Not the cheapest Drummond lathe ever produced (that was the Little Goliath of the 1920s) the "Round Bed" Drummond Type A was announced in the Model Engineer Magazine of May 21st, 1908. Publicised, with some fanfare, as the first £5 screwcutting lathe, it was labelled by its makers as the All Round Utility and was  to become famous as the most affordable machine for generations of British model engineers taking their first tentative - and usually impecunious - steps into the hobby. A simple 4" x 11.25" screwcutting lathe (but without backgear) it was manufactured until 1940 using, in that time, three increasingly strong headstock castings and modified spindle bearing - yet with only minor changes to the rest of the specification. Also offered, with production starting late in 1923, was a long-bed version with 24 inches between centres, the announcement of its availability being made in January 1924.
Although comparatively inexpensive, and designed to sell at the bottom end of the market, the lathe was not cheapened in any way but built to the usual honest, practical, Drummond standards. The heart of the lathe was the heavy, round, 3" diameter cast-iron bed, ground to within 1/1000" on an American Norton cylindrical grinder and formed with a bevelled slot along its underside to locate the tailstock and carriage. The leadscrew passed through the centre of the bed and was engaged by a dog clutch, the operating lever of which protruded through the headstock casting below the front spindle bearing. The end of the leadscrew was fitted with a large, un-graduated wheel for hand operation with the rather fine handle being made from horn or hardwood. However, because the leadscrew had a right-hand thread, turning the handle produced a cack-handed motion - rotating it anti-clockwise resulting in the carriage being moved nearer to, rather than away from, the chuck. Needless to say, after a few ruined jobs, most owners quickly adapted to this design quirk.
Needless to say, after a few ruined jobs, most owners quickly adapted to this design idiosyncrasy. With its top formed as a boring table, the "saddle" assembly could be partially rotated round the bed - which movement had the effect of altering the height of the table in relation to the spindle centre line. It was thus possible to make vertical adjustments when setting work on the boring table and make making minor changes to the setting of the height and angle of turning and milling cutters, etc.
A compound slide was never available for the Round Bed; instead, the single tool-slide (as used on early versions of the 3 ½" flat bed) was mounted on a vertical bar that passed through an extension to the front edge of the saddle. This extension, being split and formed into a clamp, allowed the tool-slide, which had a degree-graduated base, to be raised, lowered and swivelled. The toolpost was self-contained (which meant that no tool-holding strains were carried by the central bolt) and fitted with a round hole for boring tools as well as the usual rectangular slot.
On early lathes the top of the headstock was machined flat and the bronze bearings bolted to it. On the very first examples produced the bearings were in two parts, with the upper and lower sections separated by adjustment shims; later models had the bearings formed in one piece with the clearance set by a slot and pinch bolt. Another minor difference on the latter headstock was the method of locating the bearing caps, some being found with small studs screwed into the headstock casting and others with locating dowels - though it is not known which came first.  In 1920 - when the Serial number prefix changed from "MCHA" to "O" - the design underwent some revision and improvements: the left-hand headstock bearing "arm" was increased in size; the bolt-on bronze headstock bearings abandoned and the spindle made to run direct in the cast iron of the headstock with the clearance set by nipping up a clamping screw at the rear. Other changes introduced at the same time (although, as with the flat-bed lathes, there would inevitably have been some overlap) were a reduction in the number of T slots in the cross slide from 4 to 3; a modification to the tailstock where the barrel was increased in diameter and the previous use of a white metal liner abandoned; the adoption of push-in spring retainers on the changewheel studs instead of the previous slow-to-operate, screw-on rings (that tended to come undone when the changewheels were run "in reverse"); the replacement of the detachable changewheel reversing arm by a forward-facing forked bracket cast as part of the headstock - and the very overdue use of a gib strip with adjustment screws on the tool slide. Very early examples appear to have had a slightly different spindle to later versions with a thread 1.040" in diameter by 14 TPI (with no plain section to act as a register before the abutment face) and bored out to 0.325" rather than 0.375". The spindle was then modified to carry a 7/8"-diameter, 10 TPI thread on the nose and given a ground (rather than turned) finish on the bearing surfaces; as standard it was not hollow, but the 0.375" diameter hole could be ordered for a small extra cost. Later spindles (fitted to lathes where it ran direct in the headstock casting) had a 3/4"-diameter by 10 t.p.i thread with a small plain register about 0.25" long before the abutment face. If you find a Round Bed that differs from these specifications do not be surprised; over the years many have been fitted with replacement spindles and the factory did not seem averse to the occasional experimentation with altered dimensions. The three-step cone pulley on all models was in cast iron, rather heavily made with diameters of 6", 4.5" and 3.5" and driven by a 1" wide belt.
