be obscured by dark current, emission from the receiving surface, finite spectral width and the distribution of initial electron energies. The maximum emission energy is a different observable from either the average energy or the number of collected electrons. A useful test must say which one a plotted point represents.
Read the evidence: R. A. Millikan (1916), A Direct Photoelectric Determination of Planck’s h, Phys. Rev. 7, 355–388 — original scan
What the straight line actually tests
For the ideal one-photon surface balance, define the stopping potential below as a positive retarding magnitude after accounting for contact fields. The elementary charge is positive, the frequency is that of the incident light, and the work function is the minimum energy needed to escape the specified surface. The relation applies above threshold.
A frequency scan tests linearity and a slope given by Planck’s constant divided by the elementary charge. A stable surface can change the intercept without changing this universal slope. A single nonzero photocurrent, or a straight line over too narrow a frequency interval, establishes much less.
Read the evidence: R. A. Millikan (1916), A Direct Photoelectric Determination of Planck’s h, Phys. Rev. 7, 355–388 — original scan
Why two metals introduce a hidden voltage
The emitter and collector generally have different surface properties. Their contact potential changes the retarding energy experienced by an electron, so the external voltmeter reading cannot simply be substituted into the ideal energy equation. The sign of a correction also depends on how the circuit voltage is defined.
Millikan devoted substantial attention to preparing alkali-metal surfaces and measuring or stabilizing the contact electromotive force. His paper reports that the contact potential changed with surface age and absorbed gas. A constant offset mainly shifts an intercept; drift correlated with the frequency sequence can bias the slope as well.
Read the evidence: R. A. Millikan (1916), A Direct Photoelectric Determination of Planck’s h, Phys. Rev. 7, 355–388 — original scanCaltechAUTHORS, Millikan’s 1916 photoelectric paper — institutional record
Intensity and frequency are different interventions
In the weak, single-photon regime at fixed frequency and surface, increasing incident intensity mainly increases the available emission rate. Increasing frequency changes the maximum energy. Detector collection and saturation must be controlled before treating a measured current as proportional to incident intensity.
This is not a claim about every light source or material. Heating, space charge, surface changes and multiphoton emission can alter the comparison. Those effects do not invalidate a test made within its declared regime, but they do prevent the introductory intensity rule from becoming an unrestricted law.
Read the evidence: R. A. Millikan (1916), A Direct Photoelectric Determination of Planck’s h, Phys. Rev. 7, 355–388 — original scan
Agreement with a law is not agreement with every picture
Millikan’s introduction explicitly separates the success of Einstein’s energy relation from his own reservations about localized light corpuscles. The paper is valuable precisely because the investigator could accept the quantitative result while disputing a proposed physical picture. Historical evidence did not arrive as a unanimous interpretation.
The 1921 Einstein award emphasizes the photoelectric law; Millikan’s 1923 award includes both elementary charge and photoelectric work. The stopping-potential slope determines a ratio of constants. Obtaining Planck’s constant from that ratio also needs an independently established charge scale, which connects this article to the oil-drop measurement.
Read the evidence: R. A. Millikan (1916), A Direct Photoelectric Determination of Planck’s h, Phys. Rev. 7, 355–388 — original scanNobel Prize in Physics 1921 — official award recordRobert A. Millikan — official 1923 award facts
Read the Lab as an energy comparison
The photoemission analyzer lets the reader compare photon energy, work function and retarding potential. Use it to predict whether emission is energetically possible and what an ideal energy cutoff means. It does not acquire Millikan’s current–voltage records or infer a historical contact-potential correction.
A sharper reading exercise is to ask which parts of a graph survive an unknown constant voltage offset. The slope may survive; the inferred threshold need not. Then ask what happens if that offset drifts between frequency settings. This keeps the mathematical prediction connected to the measurement that would test it.
Read the evidence: R. A. Millikan (1916), A Direct Photoelectric Determination of Planck’s h, Phys. Rev. 7, 355–388 — original scan