Scientists always try to do things no one has ever done before. In the case of low temperature physicists, 'no one' even includes nature: the coldest spot in the whole universe is located in some laboratory invented by the human mind (as far as we know). See cooling techniques for more information on how really low temperatures are produced.
The race for lower and lower temperatures has led to many new insights into the secrets of nature, and the discovery of interesting effects like superconductivity, Bose-Einstein Condensation (BEC), and superfluidity1: If helium2 (He) is cooled down towards its boiling point at 4.1K (-269.1 °C) it will become fluid and behave like a quite ordinary liquid. This state is called helium-I (helium one, or HeI for short). Cooling helium-I to even lower temperatures leads to a new state called helium-II (HeII3), which is reached at the so-called lambda point, at 2.1K.
Helium-II and helium-I are both colourless, transparent fluids, so what is so special about helium-II?
The 'Two Fluid' Model
There is a theory explaining the very unusual properties of helium-II, which we will come to later. This theory basically regards helium-II as composed of two different fluids; the normal component and the superfluid component. The normal component behaves like a normal fluid (you guessed it) having a finite (ie non-zero) viscosity4 and carrying non-zero entropy. The superfluid component, on the other hand, has zero viscosity (which means it can flow without any resistance) and carries zero entropy.
The ratio between the density of the normal component and the superfluid component depends on temperature and can vary spatially5. Both components can move independently, without interaction. The normal component tends to flow towards colder regions, while the superfluid flows towards hotter regions.
This theory is a successful one, but it is not reality (theories never are) - so beware of thinking of helium-II as a mixture of two fluids or even separating the He atoms into normal and superfluid ones.
Fast Flow of Heat
In ordinary fluids, heat transport is quite slow. This, for example, causes water in kettles to boil, as the heat can not be transported from the bottom of the kettle (where it is in contact with the hot stove) to the water surface quickly, and therefore steam bubbles are produced within the liquid. The same can be seen with helium-I, but the boiling stops immediately after the transition to helium-II has occurred.
Heat transport in helium-II is fast enough to deliver the heat to the atoms located at the surface of the liquid which then evaporate, taking the heat with them. The production of bubbles within the fluid is not possible and boiling stops.
Sound usually consists of density variations which propagate through some medium. In helium-II, a variation of the ratio between the normal component and the superfluid component can propagate heat in an analogous way. The resulting heat wave is called second sound: heaters behave like loudspeakers emitting these waves, and thermometers can act like microphones, detecting the 'second sound' heat.
In everyday life you often use things with small holes to stop water from flowing. Umbrellas or tents, for example, are usually made of textile materials, and have millions of tiny holes. The reason why these devices work quite well is that water cannot flow through very small holes because of its non-zero viscosity. If you take an unglazed earthenware pot and fill it with water or helium-I, the fluid will remain in the pot. If you try this experiment using helium-II the liquid will flow through the pot faster than water through a coffee filter, as the superfluid component can flow through the tiny pores of the pot without any resistance.
Now think of two containers, connected by a pipeline which is filled with porous material (ie a 'superleak'). If we now heat one of the containers, the superfluid will flow from the colder container through the superleak into the warmer one, while the normal counterflow is blocked by the superleak. This will cause the fluid level in the warmer container to rise, while the fluid level in the colder container drops. We have constructed some sort of a pump! If the containers are suitably shaped, it is possible to generate a nice fountain!
As the superfluid component can flow without resistance, a flow of helium-II will persist for ever once successfully established. Such experiments are usually made in rotating buckets or ring shaped containers. This effect is quite similar to the persistent electrical currents observed in superconductivity.
If you fill up a bucket with helium-II close to the rim, you will observe liquid dripping mysteriously from the bottom of the bucket. This is not due to a leak (yes, there are buckets without superleaks!) but a consequence of the zero viscosity of the fluid. Many ordinary liquids can creep up container walls a small amount6. You will probably have noticed this when looking at a glass containing water, the water at the edge creeping about 1mm above the level.
The difference in helium-II is that it can flow up the edge in very much thinner films, because of its lack of viscosity. So if the walls are not too high, the film can extend up to the top and then the fluid simply flows down the outside of the wall and drips off the bottom of the bucket.
Setting the superfluid into rotation is a quite advanced topic, as there are some issues put up by theorists which are not too easy to understand without some mathematics. Basically, they say that a compact volume of the fluid can not rotate7. But as the superfluid has been shown to behave identically to ordinary ones in rotation experiments (yes, it actually rotates!), they had to solve this discrepancy. The solution was to demand vortex lines (ie holes) in the superfluid, around which it can rotate without damaging the theory. Later these lines were actually observed experimentally. This discovery has several consequences, eg the vortex lines lead to friction between normal and superfluid flow.
Superfluids and their super properties are very interesting for scientists as they hope to understand other difficult phenomena of everyday life (eg turbulence) by studying the superfluid. Useful technical applications of superfluidity on the other hand are quite rare, as it is quite difficult - and therefore expensive - to reach the low temperatures necessary for helium to become superfluid.
1 While these effects are somewhat related to each other, there remain subtle differences between them.
2 There are different flavours (called isotopes) of elements depending on the number of neutrons they have. In nature, helium has two isotopes; 3He and 4He, and both become superfluid. This entry only covers superfluidity of 4He, as 3He becomes superfluid at much lower temperatures and is somewhat more complicated.
3 That's HE-ii, not the fiery basement, for those confused by sans serif fonts.
4 Viscosity is the 'thickness' of a fluid - for example water has quite a low viscosity while honey has a rather high viscosity.
5 Of course, the sum of both components must always give 100%.
6 Depending on the combination of materials used for the wall and the fluid (eg mercury and glass) the opposite thing can happen: the surface of the fluid is lowered at the wall. This creep up or down at the edge of a liquid is called the capillary effect and is due to the interaction between fluid atoms and wall atoms influencing the surface tension.
7 In the more precise words of the theorists this fact reads: 'The rotation of the velocity vector field is zero at any point within the superfluid. The reason for this is the existence of a macroscopic wave function which describes the superfluid component.' which sounds complicated. Actually, it is complicated.