A collection of scraps of paper that I don't really understand completely. These aren't well-drawn and are presented "as-found"

This looks like some sort of lightning detector front-end. The tuned circuit is around 230 kHz. The diode time-constant is way too long, I think. No, that's wrong; it's just biased lightly. I think this was a portable VLF receiver. Not sure what I was contemplating.

Into the trash with it.

Simply no idea. 1v = 1 mr is curious, possibly not related. Oh, wait; I remember the CDV715 chambers are 17 pF. Ah, the source voltage climbs, the op-amp drops, the comparator goes low, discharging the chamber through the .001 uF.

Yes, mR is millirads, not milliradians! OK it's an early version of the resetting ionization chamber. I don't think I ever built this.

I think this was my idea for an astable flip-flop using a gas relay tube. I found a bunch of the 5823 tubes years ago and this looks like an idea for an application, just for the fun of it.
I think that's two unusual flasher circuits. The second one probably lights when power is first applied and the first circuit probably waits a period to start flashing. So, the second one would be better for a wiper kicker, since most people want a quick wipe at the beginning.

The source of the p-channel fet would be on top.

 There's an idea for a Halloween decoration, too.

This is probably a "baseband" lightning detector, detecting the pulses directly. It would probably work best if outdoors. That '3904 comment is wrong. That should be a 2N3906.
I'm guessing a fast amplifier for an RF detector. What else could it be? The DC coupling is odd. The supplies are offset toward the positive side, so maybe I wanted DC coupling for use with a negative detector.

I have no recollection of this, whatsoever! Ah, the joys of getting old.

This might be a preamplifier for a photomultiplier tube. I've had similar thoughts. The top FET provides a low-Z load to the tube and amplifies the voltage without contributing any significant noise current. That lets one use 50 ohm coax without taking a gain hit by loading the tube with a 50 ohm resistor. The actual current is developed across a 680 ohm (assuming a perfect FET)

From a noise analysis standpoint, this is probably equivalent to an inverting amplifier. However, show me a really fast op-amp with virtually no noise current and under 1 nV noise voltage like a U310 FET. I suppose the drain resistor dominates that.

Anyway, the choke seems small. I wonder if I meant 10 mH. Maybe I'm letting the other FET do the low-frequency work.

I suppose it's obviously a Schmitt Trigger since that's what it says. It was to give a clean logic output for the alarm on the lightning detector circuit. It looks like it would work. Never built.
I believe this is an alarm that sounds when a meter reading exceeds a certain voltage, set by the 10k pot. It would work for any DC voltage. The 1 megohm gives it hysteresis. I don't remember much about it, but it did run on two D-cells. The LM393 and LM339 can run on really low voltage, as low as 2 volts, if I recall correctly.
I do know what this one is, but I haven't built it like this. This is a bootstrapped AC probe. The FET, typically a 2N4416, acts as a source-follower driving the output voltage follower (my new one that I use for everything it seems). The trick is that the output voltage is coupled back to the drain to force it to follow the input voltage. So both the source and drain follow the input. The result is that the input capacitance of the circuit is very low and sensitive circuits can be successfully probed. I made one years ago with an ordinary emitter-follower output and it was quite effective.
This looks like a formulaic way to set up a meter driver circuit. The input voltage is offset and amplified to give the full power supply swing and R4 is selected to give a full-scale meter reading. This is a nice way to drive a meter since it isn't possible to peg the meter hard. I think I'd select R4 to go over full-scale by a little with Vcc. It won't hurt most meters to peg them gently.
Looks like a straightforward regulator, actually, probably right out of the application notes. Low level current limit is a good idea for many projects, so it's nice that these types of regulators make that easy. I have a box of the metal NET1928s. I should use this! I've never tried it.
This looks like an interesting single-slope A/D. The FET could be a BS170, the current diode a JFET with about 0.5 mA, and the other parts not very critical. The idea is that a micro pulls the reset line low, allowing C (maybe 1 or 2 uf) to charge. C charges in a linear fashion until it reaches Vin. The higher Vin, the longer it takes, and the more pulses make it to the microprocessor's counter input. A low from the comparator indicates the conversion is complete. I bet this could work pretty well.
This was an idea to make an analog switch turn itself on and off, using the top three switches to make an oscillator. I don't know if it works. (Or why one might want it to.) Just one of those passing ideas.