Cogs on both sides of the hubs in 30s TdF - Why?
Roger Marquis
[ Author was John - Forester ]
[ Posted on 4 Jan 1996 12:00:05 -0800 ]
The standard road-racing gearing system until 1938 was the double-sided
hub, usually "fixed and free." That is, the hub was threaded for a
freewheel on one side and a fixed gear on the other. The cogs were chosen
so that the cyclist had a relatively high fixed gear for the flatter parts
of the course and a relatively lower free wheel for the climbs and
descents. I have not heard of systems in which one could put two cogs on
one side, although, in an engineering sense, hubs could probably be
designed to accept such a system. For some events that did not involve such
steep descents, the riders preferred a double fixed wheel, with fixed on
each side. The cyclist dismounted, loosened the wing nuts that held the
rear wheel, turned the wheel around, remounted the chain, and tightened the
nuts again. The nuts were also used, probably more frequently, when
changing flats.
In those days the riders were responsible for all their own support, and
carried their spare tires slung over their shoulders in a figure-8 pattern.
When I started racing in 1949, that was still the usual way, although
derailleurs were used in all the hilly races, such as those in N.California
and in Massachusetts, the two areas in which I raced.
Tullio Campagnolo did not invent the derailleur. Derailleurs had been
used by tourists since about 1905. The typical derailleur until the late
1930s had its jockey wheel mounted on a bracket brazed to the chainstay. On
my first custom frame, from Drysdale in New York, built in 1949, I
specified such a derailleur, a Cycle Oppy, partly because it used 1/8 inch
chain when 2-mm chain was hard to find in the USA. Many of my cycling
friends used similar derailleurs, although that was the year that the
Campagnolo derailleur became standard for racers. The disadvantage of such
derailleurs was that the jockey wheel had to be sufficiently far forward to
clear the largest sprocket. Therefore, it was a long distance from the
small sprockets, so that changes between the high gears were slow and
required considerable overshoot. During those years, derailleurs were not
generally allowed in road races because derailleurs required freewheeling,
and mixing riders with fixed and free wheels produced problems on the
turns, when fixed-gear riders were limited by pedal scrape on the turns
while free-wheeled riders were not. However, there was also a series of
special races for derailleur-equipped bicycles, typically hill climbs, that
were sponsored, at least in part, by the derailleur manufacturers.
By 1938, Simplex had perfected the next generation of derailleurs whose
jockey wheel moved both in and out, to change gear, and forwards and
backwards to follow the change in sprocket sizes. The Simplex used a cage
very like those often used today, where the cage pivot is behind and below
the shaft for the jockey wheel. Therefore, as the amount of free chain
increases as smaller sprockets are selected, the jockey wheel rises to
remain close to the sprockets. However, on these derailleurs the lateral
movement was produced by the cage pivot shaft sliding laterally,
lengthwise, inside a bushing held by the "fixed" part of the derailleur. I
put fixed in quotes, because Simplex had two pivots, one at the derailleur
tang and another at the cage pivot, which worked in tandem to take up the
slack and whose tension springs were individually adjustable. The lateral
movement was powered by the cage pivot spring, that both took up slack and
pushed the cage toward the large sprockets. That spring was opposed by a
pull chain that ran inside the hollow pivot shaft and out through the
radiused boss on the outside the fixed arm, through a cable housing stop on
the fixed arm, and then up to the shift lever. The problem with this design
was that the bushing was really too short for its required diameter and
stroke. The least bit of dirt in the sliding fit and shifting went to hell.
The derailleurs were fitted with a spiral expanding dust cover (that many
people thought was the actual spring), but that did not keep out all the
debris that flew around. Cleaning and regreasing frequently was highly
desirable. I fitted Simplex 8-speed gearing on the bicycle that had come
with the Cyclo Oppy, once I had decided that replacement parts, including
2-mm chain, had become more readily available in the USA. Cyclo, which had
some financial relationship with Simplex, copied the Simplex principle with
the Cycle Benelux series, as did several other makers in time. By 1938, the
Simplex design had become so accepted that it was allowed in the Tour de
France.
After WW II, Campagnolo invented two derailleurs. The second was the
modern parallelogram movement, 1949, that replaced the sliding bushing.
