The '289 patent covers a device related to the Selectron: The Computron.

(Shift-click on each above link to allow side-by-side reading of both text and images.)

Electron bombardment of certain materials under certain conditions results in the ejection of the so-called secondary electrons. The ratio of electrons ejected versus electron received can be greater than unity, or less than unity, allowing either a net gain or a net loss of electrons. Of charge. Of potential. The curves of the emission ratio can have both positive and negative slopes resulting in a system which can attain bi-stability of potential. This bi-level characteristic lends itself to the modern "digital" concepts.

While most 21st Century techno-wonks argue whether it was Kilby or Noyce who "invented" the integrated circuit, the '289 patent anticipates both their concepts of "integration". In three-dimensions no less. This device could perform a 14-bit full multiply in 5 microseconds. OK, it would have dissipated several hundred watts requiring a steel envelope because a glass construction would melt. And the prototype was never assembled beyond a 4 by 4 bit multiplier, the concept was, and still is, remarkable.

The names of Richard L. Snyder, Jr., and Jan A. Rajchman should not be left behind just because our technlogy of SiGe stressed super-lattice semiconductors with 70 nanometre line rules is the the latest magic. Some of the ancient magic is still impressive: A true digital integrated circuit existed at the Princeton Laboratories of the Radio Corporation of America in the 1940s: The Computron.