252 ANNUAL. REPORT SMITHSONIAN INSTITUTION, 19 3 5 



the liquid nitrogen necessary for the hydrogen liquefaction was 

 generated in yet another plant. For the first experiment, on the 

 other hand, I could get the whole plant in my hand by taking a 

 torchlight. 



Now, what makes this big difference ? Heat is due to the irregular 

 motion of the smallest particles. So to heat a substance means to 

 increase its energy but still more to increase the internal disorder 

 of its particles. To cool down a substance means to diminish its 

 energy but still more to increase the internal order of its particles. 

 For example, a gas is in a very big state of disorder, the atoms 

 flying in all directions about the space according to their kinetic 

 energy, which is proportional to the temperature. Cooling it down, 

 it will first liquefy, and a liquid is already a much more ordered 

 system than a gas. Cooling it still further, it will solidify; now 

 all the mean positions of the particles are given by a crystal lattice, 

 and the system is in a state of nearly perfect order, only the thermal 

 vibrations of the atoms around their mean positions being still a 

 source of some disorder. Cooling down still further, even this 

 disorder vanishes more and more. 



So cooling a substance means bringing it into a state of order, 

 and it is always much easier to make disorder than to establish order. 

 Although you are certainly very familiar with this fact, I will illus- 

 trate it by an example. On this tray I have arranged these black 

 and white balls in order. It is very easy now to establish a dis- 

 order by shaking the tray, but it is impossible to establish any sort 

 of order again by shaking, or, to speak more precisely, the proba- 

 bility of succeeding is extremely small. Of course, I can establish 

 an order by selecting the balls with my hands; but in the case of a 

 system consisting of atoms, this is obviously impracticable and, by 

 principle, impossible, too. We have only macroscopic means at our 

 disposal. Tliis difficulty of creating order is just what makes the 

 big difference. And this is why, even 10,000 years ago, men were 

 able to generate very high temperatures — sufficient for melting 

 metals — but the low-temperature technique is of a quite recent date. 

 Even the ice-cream industry is fairly new, though the temperatures 

 involved are not very much below room temperature. 



To generate heat one must, of course, have energy at one's dis- 

 posal, but to transform this energy into the disordered form of heat 

 energy presents no difficulty at all. We saw this in the electric lamp, 

 but one can do it in a great many other ways; for instance, in a 

 candle it is a chemical energy we transform into heat energy, in the 

 brakes of a car it is a mechanical one. 



If our example with the balls were absolutely analogous it would 

 be impossible to generate a low temperature at all, but luckily it is 



