At the tender age of 14, P.J. 'John' Wallace had an epiphany at a motorcycle exhibition, and decided to build a motorcycle. He bought a set un-machined engine castings for £2 10s, and proceeded to equip a workshop in his father’s garden, teaching himself to use a few simple machine tools. He soon realized the finish work required of the castings was beyond his equipment and his ability. So he bought a frame and wheels from a local cycle maker, plus a secondhand engine, and built his motorcycle, which he promptly sold.
In 1912 (age 16), John landed an apprenticeship with Collier & Sons, of Matchless fame. After an accident at the factory, John’s father put a stop to his employment with the Collier’s. The pill was sweetened by his father buying him and his brother a pair of T.T. Rudges (see below). They both swiftly joined the British Motor Cycle Racing Club (B.M.C.R.C., or 'Bemsee') and took to racing at Brooklands. However, things did not go as planned (ah, racing!) and soon John's crashed his Rudge, which was damaged beyond repair.
With this meager Brooklands experience under his belt, in 1913 he secured a job as a test rider for the J.A. Prestwich (J.A.P) experimental department, where testing motorcycles at Brooklands was part of the job description. (Also employed thus was one Herbert LeVack, about whom more later.) When J.A.P. became aware of his age they promptly sacked him! Wallace spent the the next year studying engineering and training to become a draughtsman. With the onset of WW1, Wallace felt there would be little demand for motorcycles, so took a job at Scottish car makers Arrol-Johnston, as an aero-engine designer. This employment too was short-lived; it was over by mid-1915. However, his lengthening resumé was enough to land him a job with the design team at Westland Aircraft Company (Petters Ltd), which was to last until the end of the war.
Late in 1918 Wallace returned to his first love, and laid out an advanced high-performance motorcycle engine. When drawings were finished, he cleverly advertised his design in The Aeroplane, knowing aircraft builders would need to diversify after their war contracts had ended. One such company was the Portable Tool & Engineering Co. of Enfield, who were impressed enough to employ Wallace as Chief Designer. Their plan was to sell 'loose' engines to motorcycle manufacturers, and by September 1919 the prototype was ready for trials. Clearly, Wallace had learned a few tricks from cutting-edge aircraft technology, as his engine used Overhead-Valves and was 'oversquare' at 88.9mm bore x 76.2 stroke, giving a capacity of 475cc, using a fully-recirculating oil system with two oil pumps on the timing cover; all very advanced for 1919.
Herbert LeVack (above) had been employed during the war assembling and testing aero engines, and his services were secured by P.J. Wallace to build his new motorcycle engines. LeVack proved a valuable asset, with an uncanny ability to produce wonderful results from ill-fitting components. He built the prototype engine and got it running satisfactorily; a second engine was then fitted into a motorcycle chassis, and used by Le Vack in demonstrations to the trade & public, and in competitions. LeVack's development and riding skills produced excellent results from Wallace's design. The motorcycle was first christened the ‘Ace’, then the ‘Buzmo’, before ending up as the ‘Duzmo’ in 1920. 
LeVack won many speed events on his tuned single-speed belt-drive Duzmo, winning over 100 awards. Racing success created demand from the public, but the business plan with Portable Tool called for engine manufacture, not motorcycle manufacture, and Duzmo was barely a company! There was no chance of fulfilling orders for whole motorcycles with the small workshop that P.J. Wallace ran near the Enfield highway. Wallace suggested to the Board of Portable Tool that they take Duzmo 'public' and sell stock to raise capital for proper motorcycle manufacturing facilities, but they balked, and wound down production. A silver lining emerged when a kindly Board member loaned Wallace enough money to create his own company (John Wallace Ltd) to build his Duzmos.
Ever looking forward, in 1920 Wallace (on the right, above) and Le Vack altered their single-speed frame to fit a 3-speed Sturmey-Archer gearbox, for all-chain-drive. This machine completed the 1920 London to Edinburgh trial, and was then shipped to the Isle of Man for LeVack to ride in the 1920 Senior T.T. ... he was no stranger to the T.T., having raced there in 1914 (the last T.T. before WW1) finishing in 15th place at 45 MPH on a Motosacoche, winning a gold medal. Road conditions in the race were atrocious, more resembling motocross than road racing to modern eyes, on machines with almost no suspension, narrow high-pressure tires, and virtually no brakes.
