Each section has a pair of brass leaf springs applying just enough friction to hold the next-smallest tube in place, with a rolled crimp securing the springs and preventing the smaller section from pulling out. My first version used that short length of the largest section and the next (for Mary’s helmet) used only the two smallest tubes; it’s rapid prototyping at its finest, except that I rarely discard a prototype that actually works.
Late last year I managed to pull the shaft out of the base while adjusting the length and watched those two springs flutter to the ground beside me.
After finding both of them amid the usual roadside clutter, I swore a mighty oath that I’d epoxy the base of the middle tube into the larger one, eliminating one non-functional adjustment point:
The heatstink tubing covers most of the evidence, but you can see a fillet of epoxy around the end.