Actually... I guess I'm wrong...
http://www.princeton.edu/~lehmann/BadChemistry.html#Skating
"Common Theory of Ice Skating is all Wet!
Water has many unusual properties. One is that the low pressure solid form (what we know as ice) has a volume per mole that is ~10% larger than that of liquid water into which it melts. An everyday consequence of this fact is that ice cubes float near the surface of water, with about 10% of their volume above the water-air surface and 90% below.
Another consequence of this decrease in volume upon melting is that the melting temperature of ice decreases when one increases the pressure on the ice. This can be rationalized by Le Châtelier's Principle. An increase in pressure on a sample of ice can be partially removed by melting the ice and thus lowering the sample volume. However, the effect is a small one in that it takes a pressure of ~121 atmospheres (1.22 MPa in SI units) to reduce the melting temperature by a mere 1 degree Centigrade.
It is often claimed that one can skate on ice because the pressure of the skate causes the ice to melt, thus dramatically reducing the friction between skate and ice. While this makes a good story, it is not quite correct. If one takes the skater to have a mass of 75 kg (weight of 165 lbs), and the skate to be 3 mm wide and 20 cm long, one can calculate that entire gravitational force exerted on the area of one skate is only a pressure of about 12 atmospheres. While one can imagine that the force is concentrated in a somewhat smaller area, the effect of pressure alone is clearly enough to shift the melting temperature of the ice by at most a few tenths of a degree. Since common experience is that ice skating is possible even when the ambient temperature is well below the normal freezing point, the pressure induced lowering of the melting point clearly does not explain this every day observations.
What is responsible then? Scientists have far from a complete understanding of this everyday phenomenon. It is likely partially related to an effect known as surface melting. The stability of solids is due to the regular structure that allows for each molecule to have multiple attractive interactions. At the surface of a solid, this is not the case, since there are no molecules 'above' the surface to bind to. As a result, the surface molecules will often distort to make the best of a bad situation by trying to increase their bonding to each other and those below. This is known as surface reconstruction. It is also known that the molecules on the surface can become disordered and liquid like at a temperature below the normal melting point of a solid, this is the phenomenon known as surface melting. Bringing up another surface (such as the metal of a skate) will influence this surface melting, since now the water molecules on the surface can bind to the metal surface atoms as well. Another important effect is friction, which can generate enough heat to melt a thin layer of ice in contact with the skate. "