Summary

  • Ancker-Johnson’s 1964 experiment traced a boundary between stable and unstable electron–hole plasma states in p-type indium antimonide; the boundary changed with field direction, carrier density and sample geometry.
  • The reported hysteresis is evidence of a path-dependent physical system under specific conditions, not proof of a universal memory device or a ready-made theory of control.

The most revealing result in Betsy Ancker-Johnson’s 1964 Physical Review paper was not a single threshold. It was the gap between thresholds. In p-type indium antimonide, or InSb, an injected electron–hole plasma became helically unstable under combinations of longitudinal electric and magnetic fields. But the transition into instability and the transition back out did not occur at the same applied conditions. The route mattered.

Working at Boeing Scientific Research Laboratories, Ancker-Johnson varied the fields with their directions parallel and antiparallel and mapped the stability boundary against plasma density. At the instability threshold, the electric-field hysteresis exceeded 45 volts per centimetre and could amount to half of the applied electric field. From shifts in the loop she inferred an induced magnetic field; the magnetic hysteresis reached as much as 165 oersted, or 55 per cent of the applied field.

Those are unusually legible receipts for a nonlinear system: not metaphors, but measured differences between an upward and a downward passage through the same nominal control space.

The receipts also came with boundaries. The upper end of the magnetic-field interval was typically no more than about 600 gauss, where the plasma density approached zero. The lower end was at least about 280 gauss, associated with magnetic pinching. The largest loops appeared toward the low-field end. A combined electric–magnetic hysteresis range increased as the specimen cross-section decreased, then saturated near a radius of roughly three-hundredths of a centimetre. Material, geometry, density and sweep history were therefore part of the finding, not inconvenient details to strip away.

That distinction becomes clearer when the paper is placed beside Ancker-Johnson’s earlier work. In 1961, with Roger Cohen and Maurice Glicksman, she studied injected plasmas in p-type InSb and connected magnetic pinching with carrier-density changes, negative resistance and oscillation. In 1963, she and James Drummond analysed a thermal pinch driven by heat input and loss, explicitly distinguishing it from a Bennett pinch. These papers were building a vocabulary for several coupled mechanisms. “Pinch,” “depletion,” “instability” and “hysteresis” were not interchangeable names for one effect.

The later record reinforces that caution. A 1971 Physical Review Letters paper by Ancker-Johnson, H. J. Fossum and A. Y. Wong measured linear growth and nonlinear saturation for magnetothermal-pinch oscillations, then tested feedback that suppressed the oscillations and, to a smaller degree, enhanced them. Feedback stabilization was an intervention on an oscillatory mode; it should not be retroactively substituted for the 1964 hysteresis map. Nor does a hysteresis loop by itself establish a useful storage element.

A device claim would require retention, write and erase behaviour, endurance, noise margins, manufacturing tolerances and a credible operating envelope—evidence the 1964 experiment was not designed to supply.

Ancker-Johnson’s professional history matters because the research was neither isolated nor anonymous. Boeing provided the laboratory setting. Cohen, Glicksman and Drummond appear in the experimental and theoretical sequence; Fossum and Wong share the later feedback work. Ancker-Johnson subsequently served as US Assistant Secretary of Commerce for Science and Technology from 1973 to 1977. Archival programmes at the American Institute of Physics and the American Physical Society preserve both her own oral history and the institutional history of women in physics.

Attribution is strongest when it names the lead investigator without erasing the lab and collaborators that made a research programme possible.

The enduring insight is modest but powerful. A system can occupy different states at the same externally reported setting because its internal condition carries the consequences of the route taken. That is what the loop demonstrates. It does not authorize every later claim about semiconductor memory, plasma control or nonlinear automation. Instead, it supplies a disciplined question for any such claim: which variable carries the history, over what range, in what geometry, and with what independent evidence?

Sources