,,
ISOTOPES
BY
F. W. ASTON, M.A., D.Sc., A.I.C., F.R.S
*
FELLOW OF TRINITY COLLEGE, CAMBRIDGE
LONDON
EDWARD ARNOLD & CO.
1922
[All rights reserved]
PREFACE
I HAVE undertaken the preparation of this book on Isotopes
in response to many requests made to me by teachers of physics
and chemistry and others working in these subjects that I
should publish the results obtained by means of the Mass-
spectrograph in a form more convenient to the public than that
in which they first appeared. This is one of the reasons why
the space allotted to the inactive isotopes may appear, in the
light of the general title of the book, somewhat disproportion-
ately large. Another is that the subject of radioactive isotopes
really requires a book to itself, and I am in the hope that the
inadequacy of^my account may stimulate the production of
such a volume by hands more competent than mine to deal
with this very special and remarkable field of modern science.
The logical order of exposition of a scientific subject is to start
with the simple and from that build up the more complex.
Unfortunately the sequence of events in experimental research
is the exact opposite of this so that a compromise must be
effected, unless one is content to sacrifice historical treatment
altogether. The latter seems very undesirable in a new subject.
I have endeavoured in Chapters I, II and IV, and elsewhere
when possible, to adhere strictly to the historical order of
events even at the cost of some reiteration.
I wish to take this opportunity of expressing my indebted-
ness to Mr. C. G. Darwin for his timely criticism and unfailing
assistance throughout the work, and also to Mr. R. H. Fowler
for help with the proofs. My thanks are also due to Professor
Soddy for his diagram of the radioactive isotopes, to Mr. A. J.
Dempster for kindly sending me the illustrations of his work,
iii
iv PREFACE
to the proprietors of the Philosophical Magazine and to
the Council of the Chemical Society for permission to use the
plates and figures of my original papers, and to Messrs.
Macmillan & Co., for the diagram of the radioactive trans-
formations.
F. W. ASTON.
CAMBRIDGE,
January, 1922.
CONTENTS
CHAPTER I
INTRODUCTION PAGE
1. Introduction ......... 1
2. Hypothesis of Dalton and Prout . ... . .2
3. Crookes' Meta-elements ....... 4
4. The discovery of Isotopes ...... 6
CHAPTER II
THE RADIOACTIVE ISOTOPES
5. Chemical identities among the radioactive elements . . 7
6. Spectroscopic identity of isotopes ..... 9
7. The chemical law of Radioactive change . . .11
8. Isobares 12
9. The Radioactive Transformations . . . . .13
10. The Atomic weight of Lead 16
11. Atomic weights of Thorium and Ionium . . .18
12. Use of radioactive isotopes as indicators ... 19
13. Classification of the radioactive isotopes . . . .21
CHAPTER III
POSITIVE RAYS
14. Nature of Positive Rays . 22
15. Mechanism of the electric discharge in gases at low pressure . 23
16. The Crookes Dark Space 24
17. Methods of detecting positive rays ..... 26
18. Sir J. J. Thomson's "Parabola" method of analysis . 25
19. Secondary rays ........ 29
20. Negatively charged rays ...... 29
21. Rays with multiple charges . . . . . .30
22. Dempster's method of positive ray analysis . . .31
CHAPTER IV
NEON
23. Positive ray analysis of neon ...... 33
24. Apparatus for the determination of density ... 35
25. Method of using the density balance ., . . .36
26. Experiments on separation by distillation ... 37
27. Experiments on separation by diffusion . . . .39
28. Second attempt at separation by diffusion . . .41
29. The analysis of neon by the Mass-spectrograph . . 41
vi CONTENTS
PAGE
CHAPTER V
THE MASS-SPECTROGRAPH
30. Limitations of the parabola method . . . .43
31. Methods of increasing the intensity of the spot . . 44
32. Possibilities of " focussing " 44
i/33. Principle of the Mass-spectrograph . . . . .44
*/ 34. Optical analogue ........ 46
/ 35. The discharge tube 47
36. The slit system 49
37. The electric field 50
^ 38. The magnetic field 51
39. The camera 51
^ 40. Experimental procedure ....... 52
41. Form of the Spectrum Lines . . . . . .53
42. The distribution of the mass-spectrum over the photo-
graphic plate ........ 54
^43. Practical method of deducing the effective mass of a particle
from the position of its line on the photograph . 55
44. Comparison of masses by the method of coincidence . 57
45. The measurement of the lines ..... 69
46. Resolving power and accuracy of mass determination . 60
47. Order of results and nomenclature . . . .61