Press reports state Le Vack’s Duzmo arrived on the island via the Saturday morning boat, leaving little time to practice. Another report mentions Le Vack laboring over his machine since its arrival, working almost night and day, being rather handicapped by the lack of spare parts. Reading between the lines on these reports, it is possible Sclater’s Duzmo was sacrificed to keep Le Vacks machine alive, and might be why Sclater ultimately rode a Norton (below) in the T.T. that year.
The Senior race was held on Thursday, June 17th, in favorable conditions. Le Vack on the Duzmo had an excellent start, but on the second lap he had a bad skid at Governor’s Bridge and fell, bending his rear stand enough to rub the tyre; he was delayed eight minutes while he removed it and left it behind. He was reported passing through the grandstand on his third lap at speed, with his engine emitting a healthy bark.
The fifth lap saw 16 competitors still in the race. Le Vack tried to overtake another rider near the Bungalow, when his quarry suddenly shot across his racing line, and Le Vack was brought off, damaging the Duzmo and forcing him to retire. The name of the fellow who supposedly cut off LeVack was never mentioned - was he forced into a ditch or did the Duzmo simply blow up? It was common practice for manufacturers to disguise mechanical calamity by blaming chains or magnetos or a spill. The race was won by Tommy de la Haye on a side valve Sunbeam.
Herbert LeVack had greater ambitions, and by early August the press announced that he had severed his connection with Duzmo, joining Freddy Dixon in the Indian camp, and his track career blossomed at Brooklands where he so regularly broke speed records, he became known as 'The Wizard'. 
Wallace soldiered on racing with himself as tuner/rider, with much less success than LeVack. In a move which foreshadowed the legendary Vincent tale of 'doubling up' his single cylinder machine, a new 992cc ohv v-twin was built for racing at Brooklands in 1922, by adding another cylinder to his original design. He also penned a new single-cylinder chassis design in 1922, with a unique sloping petrol tank, and while an attractive machine, sales were poor, and the Duzmo was finished by 1923. 
Tim Walker bought his Duzmo from a friend who had saved it from the scrap yard in the late 1960s; it came with an early Triumph; the friend bought the pair not knowing what the Duzmo was - above is the Duzmo on the trailer just after he got it home. The crankcases were smashed around the timing gears, as a broken cam follower pivot had wreaked havoc. There were 2 tax discs in the holder, last dated 1928, both issued in Cornwall. It looked to have had a hard life but was basically intact, sadly missing the 2 external mechanical oil pumps that feed and return the oil to a tank beneath the saddle. Clearly, the bike had been ridden using only the hand auxiliary oil pump fitted on the side of the oil tank. Both bronze-bodied external pumps had been sawn off, the plunger holes plugged and soldered up. Tim was lucky, as the remains of the pumps indicated the diameters of the original plungers, so he was able to reproduce them. 
The forks are Druid MK 2s, which move up and down normally, but include a limited movement backwards and forwards through a complicated link, which alter the wheel base slightly. Gearbox is a close-ratio Sturmey-Archer 'CS' type with a Sturmey single-spring clutch. Carburetor is large bore Amac, and the magneto is an ML. The front brake is a Bowden scissor action lever on a dummy rim, which actually works, but only in the dry. The rear brake uses a V band attached to the spokes. The frame still has a casting to mount a second rear brake, to press the other side of the dummy rim. LeVack raced his Duzmo with both rear brakes fitted, but no front brake of any type. The back hub is an early Enfield type with a cush drive fitted. Only the saddle was missing from my bike.
The prior owner did a cosmetic restoration but never ran the Duzmo due to the crankcase damage. Tim Walker repaired the crankcases for him, but advised his friend not to ride the Duzmo until new crankcases were made. He did some research, but never found another Duzmo, only old Motorcycling reports and drawings of the new oil system from 1919 reports. The engine is the same over-square OHV with 88.9mm bore x 79mm stroke, with fully re-circling oil system - just as LeVack used.
With today's eyes, the Duzmo has quite a few design faults; weak valve gear and pivots , very soft valve springs which allow valve float at around 3500 rpm, although there are loads of power below this range. With the over-square engine, it would seem Wallace designed it to rev higher than 3500 rpm, but the design fault is in the timing chest; stronger valve springs would enable it to rev more but would load up the cam gear and lead to disaster, which must have happened to Tim's bike in 1928.
Still, the Duzmo was technically far ahead of the competition in 1919/20, and it is no surprise to find such a modest enterprise as the Duzmo had teething troubles which were never addressed, as the baby never got old enough to have teeth!
Many thanks to Nick Harrison at Duzmo.co.uk for his research and photographs!