Another "swim faster" sceme scientifically debunked!

I couldn't make this up. 2005 Ignobel prize winner onlinelibrary.wiley.com/.../full Will humans swim faster or slower in syrup? When one of us was training for the 100 m butterfly in the U.S. Olympic Trials, we began to discuss fluid mechanics of swimming. We noted that swimmers go faster in salt water than in fresh water because they are more buoyant. We argued about how drag could be minimized coming off a turn. Most of all, we wondered whether swimmers would go faster or slower if the viscosity of the fluid was increased. We discussed this with our colleagues, but found no consensus. Most, including some who were experts in fluid mechanics, felt that the swimmers would go more slowly. Some said the swimmers would go faster, because of increased drag on the hands. A few suggested that there would be no change. We decided to measure swimming as a function of viscosity. The experimental details are as follows: We chose to thicken water with guar gum because it is readily available in a food grade and causes few allergic reactions. Xanthan was not as available to us; and corn syrup, offered as a donation, must be added at such high concentrations that it would strain the municipal sewage system. We slowly poured 310 kg of guar (Aqualon Supercol, Hercules Chemical, Wilmington, DE) into a 0.15 m3 garbage can stirred with 1 kW motor through which pool water was pumped at a rate of about 0.01 m3/s. The resulting dispersion flowed into a 650 m3 swimming pool, where it was stirred for 36 h with three submersible pumps, each moving at least 0.05 m3/s. After this mixing, the viscosity of the aqueous guar solution was (1.92 + 0.05) 10−3 Pa s, or about twice that of water. This viscosity did not vary over 16 different positions in the pool. Because the viscosity at this dilute concentration (0.05%) is Newtonian, it gave the same readings in several capillary viscometers and with different spindles of a Brookfield viscometer. The density of these guar solutions was within 10−4 g/cm3 of that of water, so buoyancy changes were insignificant. We asked 10 competitive swimmers and six recreational swimmers to swim one 25 yard length in a 1,000 m3 water-filled pool, two 25 yard lengths in the 650 m3 guar-dosed pool, and (after a shower) one 25 yard length in the water-filled pool. The swimmers rested 3 min between each length. Some competitive swimmers swam several sets, sometimes with different strokes. We recorded each swimmer's lap times, gliding time off the wall, and number of strokes. We recognize that the pools have different shapes, and that smaller, shallower pools are often felt to retard swimming. We tried to minimize these effects by having only one swimmer at a time in a fairly quiet pool, swimming in a lane next to the wall. The results, summarized in Figure 1, show that swimming in guar does not change swimming speed. This figure plots the swimmer's speed in water on the abscissa, where the recreational swimmers are slower than the competitive swimmers. It plots the speed in guar divided by the speed in water on the ordinate. The standard deviation between lengths for the recreational swimmers is 3.2%, but that for the competitive swimmers is 2.4%, the same as that recorded by their coaches in normal workouts. The smaller deviation of the competitors is probably a reflection of their superior skill and physical condition. (more on web page with charts, graphs and scientific stuff)