This was later modified by Suntour, who placed the parallelogram at an
angle that approximated the angle formed by the sprocket cluster and
thereby allowed the jockey cage to be again pivoted on the same pivot as
the cage. However, few of you have ever seen Campagnolo's first derailleur,
produced in 1947. I have raced against a rider using one, and a few years
ago one showed up completely unused and was built into a racing frame by a
local framemaker as a historical exhibit. This derailleur contained no
system for taking up the slack chain. The chain went direct from chainwheel
to sprocket in all gears. How did he do it? What's more, how did the rider
use it? All the slack chain was taken up by movement of the rear wheel in
and out of the frame. To ensure that the wheel continued to run true, the
axle had a small spur gear around each end, and the upper surface of the
rear dropout was machined with rack teeth to match. Therefore, the axle
rolled in and out of the dropout in perfect alignment because both sides
had to move the same amount. The bicycle was fitted with two derailleur
control rods that extended up the right seat stay from the hub. One rod
released and clamped the quick-release hub, the other carried a fork that
pushed the chain laterally to line up with the desired sprocket. To change
gear, the rider first loosened the rear hub. Then he shifted the chain to
the desired sprocket by coordinated pedaling and movement of the fork. As
the chain shifted from sprocket to sprocket, the axle rolled forwards or
backwards under his weight (no forceful pedaling here -- probably the chain
had to be pedaled backward, because I remember that the fork was on the
upper side of the chainstay). When the chain was on the correct sprocket,
the rider reclamped the hub and resumed forceful pedaling.
Along with the parallelogram rear derailleur, Campagnolo invented the
parallelogram front derailleur. The combination proved unconquerable. For
decades, most racing bicycles (and all bicycles where price was no object)
were "all Campy." Before that, the typical front derailleur was just a fork
pivoted on a pivot that extended forwards from the seat tube, with its
upper arm extending up to a handle. The rider reached between his knees and
flipped the fork over. I used such a front derailleur with my first racing
derailleur setup. The Campagnolo front derailleur used a shift cable, so
the "all Campy" bike had two shift levers mounted at the top of the down
tube, another reason why racers adopted the setup. Simplex put out a front
shifter with cable drive to a sliding bushing to compete, but these
disappeared as less satisfactory than the parallelogram movement.
John Forester
JFor...@cup.portal.com 726 Madrone Ave.
408-734-9426 Sunnyvale, CA 94086-3041
Roger Thomas
<newc...@sable.ox.ac.uk> writes:
> Finally, the double sided hub was used briefly in I think the
> late 60's or early 70's in an attempt to liven up Paris-Tours; stopping
> it's finishing in the usual bunch sprint.
Yes. The experiment was in fact tried in Paris--Tours in 1965 and 1966. J.B. Wadley in "The
Guinness Guide to Cycling" (1977) which is essentially an edited translation of J Durry's
"Le Velo" (1976) actually says that the sprocket (or sprockets? see below) was fixed not free.
There is a fine picture in the book of Belgian national champion Guido Reybroeck taking
the finishing sprint (which turned out to be the "usual bunch sprint" again!) from Rik Van
Looy. The caption notes that an average speed of 43.3kph was managed on the alleged
fixed gear. But was it a single gear? Barry Hoban is another good source (as so often). In his
book "Watching The Wheels Go Round" (1981), he notes that in the _1965_ race the riders
were allowed one chainring and _three_ sprockets:
"To change gear, we had to jump off the bike, loosen the back wheel and put the chain by
hand onto the next sprocket. But the great plan [of race organizer Felix Levitan] backfired
and it ended with the 248 kilometres being covered at a record speed of over 45 kilometres
an hour! We were going so damn fast that I didn't have a chance to change gear and I
rode the whole race with 51 by 15, a fairly low gear of about 92 inches.
"Tom [Simpson] was keen to confirm his world championship win and the two of us
managed to get away on the roads coming into Tours. It was all well and good to get clear
by revving the pedals round, but we really needed to be able to change up to a higher gear
to consolidate the break. This we couldn't do and we eventually ran out of steam. Some of
the other teams were better organized and had fully considered the likelihood of a sprint
finish. Therefore, with about 20 kilometres to go, the whole of the Dutch team, Televizier,
with De Roo, Geldermans, De Haan and Karstens, all jumped off their bikes together
and moved the chains over to engage the 53 by 14, a gear of about 102
inches. It was impossible to compete against high gears like that, and powerful riders like
that, on our comparatively puny gears. Karstens won Paris--Tours at record speed. But I still
managed to finish fourteenth."
Unfortunately, Hoban's account of the 1966 race is brief (he just says it ended in a mass
sprint and that he finished 11th). I'd guess they were "three-speed" freewheels again (and
not double-sided gears as Tom James suggests -- three is enough gears and turning the
wheel round would be too time-consuming in a race of this character) and although
the picture in the Dury/Wadley book is not very clear from that point of view, they do
rather look like freewheel blocks to me.
Hoban also confirms that the 1967 race went back to derailleurs. (He almost won, being
edged into second place in the sprint by Rik Van Looy.)
An authoritative view of cycling history (like any other) is plainly hard to come by! I've just
looked at Noel Henderson's "European Cycling". He says of the 1965 race: "Riders used a
single freewheel with a choice limited to two chainrings, with changes made by hand"
(Quite what this ambiguous sentence means I'm not clear. Does it mean they could use
two chainrings or, more probably, does it mean that a freewheel size was designated and
then riders could choose one from two chainwheel sizes?)
