Gearing dreams and reality (another one for epicyclists)

Sometimes, you do not fully appreciate why something is the way it is till you try to do it differently (as a writer and a lover of books, I hope that may be the lasting effect of e-readers such as the Kindle:  they will make people appreciate just what a clever piece of technology a book is – but that is by the way). My Dream Roadster, as its name suggests, is an attempt to realise an ideal form of the Imperial Roadster bicycle by retaining its desirable features while overcoming its shortcomings, principally in gears and brakes.


In practice, I have found that the Dream Roadster has served mainly to deepen my appreciation of the production Imperial Roadster, in particular the two versions of it that I own, a pair of Royal Sunbeams. The matter of braking systems I will consider another day (the Dream Roadster has drum brakes, where the standard Imperial has rod-operated rim brakes) – but my excursion earlier this week ,  which involved the strenuous ascent of Necessity Brae on my 1934 Sunbeam,


revived a question I had asked before, how many gears does a man need?.  The Sunbeam’s low gear is around 54”. I had managed to climb the hill in that, though it took a deal of effort and determination – at several points I had thought about stopping, but had willed myself on.  Evidently, then, a lower gear would make things easier, but how much lower should it be? That was the question I set out to answer the next day, when I repeated the same route using the Dream Roadster.

The Dream Roadster has a five-speed rear hub (SRAM/Sachs P5) coupled to a two-speed bottom bracket gear, a Schlumpf Mountain Drive. On the present set-up, this gives nine distinct gears (two are duplicates) – a normal range of (approximately) 47”, 58”, 75”, 96” and 118” with a lower range of 19” 23” 30” 38” and 47”.  My guess, from previous experience, was that the lowest useful gear would be the 30” one so I resolved to put that to the test. Before setting out I made a couple of trial ascents of hills near by, Glenlyon Rd

Glenlyon Rd Craigie

and Quarry Rd


which confirmed my suspicions: the 19” is effectively useless – though it offers virtually no resistance, the speed at which it must be turned to make even the slightest headway requires a far greater effort than walking, for less return. The next gear up – 23” – is only marginally useful as it still requires to be spun at a higher rate than I find comfortable to achieve a forward progress less than walking pace.

So I set out to repeat the trip of the day before (map 4 here) with a minor variation at the start – a new secret way such as Craigie (the district of Perth where I live) abounds in


My aim was to use 30’’ (the mid-gear of my lower five) as my lowest gear. The ascent of Necessity Brae was still strenuous,


but less so than the day before, and at one point where the slope lessens I actually changed up to the 38” gear for a time. Though I already knew I could ascend the hill in a substantially higher gear, my aim was to find the most practical lowest gear for the Dream Roadster, and the answer to that appeared to be around 30”.

This was borne out by the fact that I continued without dismounting, save to record the occasional rarity –

red phone box

and some wild flowers here and there:

blue flowers

purple crocuses

(it’s nice, with the hedgerows so dominated by red, to find a touch blue and purple)

I even managed the sharp ascent to Craigend which had undone me the day before, though this time I was expecting it. I returned as I had previously, with occasional stops for blackberries and to record the onset of Autumn,


(a sight that recalled a line from Eliot:  ‘the communication of the dead is tongued with fire beyond the language of the living’)

and so home via another secret way:

a narrow walled lane

The one casualty of the trip was my mileometer, which I must have shed somewhere along the way (I did have to stop at one point to adjust the front wheel, which had slipped, so I think it was some time after that – I retraced my route on foot next day, but to no avail).

lucas cycle odometer

before setting out…

bicycle wheel sans odometer

…after returning home

So, the conclusion to be drawn (besides the obvious one of checking the tightness of all fixtures before setting out) is that around 30″ is as low a gear as I need. However, once I fell to my calculations, I found a familiar problem. An earlier version of the Dream Roadster had coupled the Mountain Drive to a 7-speed SRAM/Sachs hub, giving 14 gears with a spectacularly wide range of 763%. This setup was arrived at before I ever built the bike and sprang from a couple of simple-minded comparisons – the fabled (and fabulously expensive) Rohloff Speedhub had 14 speeds but a range of ‘only’ 526%, so an arrangement that was cheaper but offered the same number of gears over a wider range had to be better, surely?

