Don --
A lot of the old micrometers had their graduations on a fixed sleeve; adjusting one of these usually required that the thimble be rotated on the spindle, but some had the adjustment at the fixed anvil.
From what I've seen of these, the thimble was sometimes a close sliding fit over an unthreaded spindle end held in place with a set screw, sometimes the thimble was captured between a pair of nuts threaded onto the spindle. In either case, the adjustment required moving the thimble axially until its "lower" edge was lined up with the zero graduation on the sleeve and rotated until the thimble's zero lined up with the sleeve's index line.
There was also one fixed-sleeve micrometer (Scherr-Tumico, IIRC) than had a second sleeve on the thimble. The thimble was actually threaded onto the spindle, and rotating it moved the thimble up and down the fixed sleeve. Once the lower edge of the thimble sleeve was lined up with the main sleeve's zero graduation, the thimble was clamped in place with a lock nut. Finally, the thimble sleeve was rotated to bring its zero to the main sleeve's index line.
Most newer micrometers have a loose sleeve that can be rotated on the frame extension until the zero index line on the sleeve aligns with the zero on the thimble. The rotating sleeves usually, if not always, rely on friction between the sleeve's ID and the frame extension's OD to hold them in place, and there's usually a fair amount of stick-slip when adjusting. The rotating sleeve has a pin hole in its closest-to-frame end to accept the pin of a hook-spanner adjustment wrench. (If the adjustment wrench has two hook-spanner ends, the smaller may fit the ratchet-speeder attachment or the nut that adjusts the fit of the split-nut to the spindle screw while the larger end fits the zero-adjusting sleeve.)
On these micrometers, the spindle and thimble are still separate pieces. This means that the thimble can, in some way, be rotated on the spindle end. This allows the thimble zero to be set to the usual position on the near side of the frame, but in conjunction with a rotating sleeve it allows the "reading position" to be rotated to the top of bottom of the frame as on a crankshaft micrometer.
The Starrett spindles and thimbles are usually fitted on a taper, with a small screw or the ratchet-spinner acting as a drawbar holding the tapers together. The taper is shallow enough to be self-holding. If the thimble needs to be removed from the spindle, loosen the screw a couple of turns then, while holding the thimble firmly, smack the end of the screw with a screwdriver handle to break the taper free.
Some of the Chinese-import micrometers also use a taper-fit between the spindle and the thimble, and these are released in the same way the Starretts are.
The disadvantage to the taper fitting is that there is no simple way to move the thimble axially along the spindle. The other side of that coin is that a properly-fitted spindle should stay properly fitted. The only time I've ever seen it become an issue with a Starrett micrometer is when trying to use a spindle originally fitted to one frame to another frame. With Chinese imports, though, it might be a bigger deal; I've seen these brand new with the edge of the thimble as much as 0.015 inch away from the zero line on the sleeve.
Ok, Don, I've probably over-answered your question, especially since you asked specifically about a Starrett micrometer. I'll hope that someone else can find something useful in my essay.
Did somebody ask what time it is?
John