Apologies if this summary is a little expedient - I'm intellectually lazy! I think I can safely quote verbatim from the article where it needs it. Please note there are extensive illustrations, however this article is pulled from a database with a user agreement, and I believe I can't reproduce them without permission. I'll try to list the techniques that differ significantly from each other.
Frictional Fire-Making with a Flexible Sawing Thong.
Henry Balfour
The Journal of the Royal Anthropological Institute
Jan-Jun 1914, pp.32-64
Summary of techniques described:
The General Principle
A piece of bamboo or wood is held in a fixed position and a flexible band or thong of cane or similar is looped around it. By pulling with a sawing motion on both ends of the thing enough friction is generated to create a mass of dust, which if hot enough, will ignite and smoulder. This smouldering dust is then brought to flame with a more combustible material.
Regional variants:
Naga Hills, Assam method
"a long piece of cane passed under a dry log and pulled (i.e, the ends) alternately by the right and left hand, so as to ignite some tinder placed in a hollow or split underneath" Balfour describes a Naga fire-making set in further detail, with a 'billet' of wood made of lime, split open at one end and wedged apart with a stone. the thong is a narrow sliver of bamboo. Tinder is also bamboo shavings. the billet is put on the ground and held down with the foot. The illustrations suggest the tinder was placed down first, then the thong on top of that and the the billet on top of the thong. From the description I can't imagine what happens to the punk - does it fly out everywhere or end up in four little piles on either side?
Chittagong Hills Tipperah method
"a piece of dry bamboo about a foot long, split in half, and on it's outer round surface cut a nick or notch, about 1/8th inch broad, circling round the semi-circumference of the bamboo, shallow towards the edges, but deepening in the centre, until a minute slit about a lone in breadth pierces the inner surface" - a flexible strip of bamboo is used, first rubbed with fine dry sand and looped around the outside edge of the stick, sitting in the notch. The stick is then held by both feet on the ground while the thong is worked back and forth. Then punk collects in the notch on the inner and uppermost surface of the bamboo and once ignited, transferred to the tinder.
Annam (China?) Mois method
Same as Tipperah method with addition of toggles on the thong at each end. the author specifically notes the use of green bamboo 'rind' for the thong, which must be produced in large numbers, as 30-40 thongs are quickly expended to get a coal. These people apparently usually carried fire, or used the fire piston.
Malay Peninsula
Employed a billet of wood again, "split ... at one end so as to form a cleft of about 6 inches. In this he inserted a small stick, which formed a peg separating the two halves and standing above the surface of the billet". Tinder 'palm scurf' [which I read as palm fibre?] was stuffed into the cleft between the peg and the thong at the base of the split. The thong is rattan, with sticks fasted at each end.
Then there are described several regional variants of this technique.
Borneo method
The billet or stick of soft fibrous wood (the illustration looks more like a carved stick) is slotted instead of split. Bamboo is again used for the thong. The stick is held down on the ground with both feet on top of a bed of fine wood slivers. A couple of variants are described, including an illustration showing the stick being used in a vertical position, apparently lashed to a tree at top end, with the operators foot applying pressure to the base of the stick.
Again, more regional variants
Moving on to Papua New Guinea, where he describes a variant split stick with stone peg idea, but which is actually two sticks lashed together and then forced apart at one end with a stone. This variation is widespread. Flexible rattan cane used for a thong: "The thong saw is a very long strip of rattan, which is kept neatly coiled up in a thick ring. The tinder is usually a piece of the fibrous sheath of palm-shoot, or a piece of dry moss. The tinder is placed in the narrow part of the cleft, behind the stone".
Lots for variants on this theme - two stick, one split stick, lots of short thongs, one long coiled thong etc...
On to a slightly different New Guinea method found in the Koiari people, where (this is hard to describe) the thong, instead of being just looped around the split stick, is passed right around again so that two friction notches appear on the underside of the stick. In variants of this the stick is held down by the knee.
Jumping to Africa now... the evidence the author had to hand gets a little hazy... quickly moving on to Sweden (see i warned you I was intellectually lazy!)
Sweden and the flexible fire *saw*:" a length of rope, willow strip or strap is passed completely round a log of wood, and the ends are pulled alternately by two men until a spark is generated, and can be caught upon a piece of tinder" or "a bast rope is is wound twice round a dry branch of birch or aspen, on which the bark remains, or else round a dry tree-stump. The cord was pulled by two men" . Either way the illustrations look like hard work and might make an amusing diversion at a BCUK meet! A one man version is also described when the loop is only half way around the log.
An interesting swedish variant - the thong, material not specified, is stretched taught (one end fixed to something and the other held by a second man) and the stick is sawn back and forth with both hands.
Author then digresses into cord saws and thong drills and a lengthy description of the dispersal of these techniques around the world.
I'm sleepy now. Hopefully Match will get permission to reproduce it on his site, so you can see the illustrations. If you've a serious interest PM me.