52 



NATURE 



[September 8, 192 1 



pf the latter class, the number of hydrogen nuclei 

 Associated with the helium nuclei is invariably 

 three, except in that of nitrogen, where it is two. 

 The frequent occurrence of this group of three 

 hydrogen nuclei indicates that it is structurally an 

 isotope of hydrogen with an atomic weight of 

 three and a nuclear charge of one. It is surmised 

 that it is identical with the hypothetical " nebu- 

 Ijum " from which our " elements " are held by 

 astro-physicists to be originally produced in the 

 stars through hydrogen and helium. 

 : These results are of extraordinary interest as 

 bearing on the question of the essential unity of 

 rnatter and the mode of genesis of the elements. 

 Members of the British Association may recall the 

 suggestive address chi this subject of the late Sir 

 William Crookes, delivered to the Chemical Sec- 

 tion at the Birmingham meeting of 1886, in which 

 he questioned whether there is absolute uniformity 

 in the mass of the atoms of a chemical element, 

 as postulated by Dalton. He thought, with Marig- 

 nac and Schutzenberger, who had previously raised 

 the same doubt, that it was not improbable that 

 what we term an atomic weight merely represents 

 a mean value around which the actual weights of 

 the atoms vary within narrow limits, or, in other 

 words, that the mean mass is " a statistical con- 

 stant of great stabihty." No valid experimental 

 evidence in support of this surmise was or could 

 be offered at the time it was uttered. Maxwell 

 pointed out that the phenomena of gaseous dif- 

 fusion, as then ascertained, would seem to nega- 

 tive the supposition. If hydrogen, for example, 

 were composed of atoms of varying mass, it 

 should be possible to separate the lighter from the 

 heavier atoms by diffusion through a porous sep- 

 tum. " As no chemist," said Maxwell, " has yet 

 obtained specimens of hydrogen differing in this 

 way from other specimens, we conclude that all 

 the molecules of hydrogen are of sensibly the same 

 rnass, and not merely that their mean mass is a 

 statistical constant of great stability." ^ But 

 against this it may be doubted whether any 

 chemist had ever made experiments sufficiently 

 precise to solve this point. 



The work of Sir Norman Lockyer on the spec- 

 troscopic evidence for the dissociation of " ele- 

 mentary " matter at transcendental temperatures, 

 and the possible synthetic intro-stellar production 

 of elements, through the helium of which he ori- 

 ginally detected the existence, will also find its due 

 place in the history of this new philosophy. 



Sir J. J. Thomson was the first to afford direct 

 evidence that the atoms of an element, if not 

 exactly of the same mass, were at least approxi- 

 mately so, by his method of analysis of positive 

 rays. By an extension of this method Dr. F. W. 

 Aston has succeeded in showing that a number 

 of elements are in reality mixtures of isotopes. It 

 has been proved, for example, that neon, which 

 has a mean atomic weight of about 20.2, consists 

 of two isotopes having the atomic weights respec- 



1 Clerk Maxwell, art. " Atom," Ency. Brit., 9th ed. 



NO. 2706, VOL. 108] 



tively of 20 and 22, mixed in the proportion of 

 90 per cent, of the former with 10 per cent, of the 

 latter. By fractional diffusion through a porous 

 septum an apparent difference of density of 0*7 

 per cent, between the lightest and heaviest frac- 

 tions was obtained. The kind of experiment which 

 Maxwell imagined proved the invariability of the 

 hydrogen atom has sufficed to show the converse 

 in the case of neon. 



The element chlorine has had its atomic weight 

 repeatedly determined, and, for special reasons, 

 with the highest attainable accuracy. On the 

 oxygen standard it is 35'46, and this value is 

 accurate to the second decimal place. All attempts 

 to prove that it is a whole number — 35 or 36 — have 

 failed. When, however, the gas is analysed by 

 the same method as that used in the case of neon 

 it is found to consist of at least two isotopes of 

 relative mass 35 and 37. There is no evidence 

 whatever of an individual substance having the 

 atomic weight 35'46. Hence chlorine is to be 

 regarded as a complex element consisting of two 

 principal isotopes of atomic weights 35 and 37 pre- 

 sent in such proportion as to afford the mean 

 mass 35'46. The atomic weight of chlorine has 

 been so frequently determined by various observers 

 and by various methods with practically identical 

 results that it seems difficult to believe that it con- 

 sists of isotopes present in definite and invariable 

 proportion. Dr. Aston meets this objection by 

 pointing out that all the accurate determinations 

 have been made with chlorine derived originally 

 from the same source, the sea, which has been 

 perfectly mixed for aeons. If samples of the ele- 

 ment could be obtained from some other original 

 source, it is possible that other values of atomic 

 weight would be obtained, exactly as in the case 

 of lead, in which the existence of isotopes in the 

 metal found in various radioactive minerals was 

 first conclusively established. 



Argon, which has an atomic weight of 3988, 

 was found to consist mainly of an isotope having 

 an atomic weight of 40, associated to the extent 

 of about 3 per cent, with an isotope of atomic 

 weight 36. Krypton and xenon are far more com- 

 plex. The former would appear to consist of six 

 isotopes, 78, 80, 82, 83, 84, 86; the latter of fivcv 

 isotopes, 129, 131, 132, 134, 136. 



Fluorine is a simple element of atomic weight 

 19. Bromine consists of equal quantities of two 

 isotopes, 79 and 81. Iodine, on the contrary, 

 would appear to be a simple element of atomic 

 weight 127. The case of tellurium is of special 

 interest in view of its periodic relation to iodine, 

 but the results of its examination up to the present^ 

 are indefinite. 



Boron and silicon are complex elements, each 

 consisting of two isotopes, 10 and 11, and 28 and 

 29, respectively. 



Sulphur, phosphorus, and arsenic are apparently 

 simple elements. Their accepted atomic weights 

 are practically integers. 



All this work is so recent that there has been 



