Oh, and just in terms of sideways pressure, I think this is how you'd go about figuring it out - similar derivations are often used when looking at pressure differences and the speed of sound. This ignores gravity and the like, and it's probably only useful for considering ideal gases (not liquids!), seeing as I'm not accounting for how the particles with velocities on the z axis (defined as the "upwards" direction) undergo a change in momentum (this would require the surface of the water to exert a force downwards, and for the surface to be static).
Consider a particle moving at the average velocity of a substance v. When it hits a wall, a change in momentum occurs as it changes direction, and hence a force is exerted. The magnitude of this force = 2mv/t. However, as pressure is F(av)/t, you can take the time interval between hits, which should be 2x/v, where x is the length from one end of whatever it's in to the other. If you multiply this by the number of particles in the mixture, you should get the total force exerted, and if you then divide by the area, you'll find the pressure.
However, how can the sideways velocity be calculated? If the temperature is known, we can then find out the kinetic translational energy, as only the translational energy contributes to the temperature. Relationships between internal energy and degrees of freedom could then be used to find out the velocity of the water, and presuming the equipartition of velocity, the x direction of the velocity^2 = v(total)^2/3. Or something like that.
I don't think this even holds for liquids, because the intermolecular forces are too large and there's the whole problem of the downwards force at the top (it could be caused by surface tension, but I don't actually know anything at all about that, so I'm not going to make any assumptions). You could probably make a model taking these into account, but it seems like it'd end up being really fiddly...
Edit: I don't even know if gravity needs to be taken into account, because the whole pressure equation for downwards pressure takes F(net) = 0 (dp/dx = -rho * g or something...)