Whatever, I think Henderson's wrong -- application of Occam's Razor leads me to accept
Barry Hoban's account over Henderson's. Hoban's memory might be defective, but were
that so we'd also have to explain away as a very peculiar aberration of memory his vivid
stories of the mass synchronised gear change of the Dutch and his and Simpson's vain
escape.
(Another possible source might be Tom Simpson's autobiography -- unfortunately I
have no copy of this.)
--Roger Thomas
Tom James
fixed- on each side of the hub, giving four gears in total which had to
be changed by hand.
Henri Desgrange, the director of the Tour de France from its
inception until 1937, was nothing if not a hard man. In an attempt to
stop trade sponsors getting "free" publicity from the tour, by the 1930s,
when the tour was run with national teams, he had all riders riding
anonymous yellow bikes provided by the organisation. He also forbade the
use of the early derailleur gears available at the time, partly because
each marque was highly recognisable to the fans of the day even if
unbranded and he wanted to avoid publicity for the component
manufacturers; and partly because he thought gears would make especially
the mountain stages too easy (!) However, this led to an anomalous
situation, whereby the "Touriste Routier" class of rider, who were
allowed derailleur gears, started to beat on some stages the
professionals, who had to labour with stopping to change gear. Clearly,
the manufacturers who paid the riders could not accept this situation.
Additionally, in many of the other races of the time, gears were allowed
and this led to their being run off at a faster speed. The Tour
organisation could not accept for long this situation - the Tour had to be
the longest, hardest, fastest race. Nonetheless, no derailleurs remained
the rule until Victor Goddet replaced Desgrange halfway through the 1937
race. The following year, derailleur gears were allowed for the first
time, the winner being Gino Bartali.
The immediate result was an increase in the overall speed of the
peloton, and this led to a gradual change in tactics. Whereas previously
the strongest rider won by a process of attrition, wearing out his rivals
over a long distance of sustained hard riding, the new style involved
treating the peloton as a mobile base for operations. Attacks would be
essayed if the attacker thought he had a chance, but otherwise the
priority was to stay in the peloton and not lose time to a rival, even if
you didn't gain time either. Thus the modern pattern developed, with the
classification remaining reasonably static from day to day with
occasional upheavals on the big "set piece" days. The new pattern also
led to increasing importance for team riding, with a designated leader
and a number of domestiques.
Cycle Sport magazine (in Britain) a few months ago (I haven't got
it to hand, so I can't remember exactly when) had pictures of some old
bikes. One of them was Bartali's 1938 tour bike, with a derailleur in
which the chain tensioning was achieved by the wheel rolling back and
forth between the dropouts, rather than the modern system.
Finally, the double sided hub was used briefly in I think the
late 60's or early 70's in an attempt to liven up Paris-Tours; stopping
success, and it very quickly returned to allowing derailleur gears.
Tom
N.Le...@massey.ac.nz
> Henri Desgrange, the director of the Tour de France from its
>inception until 1937, was nothing if not a hard man.
The art of understatement is alive and well!
> In an attempt to
>stop trade sponsors getting "free" publicity from the tour, by the 1930s,
>when the tour was run with national teams, he had all riders riding
>anonymous yellow bikes provided by the organisation.
I think that they were made by a company called Kingfisher, but that
could be a complete load of bollocks.
I've asked this before, but had no reply. Once upon a time the
"jerseys" of the TdF had the initials HD on them as a tribute to the
old sadist. Does anyone know why this stopped?
Neil Leslie Email: N.Le...@massey.ac.nz
Department of Computer Science Tel: +64 6 350 4185
Massey University Home: +64 6 354 8130
Palmerston North Fax: +64 6 350 5611 Attn. N Leslie
New Zealand
Keep your hand upon your wages, and your eye upon the scale
Tom James
On Mon, 8 Jan 1996, Roger Thomas wrote:
(Lots of interseting stuff about Paris-Tours, and the lack of derailleur
gears cut)
> (Another possible source might be Tom Simpson's autobiography -- unfortunately I
> have no copy of this.)
>
> --Roger Thomas
>
>
It's a while since I read Simpson's book (and, as usual it's not to hand
- the problem of studying full time in Oxford but living elsewhere!) but
as I remember it says nothing about Paris Tours. More or less the last
race covered is the 65 world championship - I don't even think that
Simpson's win in the Tour of Lombardy is covered, which was one of
Simpson's most emphatic wins.
Apart from that, it is a well written book, worth seeking out if
you can.
Tom
Carlsun
which could be interchanged from front to rear (there were splines to fit
a special freewheel, as I recall).