264 ANNUAL REPORT SMITHSONIAN INSTITUTION, 193 5 



gas exists, because there are forces between the atoms that cause the 

 system to pass over to a state of order — the crystal — and there is no 

 longer any disorder that we could make use of for lowering the tem- 

 perature. In the paramagnetic salt the same thing happens, only at 

 temperatures about 100 times lower. In them there also exist 

 interaction forces which have the effect of establishing an order 

 within the system without an external magnetic field, and so at these 

 temperatures the paramagnetic salts cease to be of use. These tem- 

 peratures differ, of course, from substance to substance, and we saw 

 in our experiments that the manganese ammonium sulphate passed 

 over into this state earlier than the iron ammonium alum. Of 

 course, one could get a bit further by using still stronger magnets, 

 but I think the practicable limit has been reached with the dimen- 

 sions of the Leiden magnet. 



A more hopeful way seems to be to work in two stages. That 

 means, first work down to about 0.05°, and then, starting at this tem- 

 perature, go still lower with a new procedure, like the cascade for 

 liquefying gases, which was used so much in the past. Of course, to 

 get to appreciably lower temperatures for the lower stage it will be 

 necessary to find a substance in which these interaction forces, which 

 tend to bring the system into an ordered state, are still smaller than 

 within the substances hitherto used. I have already said that the in- 

 teraction forces between the nuclear moments and their surroundings 

 are very small, and in the second stage one will have to tr}?^ to work 

 with a substance that exhibits nuclear paramagnetism. 



But even here there are some interaction forces, and so this will 

 only work for a bit of the way to absolute zero. And with all other 

 phenomena that may still happen in the new temperature region, it 

 will certainly be the same. There exists a law, Nernst's Theorem, also 

 called the third law of thermodynamics', which is confirmed by all 

 experiments. It postulates that at absolute zero all substances are in 

 a state of perfect order, or, in other words, that the state of lowest 

 energy must be a state of perfect order. And we know now that this 

 means that it will be impossible ever to reach the zero of temperature 

 absolutely. But this does not mean that one cannot get below a cer- 

 tain limit, say 1/10,000°. It will be possible to reach any finite tem- 

 perature, be it as small as you like. But the technique of reaching 

 such a temperature will always be dependent on finding a phenome- 

 non, connected with only a very small energy change, happening 

 within a system. And so you see that this last degre-e, or, as we may 

 say now, the last 1/100° to absolute zero, though absolutely very 

 small, stretches in reality an infinite distance before us. And this 

 infinity is not an empty one, but one that is filled with phenomena 

 worth investigating. 



