November 25, 1922] 



NA TURE 



703 



solid, these decimetre strings gut end-to-end will reach 

 6-3 million million miles, the distance light will travel 

 in a year, a quarter of the distance to the nearest fixed 

 star. If the atoms are spaced but one millimetre apart 

 the string will be three and a half million times longer 

 yet, spanning the whole universe. 



Again, if an ordinary evacuated electric light bulb 

 were pierced with an aperture such that one million 

 molecules of the air entered per second, the pressure 

 in the bulb would not rise to that of the air outside 

 for a hundred million years. Perhaps the most striking 

 illustration is as follows : Take a tumbler of water 

 and — supposing it possible — label all the molecules 

 in it. Throw the water into the sea, or, indeed, 

 anywhere you please, and after a period of time so 

 great that all the water on the earth — in seas, lakes, 

 rivers, and clouds — has had time to become perfectly 

 mixed, fill your tumbler again at the nearest tap. 

 How many of the labelled molecules are to be expected 

 in it ? The answer is, roughly, 2000 ; for although 

 the number of tumblerfuls of water on the earth is 

 5 x io 21 , the number of molecules of water in a single 

 tumbler is io 25 . 



From the above statements it would, at first sight, 

 appear absurd to hope to obtain effects from single 

 atoms, yet this can now be done in several ways, and 

 indeed it is largely due to the results of such experiments 

 that the figures can be stated with so much confidence. 

 Detection of an individual is only feasible in the case 

 of an atom moving with an enormous velocity when, 

 although its mass is so minute, its energy is quite 

 appreciable. The charged helium atom shot out by 

 radioactive substances in the form of an alpha ray- 

 possesses so much energy that the splash of light 

 caused by its impact against a fluorescent screen can 

 be visibly detected, the ionisation caused by its passage 

 through a suitable gas can be measured on a sensitive 

 electrometer and, in the beautiful experiments of 

 C. T. R. Wilson, its path in air can be seen and photo- 

 graphed by means of the condensation of water drops 

 upon the atomic wreckage it leaves behind it. 



In the first complete Atomic Theory put forward 

 by Dalton in 1803 one of the postulates states that : 

 " Atoms of the same element are similar to one another 

 and equal in weight." Of course, if we take this as 

 a definition of the word " element " it becomes a 

 truism, but, on the other hand, what Dalton probably 

 meant by an element, and what we understand by 

 the word to-day, is a substance such as hydrogen, 

 oxygen, chlorine, or lead, which has unique chemical 

 properties and cannot be resolved into more elementary 

 constituents by any known chemical process. For 

 many of the well-known elements Dalton's postulate 

 still appears to be strictly true, but for the others, 

 probably the majority, it needs some modification. 



Throughout the history of science philosophers have 

 been in favour of the idea that all matter is composed 

 of the same primordial substance, and that the atoms 

 of the elements are simply stable aggregations of atoms 

 of this substance. Shortly after Dalton's theory had 

 been put forward Prout suggested that the atoms of 

 the elements were composed of atoms of a substance 

 he called " protyle," which he endeavoured to identify 

 with hydrogen. 



If Dalton and Prout were both right the combining 



NO. 2769, VOL. IIO] 



weights of the elements should all be expressible as 

 whole numbers, hydrogen being unity. Experimental 

 evidence showed this to be impossible in many cases. 

 Chemists therefore wisely preferred Dalton's theory, 

 which was in accord with definite though fractional 

 atomic weights, to Prout's, which would necessitate 

 the elements of fractional atomic weight being hetero- 

 geneous mixtures of atoms of different weight. 



The idea that atoms of the same element are all 

 identical in weight could not be challenged by ordinary 

 chemical methods, for the atoms are by definition 

 chemically identical, and numerical ratios were only to 

 be obtained in such methods by the use of quantities 

 of the element containing countless myriads of atoms. 

 At the same time it is rather surprising, when we 

 consider the complete absence of positive evidence in 

 its support, that no theoretical doubts were publicly 

 expressed until late in the nineteenth century, first by 



Fig. 3. — Cube 26 showing atoms with scale of reference. 



Schutzenberger and then by Crookes, and that these 

 doubts have been regarded, even up to the last few- 

 years, as speculative in the highest degree. In order 

 to dismiss the idea that the atoms of such a familiar 

 element as chlorine might not all be of the same weight, 

 one had only to mention diffusion experiments and the 

 constancy of chemical equivalents. It is only within 

 the last few years that the lamentable weakness of 

 such arguments has been exposed and it has been 

 realised that the experimental separation of atoms 

 differing from each other by so much as io per cent, 

 in weight, is really an excessively difficult operation. 



There are two ways by which the identity of the 

 weights of the atoms forming an element can be tested. 

 One is by the direct comparison of the weights of 

 individual atoms ; the other is by obtaining samples 

 of the element from different sources or by different 

 processes, which, although perfectly pure, do not give 

 the same chemical atomic weight. It was by the 

 second and less direct of these methods that it was 

 first shown by the experiments of Soddy and others 

 on the atomic weight of lead from different radioactive 



