Looking similar to the well-known "Relm" lathes, the RSB might have been made by Thomas Breed, in Bramley, near Leeds, his Company, Breeds Lathes Ltd., having a connection to Relm and whose lathes were sometimes branded "Atlas". Just three surviving examples are known, the one shown below, another, in better condition (with simple "split" headstock bearings closed down by a screw at the rear) and one in New Zealand.
The pictures on this page show a lathe of 3-inch centre height and 12 inches between centres that would have been designed - judging by its appearance and specification - during the first decade of the 20th century. Intended for the amateur market, production could well have continued into the 1920s though unfortunately, so far, no contemporary advertisements have been found. The bed, very lightly constructed with thin walls, was of typically English style with a flat top, V-edged ways and provided with a decent gap.
Cast as-one with bed, the headstock lacked a speed-reducing backgear - though the 1-inch wide 3-step pulley unusually generous diameters of 2.5", 4" and 5.5" so helping to improve belt grip. Where simple and cheap split headstock bearings might have been expected (and are in place on the other example) this RSB had a proper 2-bolt cap type with thin, split bronze shell bearings that, on the surviving example, have survived the decades unworn. The "split" type, though much cheaper to produce, have that inevitable pitfall of the type, the possibility of cracked castings as a result of enthusiastic over-tightening of the adjustment screw.
Screwcutting was by changewheels - these being mounted on an awkward-to-set-up, L-shaped bracket - and through an unusually large diameter, fine-pitch leadscrew fitted with a large balanced handwheel at its tailstock end, a dog-clutch by the headstock and a "full nut" bolted to the underside of the saddle - there being no necessity for a conventional apron. The full nut appears to have been split and fitted with an adjustment bolt at its lower end - an unusual addition on a lathe of this class (indeed, on one of any type). Unfortunately, where the tool thrust was taken out against the back of the bed, instead of a solid-to-solid contact, an adjustable gib strip was inserted, this introducing unwanted flexibility - the strip and its adjuster screws should have been at the front - through with the light forces involved in turning on such a small machine, the arrangement would have been of little consequence.
A single swivelling tool-slide was provided, carried in a traverse T-slot in the saddle - with two more slots that ran longitudinally down the right-hand saddle wings -  so providing the makings of a handy boring table. Judging from the thickness of the saddle and top-slide castings, it's likely that sufficient room was available to fit a proper compound slide rest, something that the maker may well have offered at extra cost.
Of the simplest possible kind, the tailstock was a single casting - and so with no base on which to set over the top for taper turning. Fitted with a No. 1 Morse taper nose the spindle was locked by that horribly crude method, a slit in the casting closed down by a bolt.
If any reader has an RSB lathe (swan-neck drills with the same branding have also been discovered) the writer would be interested to hear from you..