Hi-E "Siamese" spokes are arguably the lightest spokes ever offered for sale. "Arguably" because a single spoke runs from rim to hub to rim, so to reach the "lightest spoke" weight you need to compare two ordinary spokes to one "Siamese" spoke.

The basic idea is clever: Harlan Meyer (owner-operator of Hi-E) wanted to build lighter spokes. The usual way to make a lighter spoke is to swage (thin) the center section, which needs expensive machinery not available to a small maker. Forming spoke heads and elbows also typically uses expensive machines.

Meyer bought 1.27 mm (0.050 inch) wire, of a sort suitable for spokes. He formed threads on the outside of brass(?) tubing, which he slid on and then silver-brazed to the wire. He formed heads by... doing away with the head entirely, and replacing each pair of spokes (and heads!) with one long spoke, run rim-to-hub-to-rim.

                 

The first picture of spokes laced in to a wheel are from http://bulgier.net/pics/bike/Parts/Wheels/Siamese_Spokes.JPG as of 2012/07. The second is from http://bikeville.blogspot.com/2012/06/bikes-from-cirque-part-4.html as of 2012/07, but ignore the description of construction and assembly, which is wrong. Note in the second picture, at least one spoke hole on the left flange is empty, suggesting a broken spoke pair.

Half of a Hi-E spoke (equivalent to a single conventional spoke) is 3.65 g for 311 mm equivalent, which would make a 260 mm equivalent about 3.05 g. In comparison, Sapim lists their Super Spoke as 3.61 g for 260 mm, or about 18% heavier than a Hi-E spoke.

The Hi-E "Siamese" spoke winds up in the Bicycle Museum of Bad Ideas because:

• if you break one spoke/nipple/etc., you have two slack spokes, so spoke failures are more of a problem than with conventional spokes; and,
• the way they were built, they had poor durability.

On durability, Harlan said (circa 1990) he had stopped selling the Siamese spokes. I asked if that was because the spokes wound up too much during tightening and broke? He said no, and described a truing stand made of an old barber's chair with a foot treadle to push radially inward on the rim to de-tension the spoke while adjusting it.

Instead, he said they broke fairly quickly where the threads were brazed on to the wire. That is probably because the wire forming process draws the wires through dies to shape it, and in doing so aligns the metal grains for high strength, but heating tends to change the grain structure, weakening it. Even at relatively low temperatures, there is some change in grain structure — as demonstrated, most likely, but the short life of the spokes.

As an aside, Hi-E siamese spoke elbows appear to be slightly wider than a conventional headed spoke. If so, and if spoke threads were not the main source of premature failure, the "elbows" might have been. It is unclear if the bends could be placed closer and still fit through a spoke hole, but probably would be possible with nonstandard spoke holes, and thus need not be an intrinsic problem.

Hi-E spoke wire is so thin that ordinary truing tension can make it wind up and snap off. That's why Harlan built the truing stand which de-tensioned the spoke to adjust it. For field failures, it may be possible to push or pull sideways on the rim to achieve a similar effect, but there is also some risk of snapping spokes. And, as noted above, with "Siamese" spokes, a single failure counts for two.

As another aside, Hi-E also offered the lightest nipples, about 0.19 g per nipple:

        

(The picture of nipples with a wrench is from http://velobase.com/CompImages/SmallParts/D674A10D-134E-41CD-B536-72A6241B8A83.jpeg as of 2012/07.)

The nipples are made from hex (six-sided) high-strength aluminum. They weigh less and have better aerodynamics than most nipples, but do require dismounting the tire to adjust spoke tension.