Velocette made a series of production racers from 1928 to 1951, which were essentially replicas of their successful Works racers; other manufacturers would have called them 'TT Replicas', but Velo were early adopters of overhead-camshaft racers, and called them the 'KTT' ('Kamshaft Tourist Trophy' is the usual translation). After that first KTT of 1928, which was a very close copy of Alec Bennett's '27 Junior TT-winning machine, subsequent iterations of the model had a series of "marks" to designate the major changes, although during the first 5 years of production they were all simply 'KTT' in factory literature and advertising.
The first major change to the line came in 1932, with what the factory itself named the 'Mk4' KTT; thus the earlier machines gained retroactive 'Mark' status of Mk1 thru Mk3. Later models were generally given a new 'mark' with a significant change - thus the Mk5 in 1935 had a new frame and bronze cylinder head, the 1937 Mk7 had a completely new all-alloy engine and new frame, then the final 1938-51 Mk8 had the Velocette invention of a swing-arm rear end with two suspension units - a setup recognizable to any motorcyclist today.
Ah, but what of the 'missing' Mk6 KTT? For whatever reason, the factory skipped over the number 6 as a machine offered for sale, but like Nessie, rumors have swirled since 1936 regarding the existence of such a machine. Contradictory reports from authoritative sources haven't done the historical record any good, and several published works give contradictory evidence.
Readers of 'Always in the Picture' - the original Velocette 'biography' written by Burgess and Clew, Dave Masters' first-edition booklet 'An Illustrated Profile of Models 1905-71', the "The Velocette Saga" series by Titch Allen included in "MotorCycle Sport", and Jeff Clew's 'Classic Motorcycles', all repeated the line 'there was only one Mk.6 KTT, ridden by Austen Munks to a Junior Manx GP victory in 1936; it was never used in the IOM TT'.
Unfortunately all of the main characters involved with the factory's project are now dead, so we rely on photographic research, race entries, and clues from the Works records, to find out more about this elusive beast. And from all this, we can conclude that there was not one but THREE Mk.6 KTTs, all ridden in the 1936 Isle of Man Junior TT.
These seem to be in fact three prototype Mk.6 KTTs, for the production of the Mk.5 KTT finished in October 1935 with engine number KTT620; there was a gap of 17 months until the first of the Mk.7 KTTs, starting with engine number KTT700 [I have engine #702 - pd'o], was dispatched to Australia.
The search for the Mk6 began with a physical description, as from a crime scene; these Mk.6 KTTs looked like the Mk5 KTT, onto which Harold Willis, Veloce's race shop genius, had grafted the aluminum cylinder head from the new KSS mk2 roadster model. Willis, ever the masterful nick-namer (having invented the term 'knocker' for overhead camshaft, among dozens of others still in use), dubbed his creation "The little rough'un".
Another useful snippet of information came to light in Les Higgins' delightful book 'Private Owner', about his years racing Velocettes. Regarding his experiences during 1936, Higgins says on page 41:
"another machine was difficult to come by, because Veloce Ltd. had temporarily ceased manufacture of the KTT model. The last machine made in any quantity were the Mk.5 models. A few Mk.6 machines went out to approved customers and the concern was now busy evolving something new to provide an answer to the International Norton".
The 'approved customers' of the three Mk.6 KTTs at the Isle of Man in 1936 were H.E. Newman, Billy Tiffen and Roger Loyer. Photos from 1937 show Ernie Thomas, Les Archer and his son, and one unidentified rider, all circa June 1937; because of the date these are likely from the '37 IoM TT races.
An extract (below) from the Velocette factory KTT records gives the following notations:
June 1936 Veloce for H.E. Newman for TT KTT621......frame 6TT3
June 1936 Veloce for W.T.Tiffen Jnr for TT...KTT622.. .frame 6TT4
June 1936 Veloce for Loyer for TT....................KTT623 frame 6TT5
The reference "Veloce for" or "Veloce Ltd for TT" is used throughout the factory KTT records to indicate a Works-prepared racer (and to clarify, of course all KTTs came from the factory, but a separate Race Shop created the 'Works' machines for use by designated riders, as opposed to the standard KTTs offered to the general public). Similarly the frame designation 6TT was used on the factory's rigid-framed TT racers.
Interestingly, in Das Motorrad of April 18, 1936, these two photos of Ted Mellors' machine are included. The engine number is obscured, but the publication date of the magazine predated the mk6 factory notations above. Mellors was occasionally engaged by Veloce to race their equipment - and the machine in these photos certainly looks like it came from the factory (the bevel box drain is as per works, not as per KSS mk2). The mystery deepens...
could this have been the first testbed of the new cylinder head?The weekly motorcycle paper "MotorCycling" of June 17, 1936, in the "Straight from the Island" article, included this interesting snippet, previously overlooked by historians:
"One or two of the lads who are riding Velocettes, notably H.E Newman, will be mounted on advance KTT jobs for 1937. These are standard products, but they can motor very rapidly".
