Harro, That's still not right. Your billiard ball model does not explain radiative heat transfer, but it does explain conductive and convective heat transfer. It appears that you are making a typical mistake. Heat is not stored. An object does not contain heat. Heat is instead what is called a process variable. Suppose you want to bring a gas from some initial pressure and volume (and hence temperature, since PV=nRT for an ideal gas) to some final pressure and volume. There are many ways to do this. For example, you can Change the volume to the final volume while keeping the pressure constant at the initial pressure, and then change the pressure to the final pressure while keeping the volume constant at the final volume, or Change the pressure to the final pressure while keeping the volume constant at the initial volume, and then change the volume to the final volume while keeping the pressure constant at the final pressure, or Make the gas follow a straight line path from the initial state to the final state, or Make the gas follow some curved path through the pressure-volume state. The amount of heat transfered to/from the system depends on the path followed. If objects contained heat, the amount of heat transfered would depend only on the difference between the final and initial state.