In practice, I soon encountered the difficulty I describe above – the lower end of the gear range was something like 17, 20 and 24 inches – of which only the last was even marginally useful. This led me to conclude that the SRAM/Sachs 7 speed hub offered an adequate range in itself, so I swapped it for the SRAM/Sachs 5 speed I had installed in my daughter’s bike, giving us both (I hoped) a more useful range of gears.

However, setting 30” as a bottom gear with the P5/Mountain Drive combination and a 28” wheel implies a direct drive  of some 119”, not only ridiculously high in itself and giving an unfeasibly huge 188″ top gear, but requiring a gear ratio (front to back) of 4.25:1, meaning that a 48T chainwheel would need an 11T sprocket (impossible to obtain for a hub gear – 13T is the smallest I have come across). With 13T at the back a 55T chainwheel would be required to maintain the same ratio.

The upshot is that I have ordered a 14T sprocket which, with the present 48T chainwheel, would give a gear ratio of 3.4:1, meaning a direct drive of c95” and a lowest gear of c24” – still barely useful. The full range would be (approximately)

lower: 24, 30, 38, 49, 60        upper: 60, 75, 95, 122, 150

I look forward to testing that – the long-striding twelve-and-half foot top gear should be fun* but I have to admit that I am already toying with the notion that for my purposes six gears might be enough – using the same 14/48 set up with a typical 3 speed would give me approximately

lower: 28, 38.5, 52        upper: 70, 96, 131.

That gives a bottom gear much closer to my tested useful minimum and makes an interesting comparison with the fabled six-speed Sunbeam of 1908, which offered 49, 66, 72, 88, 96, 129:

Sunbeam A6 crop

(photo by kind permission of Leon Wise)

Some might object to the increasingly wide gaps in my upper range but I think that is a matter of taste and cycling style. The justly-celebrated Rohloff offers 14 evenly-spaced gears (at 13.6% intervals) but for me that is an epicyclic hub brilliantly conceived to do the same job as a derailleur, only better – it’s all about maintaining cadence, keeping the same input and varying the gear to suit.

Riding a roadster bicycle is about varying input to suit the conditions: you are prepared to labour up a hill, or even dismount and walk, knowing that eventually you will reach the top and be able to freewheel down the other side; and on the flat, there is no more lordly feeling than to sweep along at great speed by turning a tall gear at a dignified, leisurely pace. Large steps between high gears are not a problem as you only use them when you are already travelling at speed; it is in the low gears that you want to avoid the jarring shock of too wide a gap.

Around 70 inches was the common single-speed gear for the Edwardians, who liked to calculate at 10 gear inches for each inch of crank, and regarded 7” cranks as the standard. I reckon that a future 6-speed Dream Roadster with the set-up above would give me in one bike the equivalent of two three-speeds, one well-suited to the hill country, the other formidable on the flat.

I have to say I’m sorely tempted…

* 13 yards on the road for a single turn of the pedals, or better than 26mph at a modest 60 rpm – though I expect wind-resistance would be a factor.

Education compared to a Penny-farthing Bicycle

‘A growing body of literature has linked the ability to delay gratification to a host of other positive outcomes, including academic success, physical health, psychological health, and social competence.’ – Wikipedia article

I don’t know about the rest, but I find the link with academic success interesting: I can see why it might be the case, but I wonder if it is an encouraging sign.

There was a time –  for roughly two decades, between 1870 and 1890 – when a man’s inside leg measurement was strongly linked with positive outcomes in cycling – the longer it was, the more likely he was to be successful. The reason was simple: on the bicycle of the day – now variously called the Ordinary, the Penny-farthing or the Hi(gh)-wheeler – the rider sat astride a huge directly-driven wheel of anything from from 55 to 60 inches (140-152 cm) in diameter, so that a single rotation of the cranks would propel him some five yards and more (4.7m) along the road; the hour record for these formidable machines (paced) is a remarkable 23.72 miles. A man called Tom Stevens rode one round the world between April 1884 and December 1886.