48. Lines of the first, second and higher orders . . . 61
49. Negative mass-spectra ....... 62
CHAPTER VI
ANALYSIS OF THE ELEMENTS
50. Arrangement of results . . . . . .63
51. Oxygen and carbon ....... 63
62. Neon 64
63. Possibility of a third isotope of neon . . . .65
64. Chlorine . . . .... .65
55. Argon ........ .66
66. Nitrogen .67
67. Hydrogen and helium ....... 67
58. The determination of the masses of atoms of hydrogen and
helium by the method of " Bracketing " . . . 69
59. Triatomic hydrogen ..... .70
60. Krypton and Xenon ... .70
61. Mercury 72
62. Boron, Fluorine, Silicon . . 72
63. Molecular lines of the second order . . .75
64. Bromine . . 76
65. Sulphur . .76
66. Phosphorus. Arsenic ..... .77
67. Selenium. Tellurium 77
68. Iodine .78
69. Antimony 78
70. Tin 78
71. Nickel . 79
CONTENTS vii
CHAPTER VII
ANALYSIS OF THE ELEMENTS (Continued) . PAGE
72. Positive rays of metallic elements . . . .80
73. Dempster's analysis of Magnesium ..... 80
74. The Mass-spectra of the alkali metals . . . .83
75. Experiments with the parabola method of analysis . 84
76. Lithium 86
77. Sodium 86
78. Potassium 87
79. Rubidium 87
80. Caesium .87
81. Thomson's work on Beryllium ..... 88
82. Calcium and Strontium . . . . . . .88
83. Table of Elements and Isotopes . *. . .88
CHAPTER VIII
THE ELECTRICAL THEORY OF MATTER
84. The whole number rule ....... 90
85. The unitary theory of the constitution of matter . . 90
86. The atom of negative electricity, or electron ... 91
87. The atom of positive electricity, or proton ... 92
88. The nucleus atom 92
89. Moseley's atomic numbers ...... 93
90. The Bohr atom 95
91. The Lewis Langmuir atom ...... 95
92. Diagrammatical representation of atoms of Isotopes and
Isobares ........ 96
93. The relation between Isotopes and Elements in the same
Group 98
94. Abnormal compounds formed by charged atoms . . 98
95. The failure of the additive law in respect to mass . . 99
96. The explanation of the fractional mass of the hydrogen
atom by the hypothesis of " packing " . . 100
97. The structure of the nucleus . . . . . .101
98. Cosmical effects due to change of mass . . . .103
99. The stable systems of protons and electrons known to
occur ......... 105
CHAPTER IX
ISOTOPES AND ATOMIC NUMBERS
100. The relation between chemical atomic weight and atomic
number ......... 108
101. Statistical relations exhibited by elements and their isotopes . 109
102. The preponderance of elements of even atomic number . Ill
103. The constancy of chemical atomic weights . . .112
104. The agreement between the chemical atomic weight and
the mean atomic weight deduced from the mass-
spectrum . . . . . . . .113
105. The meaning of the word "element" . . . .115
106. Disintegration theory of the evolution of the elements . 116
107. Crookes' theory of the evolution of the elements . .117
viii CONTENTS
PAGE
CHAPTER X
THE SPECTRA OF ISOTOPES
108. The spectra of isotopes . . . . . . .121
109. The magnitude of the gravitational effect . . .121
110. Deviation of the Bohr orbits due to change in the position of
the centre of gravity of the rotating system . .122
111. Later experiments of Aronberg and Merton . . .123
112. "Isotope " effect on the infra-red spectrum of molecules . 125
CHAPTER XI
THE SEPARATION or ISOTOPES
113. The separation of isotopes 127
114. Separation by diffusion . . . . . .127
115. The separation of the isotopes of chlorine by the diffusion
of HC1 . . 129
116. Separation by thermal diffusion ..... 129
117. Separation by gravitation or "pressure diffusion" . .131
118. Separation by chemical action or ordinary fractional
distillation 133
119. Separation by evaporation at very low pressure . . 134
120. Separation of the isotopes of chlorine by free evaporation . 136
121. Separation by positive rays 136
122. Separation by photochemical methods . . . .137
123. Other methods of separation and general conclusions . 138
APPENDIX I. TABLE OF ATOMIC WEIGHTS AND ISOTOPES OF
THE ELEMENTS . . . . .141
II. THE PERIODIC TABLE OF THE ELEMENTS . 144
III. RECENT RESULTS OBTAINED BY DEMPSTER 146
LIST OF PLATES
PLATE I. Positive Ray Parabolas . . . To face page 28
II. Original Mass-Spectrograph . . . ,,46
III. Mass Spectra . . . . . ,,66
IV. Mass Spectra ,,72
CHAPTER I
INTRODUCTION
1. Introduction. Towards the end of the last century
the attitude of science in relation to the atomic theory started
to undergo a complete and radical change. What had been
before regarded as a convenient working hypothesis became
with remarkable rapidity a definite statement of fact.
