[AndrewC]

The resulting diode, approximately 12 atoms wide, could be the first in a move toward smaller, simpler, and thus faster transistors—which are made by placing two diodes back to back.

Yeah, right. So place two diodes back to back and see what you get. No amount of wishing is going to make it a transistor despite the incomplete use of words in a "technical" journal such as the University of Chicago magazine. It is a simple demonstration. Go to radio shack. Buy 2 1N4001 diodes. Twist their cathodes or anodes together(or any d*** way you please). Now try to make the contraption amplify a current like a transistor. Good luck.

[Ichneumon]

It is a simple demonstration. Go to radio shack. Buy 2 1N4001 diodes. Twist their cathodes or anodes together(or any d*** way you please). Now try to make the contraption amplify a current like a transistor. Good luck.

Why are you dishonestly trying to change the subject to "twisting" leads together, when you know full well that the discussion concerns mating the semiconductor layers?

In any case, does this mean that you're now retracting your original claim that, "Which, of course, makes my original statement, that a transistor was in a sense two diodes back-to-back, entirely correct"?

[Nataku X]

AndrewC, if the diodes in the patents are not also the transistors, then why does the detailed description of the patent refer to transistors Q1-Q8? Where are Q6,Q7,Q8 in the diagram?