Silicon (Si) transistor and integrated circuit (IC) technology has grown so big, and become so important, that it is now hard to recognize where, apart from the invention of the transistor itself (Bardeen and Brattain, Dec 16, 1947), it had its origin. In spite of obvious differences in Ge and Si, in 1950–55 it was not evident in many laboratories, concentrating only on Ge, what form of Ge transistor (grown, alloyed, jet‐etched, etc.) might be expected to prevail, with Si not even being considered (or being dismissed outright). What was the need for Si and, at the time, such a seemingly intractable peculiar new technology? The requirement on switching devices of low leakage, and thus the need to leave Ge in favor of Si, led directly in 1954–55 (Bell Telephone Laboratories, BTL) to the exploration of impurity‐diffusion and metallization technology to realize Si transistors and p‐n‐p‐n switches. This technology, a more or less ideal thin‐layer technology that can be referenced from a single surface (and which indeed has proven to be basically invariant and constantly growing), led further to the discovery (1955) of the protective Si oxide, oxide masking and patterning, and the fundamental basis of the integrated circuit (i.e., device‐to‐device interconnection by patterned metallization across the oxide). We recount some of the exploratory diffused‐impurity Si device development of 1954–55 at BTL, particularly the work in and near Moll’s group, that helped to establish the basis for today’s electronics. The Si diffused‐impurity devices of 1954–55 are described, including work and data not previously reported or broadly known—in fact, much work and data (a new technology) that was carried across the Country to a place that became known as Silicon Valley. For further perspective, an appendix is included of independent early suggestions of Bardeen (Urbana notebook, Feb 1952) to leave Ge in favor of diffused Si devices. © 2003 American Institute of Physics