This transformation is now complete and in any well-
equipped laboratory to-day not only can individual atoms be
detected but the movements of the swiftest of them can be
tracked and made visible even to the untrained eye.
The causes of this remarkable advance are to be ascribed in
particular to the discovery of radioactivity, which has provided
us with atomic projectiles possessing enough energy to produce
visible and measurable effects individually, and in general to
the steady and continuous improvement in technical methods.
Subject to such unprecedented scrutiny it was to be expected
that the fundamental physical theories which underlie the
applied science of chemistry and form a solid mathematical
foundation for its formulae, might show hitherto unsuspected
flaws. Such expectations began to be realised when, among
the radioactive elements, Boltwood failed to separate ionium
from thorium, and, among the inactive elements, when Sir
J. J. Thomson a few years later observed the anomalous
behaviour of neon when subjected to positive ray analysis.
Further and still more delicate and careful scrutiny of these
flaws revealed them, as it must always do, if they are real, not
as fortuitous and disconnected but as a definite and ultimately
intelligible pattern. It is with the interpretation of this
pattern, so revealed, that this volume is concerned, so that it
will be of interest to look back rather over a century to the
1 B
2 ;: : .V.ISOTOPES
beginning of the theories which form the background against
which it was jfirst observed.
2. Hypotheses of Dalton and Prout. In the generalisa-
tion, known as the Atomic Theory, put forward by Dalton in
1803, which laid the foundations of the whole of modern
chemistry, five postulates were laid down, and it is a striking
tribute to the shrewd intuition of that observer that, of those
five, to this day, the validity of one only is in any question.
This postulate is that : Atoms of the same element are similar
to one another and equal in weight. It obviously consists
of two parts and if we combine both as a definition of the word
element the whole becomes a truism ; this aspect of the matter
will be considered later on. For the present we shall take the
word " element " to mean what Dalton evidently intended it
to mean, and what we generally consider it to mean to-day,
namely a substance such as chlorine or lead which has constant
chemical properties, and which cannot be resolved into further
components by any known chemical process. The first half
taken together with the other four postulates is then
sufficient to define the word " element " and the second
becomes a pure hypothesis.
About ten years later Prout suggested that the atoms of the
elements were all made up of aggregations of atoms of hydrogen.
On this view the weights of all atoms must be expressed as
whole numbers, and if, as postulated by Dalton, the atoms of
any particular element are all identical in weight, the atomic
weights and combining ratios of all elements must be whole
numbers also. Chemists soon found that in the case of many
elements this was certainly not in agreement with experiment ;
the more results they obtained the more impossible it was to
express the atomic weights of all the elements as whole numbers.
They therefore had to decide which hypothesis, Dalton's or
Prout's, they would adopt. There was little doubt as to the
resul^ of the decision and in due course Prout's theory was
abandoned.
It is interesting to consider the reasons which led to a decision
which the subsequent history of science proves to have been
as wise in principle as it was wrong in fact. The alternative
INTRODUCTION 3
views were either an element was composed of atoms of iden-
tically the same weight, when in certain elements the weights of
the individual atoms must be fractional, or these particular
elements were composed of atoms of different weights mixed
together, so that though the individual weights of the atoms
would still be whole numbers their mean would be a fraction.
It is almost inconceivable that the second alternative never
occurred to philosophers during the time when the decision
hung in the balance indeed it was far more likely to be
considered then than years later when Dalton's view had been
generally accepted but the objections to it were immediate
and formidable. The idea that particles could behave in a
practically identical manner even though they had different
weights is not one that commends itself, a priori, to common
sense, and as a working hypothesis for chemists it is as hopeless
and indefinite as the simpler alternative is distinct and inspiring.
Also it could be urged that the objections to the fractional
weights of atoms were rather philosophic than practical.
