I should clarify that the sense of energy flow (from larger temperature to smaller temperature) is true for positive temperature systems. If you were to bring a negative temperature system into contact with a positive temperature system then energy will flow from the negative temperature system to the positive one - the equilibrium will be when both systems have a positive temperature. So, in a sense, negative temperatures are HOTTER than positive temperatures. If you think about why they are negative in the first place: entropy *decreases* as you increase the internal energy, it makes sense. For normal, positive temperature systems, if you increase the energy of a low temperature system its entropy increases more than the amount of entropy that is lost when the hotter system loses the energy - so energy flows from the more positive temperature system to the lower temperature system. But for a negative temperature system the entropy will increase if it loses energy.<br /><br />In the sense of how energy will flow, the cold to hotness scale runs from:<br /><br />+0 K, .... , +1000 K, ..... , + infinity K, - infinity K, .... , -1000 K, -0 K<br /><br />where -0 K is the "hottest" something can be (At -0 K, 1/T = d(entropy)/d(energy) = -infinity, so the system gains infinite entropy by undergoing a slight decrease in energy).<br /><br />If you're interested, this is taken from the appendix of "Thermal Physics" by Kittel & Kroemer. <div class="Discussion_UserSignature"> </div>