With the advent of the ‘safety’ bicycle and its geared chain drive, the fact of being long-legged ceased to be an advantage – on a six-speed Sunbeam (available in 1907) a rider of average leg could drive the equivalent of a 129” (328cm) wheel (the legacy of the penny farthing is that (in Britain, at least) bicycle gears are still measured in inches, as the equivalent of a directly-driven wheel of that diameter). If you were to turn such a gear at a modest sixty rotations per minute for an hour – no great feat on a level road – you would travel a shade over 23 miles. Thus a different way of doing things can bring feats once reserved to the few within the compass of the many.*

This might tell us something about education. Despite great advances in the way we understand teaching and learning, the high school system in this country is still – like the diamond-frame ‘safety’ bicycle that dethroned the Ordinary – essentially a Victorian design. While the diamond frame bicycle might be likened to the shark – having early evolved a form perfectly adapted to its purpose, there has been no need to alter it – I do not think the same can be said of our high school system.

That it is a system is perhaps the first point to note: it all hangs together, from the design of the buildings, the division and delivery of the curriculum, the staff structure, the central importance of texts – which is why it has been so difficult to change. It is, in essence, conceived as an economic method of knowledge transfer: large groups of students are taught by single teachers in rooms designed expressly for that purpose. The curriculum that is delivered is divided into separate subjects, each with its expert, and the content is ordered to allow a graded progress over a period of years. Language, mainly written language, is the principal vehicle of instruction. It is, in a word, rational. It is a system that works best (and it can work very well) when the students are grouped according to ability, literate, biddable and with a capacity for deferred gratification.

The deferred gratification is needed because learning in these conditions offers little in the way of enjoyment and requires a fair degree of self-denial: it is something of a slog, and although it is rational (indeed, perhaps because it is rational) the point of it is not always obvious, and to keep at it you must believe what your teachers and parents tell you when you complain, that ‘it will all be worth it in the end’ and ‘some day you’ll be grateful’.

Yet learning can be enjoyable and exciting in itself, when it rouses the curiosity and feeds the passions – but reason and logical progression are seldom key motivators: they facilitate learning for the experienced learner, i.e. the person who has already (through years of deferred gratification) learned how to play the game. The illuminating analogy here is to consider how we learn language, and indeed how language learning has changed.

The grammar I learned in primary school was largely derived from Latin grammar,  (which is the grammar for which Grammar schools are named) despite the very considerable differences in character between Latin, a highly inflected language where word-order is relatively unimportant, and English, a largely uninflected language where word order matters a lot. We did parsing and analysis – dividing sentences into the various parts of speech (noun, verb, adjective, etc) clauses (main, co-ordinate and subordinate) and identifying their grammatical relations (subject and predicate, direct and indirect objects). 

All this gave me a command of written English and stood me in good stead when I went on to high school and actually learned a little Latin and rather more of its modern descendants, French and Italian – so I’m not complaining: it was an approach that served me well, though it has still left me a lot better at reading French and Italian (and writing them to some extent) than I am at speaking either and – particularly – understanding them when they are spoken. And even now, if I consider learning a language, my first impulse is to buy a grammar book: I feel safe with that, I know my way around. I shy away from speaking to people though: I’d rather acquire some degree of expertise first.

Yet I learned my own language when I was too young to be aware of doing it, without the aid of books and with no knowledge of grammar; and I learned it by talking to people who talked to me; and if my observation of young children since is anything to go by, I think the experience probably afforded me a great deal of enjoyment and even outright hilarity. Thankfully, language teaching now makes more use of these ‘natural’ methods than in my day.

Might it not be better if we devised an education system that was geared to our natural propensities for learning, rather than one which – however effective it might be for some (like me) – achieves its end by stifling those natural propensities, and with them, spontaneity and enjoyment? Must we defer gratification to learn?

*always provided you had the 19 guineas (£19-19s-0d or £19.95) that a Sunbeam A6 would cost you – the equivalent of about £1100 today.