They were concerned with the structure of individual atoms
and so might be, and wisely were, set aside till the time, distant
enough it would then have seemed, when these hypothetical
entities could be dealt with experimentally.
The idea that atoms of the same element are all identical in
weight could not be challenged by 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 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
4 ISOTOPES
separation of atoms differing from each other by as much as
10 per cent, in weight, is really an excessively difficult operation.
3. Crookes' meta- elements. The chemist who above
all others urged the possibility of the heterogeneity of atoms
was the late Sir William Crookes, to whom we are indebted
for so many remarkable scientific prophecies. His address to
the Chemical Section of the British Association at Birmingham
in 1886 1 is a most amazing effort of reason and imagination
combined and should be read by all those interested in the
history of scientific thought. In it he says : " I conceive,
therefore, that when we say the atomic weight of, for instance,
calcium is 40, we really express the fact that, while the majority
of calcium atoms have an actual atomic weight of 40, there
are not a few which are represented by 39 or 41, a less number
by 38 or 42, and so on. We are here reminded of Newton's
* old worn particles.'
" Is it not possible, or even feasible, that these heavier and
lighter atoms may have been in some cases subsequently sorted
out by a process resembling chemical fractionation ? This
sorting out may have taken place in part while atomic matter
was condensing from the primal state of intense ignition, but
also it may have been partly effected in geological ages by
successive solutions and reprecipitations of the various earths.
" This may seem an audacious speculation, but I do not
think it beyond the power of chemistry to test its feasibility."
Later 2 he developed this idea in connection with his pioneer
work on the rare earths. By a laborious process of fractional
precipitation he subdivided the earth yttria into a number of
components which had different phosphorescent spectra but
resembled each other very closely in their chemical properties.
Pointing out that at that time yttrium was considered to be
an element he says : " Here, then, is a so-called element whose
spectrum does not emanate equally from all its atoms ; but
some atoms furnish some, other atoms others, of the lines and
bands of the compound spectrum of the element. Hence the
atoms of this element differ probably in weight, and certainly
1 Nature, 34, 423, 1886.
2 Trans. Chem. Soc., 53, 487, 1888 ; 55, 257, 1889.
INTRODUCTION 5
in the internal motions they undergo." He called such com-
ponents " Meta-elements " and suggested that the idea might
apply to the elements generally, for example referring to the
seven series of bands in the absorption spectrum of iodine,
" some of these molecules may emit some of the series, others
others, and in the jumble of all these molecules, to which is
given the name * iodine vapour,' the whole seven series are
contributors."
In so far as they differed a little in atomic weight and a
mixture of them constituted a chemical element, these hypo-
thetical meta-elements may be said to have offered the first
feasible explanation of the fractional atomic weights. But
as more and more refined chemical methods were applied, the
rare earths one after another yielded to analysis and the
different spectra observed by Crookes were shown to be due
to the fact that he was dealing with a mixture of real elements,
each of which' had a characteristic spectrum and a definite
atomic weight. The theory of meta-elements was therefore
abandoned and the problem of fractional atomic weight
remained unsolved.
4. The discovery of Isotopes. As time went on the
numbers representing the atomic weights grew more and more
accurate and consistent. Significant figures one after another
were added by one worker, confirmed by others, and finally
approved by an International Committee. Small blame to
the student therefore if, when studying the imposing list of
numbers called the International Atomic Weights, he fell into
the very natural error of confusing " atomic weights " with
" weights of atoms," and considered that these figures did
actually represent the relative weights of the individual atoms
themselves.
Why so many of the atomic weights should be very nearly
integers when expressed on the scale 0=16 was still a very
difficult question to answer, for the probability against this
being due to pure chance was enormous, but it was not until
the discovery of radioactivity that the true reason for this
curious jumble of whole numbers and fractions was suggested,
and later confirmed generally by positive ray analysis. It is
6 ISOTOPES
worth noting that the first experimental proof that the
atoms of an element might be even approximately of the same
weight was given by positive ray parabolas. 1
The results given by the radioactive elements introduced a
wealth of new and revolutionary ideas. One of these was
that elements might exist which were chemically identical
but yet differed in radioactive properties and even in atomic
weight. By 1910 this idea had gained ground and was seriously
put forward and discussed by Soddy. At about the same
period the technique of positive ray analysis was rapidly being
improved, and in 1912 the first results were obtained from neon
which were later to support this new idea and carry it into the
region of the non-radioactive elements. From this time
onwards advances were made in the two fields side by side,
and so it happened that at the meeting of the British Associa-
tion in 1913 2 papers were read in different sections, one on
the Radio-elements and the Periodic Law, the other on the
Homogeneity of Neon, both of which tended to prove that
substances could exist with identical, or practically identical,
chemical and spectroscopic properties but different atomic
weights.