From the start of production until 1925 a set of eight changewheels was provided as standard: 20t, 24t, 28t, 32t, 36t, 40t 44t and 64t with an extra set to generate metric pitches (at additional cost) of  25t, 35t, 45t 50t and 63T. Thereafter a set of 9 gears was supplied - which included those necessary for metric threading: 20t, 25t, 26t, 30t, 35t, 40t, 45t, 50t and 66t giving a threading range of: 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 16, 18, 20, 24, 26, 28, 32, 36 and 40 t.p.i. Because early headstock castings lacked the cast-in slot provided on later examples (that accepted a stud on which to run an extra changewheel to produce left-hand threads) the makers offered (at extra cost) a slotted bracket that bolted to the left-hand face of the headstock. This positioned the extra "reversing" changewheel in a rather distinctive fashion above the others. As on all small Drummond lathes the changewheels were 14 DP with a 14.5 degrees pressure angle; today, unfortunately, the standard pressure angle is 20 degrees so "off-the-shelf" gears will not match and if new gears are required they have to be specially made. To connect the gears in pairs for a compound train each gear carried a pin and a pin hole (a useful tip is to know that the pins are pressed into tapered holes and, if driven out the wrong way, will fracture the gear). By combining extra changewheels a large number of odd pitches and metric threads could also be cut - however, one serious drawback with the lathe was its lack of slow speeds to help with screwcutting and large facing jobs; no backgear was fitted and the bottom speed with the maker's countershaft was approximately 120 r.p.m. To some extent this could be overcome by re-jigging the belt drive to give lower ratios or, more effectively, by resorting to various third-party accessory suppliers who offered two solutions: the ingenious WALRAM unit or some form of backgear. The (now very rare) Walram bolted in place on the end of the headstock casting using the reversing stud bracket as one of its mounting points. It provided several features: a means of driving the headstock spindle, through gears, either faster or slower than normal; allowed an ultra-fine carriage feed to be set up and permitted the generation of left-hand threads. The backgears offered were either of a conventional design - that bolted on in some way - or rather splendid epicyclic mechanisms built into a slightly larger-than-standard headstock pulley.  At least two types of the latter were offered: the first, by George Gentry, being published in the "Model Engineer" magazine during 1912 and the second, by A. E. Bowyer-Lowe complete with illustrations and detailed drawings, in the same magazine for April 1st, 1915.
A useful range of accessories was offered including a vertical slide that, when mounted on a swivelling adapter with a graduated base, could also be used to mount a neat and effective Indexing and Gear-Cutting Attachment. The swivelling adapter part must have been either very popular as an accessory, or later supplied as standard, for many used machines come with one in their stock of kit.
Over the years various drive arrangements were available including the usual type of fixed and fast-and-loose countershafts for wall and ceiling mounting - but the most popular and common arrangement was a pair of legs and a chip tray (all in cast iron) fitted with a 100 lb flywheel running on a plain bearing and with operation by treadle.
Though of simple design and construction the Round-bed Drummond was, in sympathetic hands, capable of accurate work and many examples are still in use today. In standard short-bed form its overall length was 2 feet 11 inches (889 mm) and the basic machine weighed approximately 105 lbs. (48 kg).
Colour: the writer of often asked what the original colour was. Evidence from removing the maker's badge from an early (detachable headstock bearings) models shows this to have been a very, very dark blue, so dark that it appeared, at first a first glance to be black. On exposed surfaces this original paint would have oxidized to a an even darker shade, so giving the impression that black was the original colour.
How many Round Bed lathes were made? The only evidence stems from a comment in a magazine article of 1921 quoting: Fifty of the new type 31/2-in lathes are made each week, and nearly that number of the popular 4-in (Round-bed) model". Reducing that figure to allow for slower sales in the depression of the early 1930s and during the approach to WW2 a figure of around 50,000 would seem a reasonable estimate..