How did they fare in the TT itself? Billy Tiffen retired at the end of lap three with a broken front fork spring and Roger Loyer (usually entered by Boudene, the Velo agent from Paris) stopped soon after his 5th-lap pit stop and retired with "lubrication trouble of an irremediable kind".
That left Newman, who came in 10th, the third Velocette to finish, with Ted Mellors in 3rd place and Ernie Thomas in 4th place, both mounted on 'Works' DOHC 350 Velos; definitely not 'production prototypes'!
Newman's lap times are not spectacular, but consistent at around the 30 minute mark, with a fastest lap of 28 min.58 secs. He was marginally slower than the 'works' DOHC Velos, and faster than the Mk.5 KTTs ridden by Noel Pope, Gledhill and Chas. Goldberg from New Zealand.
Development of the Mk6 seemed to continue, for the photo of Ernie Thomas mounted on a Mk.6 KTT in the 1937 Junior IOM TT seems to be Newman's 1936 machine, now with a different type of seat. And as mentioned, the Archers, well-known Velocette agents, also rode a Mk.6 KTT in the 1937 TT. Obvious changes to the model included the seats - the "dual" seat, dubbed the 'Loch Ness Monster', was introduced on factory racers in 1934 (and can be seen clearly below)- and the larger petrol tank, which would be carried over to the Mk.7 model, was used on factory bikes starting at the '36 TT. The rear wheel of Archer's Mk.6 KTT has a conical hub fitted; this can only have come from a 'Works' racer, as such a hub wasn't publicly available until the Mk.8 KTT appeared in April 1939.
With reasonable results in the TT, why didn't the Mk.6 continue into full production? Perhaps with the introduction of the massive square-finned alloy SOHC engine ( 10" x 10") to the factory bikes, and the subsequent improvement of lap times, Harol Willis felt the all-alloy engine with big fins was the future. Stanley Woods was lapping over 1min.17secs. a lap faster than Ernie Thomas in the 1937 Junior TT, despite the development time spent on the Mk.6 during the intervening year - Thomas's best time was only marginally better than Newman's the year before. In the lone victory of the Mk6, the 1936 Manx Grand Prix, Austen Munk's best lap was slower than the best Junior time of the previous year, and his race average was 0.1mph faster than Newman's '36 TT time (below, Noel Pope in '37).
Why so much slower than the 'Works' bikes? A more detailed review of factory literature gives drawing 'KO2716', part no. K27/16, a piston, described on the drawing as for the KTT Mk.6, plus drawing KO2805, part no. K28/5 a connecting rod, also for the KTT Mk.6, give substantial clues to the engine's weak points. Inspection of a K27/16 piston from a Mk.8 KTT revealed it to be, as expected, identical with the drawing. The instruction sheet published by Veloce Ltd. for the Mk.7 and Mk.8 KTTs list the K27/16 piston as the Mk.7 'alcohol piston' and the Mk.8 KTT 'petrol-benzol' piston, ie, a high-compression item for use with very high-octane fuel.
After a long look at the period photographs, it seems the cylinder barrel of the Mk.6 is the same alloy barrel of the Mk.7 and Mk.8, making the Mk.6 in effect a Mk.7 engine type, but with a KSS Mk.2 cylinder head fitted. After testing in the TT and MGP of 1936 and 1937, together with a few Continental GPs (according to the late Bruce McNair's conversations with Roger Loyer), the Mk.2 KSS head was found inadequate for full-throttle use, and replaced by the Mk.7/8 type with larger finning. These cylinder heads were nearly the same as the 'Works' 1937 cylinder heads, although only 9" x 9" square; the resultant machine was then offered to the public as the 'Mk.7' in 1938.
What was the problem of the KSS mk2 cylinder head? The combustion chamber of the KSS head is virtually identical with the Mk.7 head; the valve sizes are the same, but the inlet tract is a bit larger. Perhaps there was trouble with the coil valve springs inside the Mk.2 head at the sustained higher rpm of racing; the re-adoption of hairpin valve springs to overcome this, as used on the Mk5 and Mk7, would have meant a major redesign. Similarly the relatively scanty finning on the Mk.2 head may have caused overheating, especially as racers used a hot-burning petrol-benzol blend. The use of the 'Works' big square cylinder head, as seen on the Mk7 and beyond, must have been the clear path forward.
Many thanks to Dennis Quinlan for his primary research, and allowing me to substantially edit his authoritative work.