260 ANNUAL, REPORT SMITHSONIAN INSTITUTION, 1935 



But in order to go lower, how can we proceed? As we discussed 

 before, to get a low temperature one must have a system at one's 

 disposal which is still in a big state of disorder, and one must be 

 able to change this disorder by changing an external variable. At 

 these temperatures we no longer have gases. What other disordered 

 systems still exist? Well, we have just seen one, namely, the para- 

 magnetic salt which still follows Curie's law; and about 10 years 

 ago Debye and Giauque proposed using this for the generation of 

 still much lower temperatures. 



To understand the principle, look at figure 5, which represents a 

 paramagnetic salt. The circles represent the atoms arranged in a 

 crystal lattice, and the little arrows represent the magnetic mo- 

 ments attached to every atom. Without a field (fig. 5, A) there ex- 

 ists, as we saw before, a random distribution of the directions of the 



000© 000® 

 0000 0000 



B D 



Figure 5. — Principle of tbe magnetic cooling method. 



dipoles, so the disorder in this system consists of two parts; one is 

 due to the distribution of the directions of the dipoles, and the 

 other to the thermal motion of the atoms. (The diameters of 

 these circles indicate the magnitudes of the vibrations and give 

 also in this way a measure for the disorder due to the thermal vibra- 

 tions.) Applying the field now (fig. 5, B) it will try to order 

 the directions of the dipoles. Making it adiabatic — that is, having 

 the system isolated thermally from its surroundings — the whole state 

 of order should remain constant, which means that the disorder due 

 to the thermal movement must increase, in other words, the tem- 

 perature rises (corresponding to the compression of the gas in our 

 former example). Making heat contact with the surroundings, the 

 system cools down to the initial temperature, so that now the state of 

 order has increased (fig. 5, C). Now, isolating the substance from 

 the surroundings and taking the field away, the dipoles try to dis- 

 tribute their directions at random again, the disorder due to this 



