As Joel pointed out, the bikes in the trials were weighed by the _organizers_, not the _builders_. Before the war, the entire bike was weighed. After the war, good tires were not available, so the bikes were weighed minus tires and bags. It's not rocket science to make a bike light - Singer had a guy who did nothing but make stems, racks (for production bikes) and parts for trials bikes. Full time. Also, a good Singer rack weighs about 100-120 grams (front), while the fenders don't add more than 2 lbs. in standard form. Generators are heavy, but the lights themselves are featherweight. So if you can build a 12 lb. racer, a 16 lb touring bike is doable as well.
To quote from Joel's charts: The lightest bike in the 1938 Concours du Duraluminum, the Narcisse of Rene Herse, weighed 7 kg 940 for a 55 cm frame (Le Cycliste, October 1938). That translates into 17.468 lbs. for a bike with fat 650 B tires. I haven't seen the official results for 1939, where Singer won with a bike that beat all the others - check Raymond Henry's story in the proceedings of the Bicycle History Conference for the exact weight.
Take a Stronglight crank, already quite light, file off 30% of the metal on the arms, then file off the protrusion with the threads for the crank puller, and so on, and do the same for every part, and soon you get there.
I have seen a few parts that either were for trials bikes or modified by somebody who wanted to recreate a trials Singer in the late 1940s, and it is amazing how much was taken off many of the parts.
I would like to thank Joel Metz for sharing his priceless library for the upcoming newsletter - which (shameless self-promotion) is coming along nicely - first issue due in August.
Jan Heine, who still doesn't care all that much about weight but
would like to organize technical trials some day in Seattle - any
takers?
>From: "joel metz, ifbma/sfbma" <magpie@messengers.org>
>Subject: Re: [CR]Technical trials > scepticism regarding weights...
>
>weights are listed to the
> nearest .005 kg, km/hr to 2 decimal places, etc. there were
> formulas to calculate weight penalties based on frame
weight compensated for frame size... things like that.