Further to….
I have done some photo-faffing (saved it, then lightened and enlarged the image in MS 'Photo Editor' ).
The results do make it seem my first suggestion – a heavy-duty jenny – was indeed wrong. I had arrived at that by comparison to my own jenny, and to similar machines from its manufacturers.
There are no names or other identifiers on the parts visible to us.
The long operating-handle is connected to the shafts by a pinion and gear giving a 3 or 4 to 1 mechanical advantage, so clearly this thing needed considerable forces at the business end.
The two main gears appear nearly out of mesh but that may be a result of heavy wear-and-tear; or more charitably happened when the rollers were opened to release the work-piece.
The small T-handle in the middle appears to enter a U-bracket, but what it adjusts or did otherwise is not at all obvious.
The small pulley-shaped wheel most likely IS a pulley; an idler on a small stub-axle in its own casting screwed to the main frame.
So the central adjuster probably set something allied to that pulley.
–
The top T-handle is almost certainly a gap or pressure adjuster between the rolls.
The top roll is held by what looks like a circular nut, but if that was manipulated by a pin-spanner the holes are not visible. It may simply be a large washer, with what could be a taper-pin just visible in the shaft outboard of it.
The lower roll's nut might have a locking cotter or split-pin but excrescence suggesting that, is not definitive enough. It might just be a particle of the gloom and muck on the rest of the machine.
It is strange that the rolls are so different in width, which was what made me wonder if they are not the matched pair they would have been in normal use. Alternatively they matched the profile of whatever was passed between them.
The two main gears are of different diameters but that may be to equalise the rolls, for the lower does look a little larger then the upper.
There appears to be a fulcrum between the two shaft gears and the driving-gear – though I cannot see what it did. There seems another on the top arm. They might been associated with the pressure-adjustment, perhaps conjoining to keep the shafts parallel. The gear teeth look about normal depth so the work-piece thickness was only ever within a very small range.
A curious detail is the two nuts on what I take to be the adjustable upper roll's journal. They are square and either lock-nuts set corner-to-corner, or double thickness, The rest of the fastenings are hexagonal bolts / nuts or cheese-head screws, except the square-headed set-screws for the gears – as was common practice. So either they are replacements in the dim and distant, or were perhaps further adjusters operated by a key – possibly to take up bearing wear.
'''
As for approximate size, a guide to that is the proportions of the controls, designed for human hands. The lever has an elegant hand-grip we can assume is about 4-5 inches tall, maybe a little more. So by ISO-9001 calibrated finger-and-thumb caliper on the enlarged picture, and allowing for perspective, the whole machine is about 6 or so hands long – so two feet or so, excepting the driving-handle.
Do we have any idea where this handsome old beast came from, or was it found lurking in some shed?