The need for a specific name for such substances soon became
imperative and Soddy suggested the word Isotopes (loot;
equal, ronoq, place) because they occupied the same place in
the periodic table of the elements.
1 V. p. 29.
2 Oddly enough this was the first meeting of the Association at
Birmingham since the one twenty-seven years before at which Crookes
made his prophetic remarks about atomic weights already quoted.
CHAPTER II
THE RADIOACTIVE ISOTOPES
5. Chemical identities among the radioactive ele-
ments. Apart from the purely speculative considerations
which have already been detailed, the theory of isotopes had
its birth in the gigantic forward wave of human knowledge
inaugurated by the discovery of radioactivity. It can admit-
tedly be argued that, even if no radioactive elements existed,
isotopes would inevitably have been discovered by the method
of positive rays. But progress must then have been exceed-
ingly slow, and the arrival at the real interpretation of the idea,
depending as it does on Sir Ernest Rutherford's theory of
the " nucleus " atom, almost impossible.
In 1906 Boltwood at Yale discovered a new element in the
radioactive group which he called Ionium, 1 and described
as having chemical properties similar to those of thorium.
So much was this the case that if, by accident, salts of these
two elements were mixed, he found it impossible to separate
them again by any of the chemical processes.
Boltwood, being occupied in the experimental proof that
ionium was the parent substance of radium, did not pursue
this line of investigation further at the time, but the work
was later taken in hand by Marckwald and Keetman of Berlin. 2
Thanks to the rapid advance in radioactive methods there
were now at command means of detecting change in concen-
tration of a delicacy unheard of in the previous work on the
rare earths, but yet, after years of patient and laborious work,
not the slightest sign of separation of ionium and thorium
could be observed. The chemical similarity between these
1 Boltwood, Amer. J. Sci., 22, 537, 1906 ; 24, 370, 1907.
* Keetman, Jahr. Radioactimtat, 6, 269, 1909.
7
8 ISOTOPES
two bodies was therefore of an order entirely different to that
exhibited by the rare earth elements, and came as near absolute
identity as the most critical mind could require.
This result was confirmed in the most rigorous manner by
Auer v. Welsbach, 1 who was able to apply to the problem
his valuable experience in work on the rare earths.
Furthermore, Mesothorium, discovered by Hahn in 1907,
was shown to be chemically inseparable from radium by
Marckwald 2 and Soddy 3 and similar chemical identities were
shown to be exceedingly probable in many other cases of
radioactive products. Certain regularities in the occurrence
of these were pointed out by Hahn and Meitner. 4
The situation was admirably summed up by Soddy in his
report on radioactivity for the year 1910 5 in the following
words :
" These regularities may prove to be the beginning of some
embracing generalisation, which will throw light, not only on
radioactive processes, but on elements in general and the
Periodic Law. Of course, the evidence of chemical identity
is not of equal weight for all the preceding cases, but the
complete identity of ionium, thorium and radiothorium, of
radium and mesothorium 1, of lead and radium D, may be
considered thoroughly established. . . . The recognition
that elements of different atomic weights may possess identical
properties seems destined to have its most important applica-
tion in the region of inactive elements, where the absence of
a second radioactive nature makes it impossible for chemical
identity to be individually detected. Chemical homogeneity
is no longer a guarantee that any supposed element is not a
mixture of several of different atomic weights, or that any
atomic weight is not merely a mean number. The constancy
of atomic weight, whatever the source of the material, is not
a complete proof of homogeneity, for, as in the radioelements,
genetic relationships might have resulted in an initial constancy
of proportion between the several individuals, which no sub-
1 A. von Welsbach, Wien. Ber. iia, 119, 1011, 1910.
2 Marckwald, Ber. d. Chem. Ges., 40, 3420, 1910.
3 Soddy, Trans. Chem. Soc., 99, 72, 1911.
4 Hahn and Meitner, Physical. Zeitsch., 11, 493, 1910.
6 Soddy, Chem. Soc. Ann. Rep., 285, 1910.
THE RADIOACTIVE ISOTOPES 9
sequent natural or artificial chemical process would be able to
disturb. If this is the case, the absence of simple numerical
relationships between the atomic weights becomes a matter of
course rather than one of surprise."
6. Spectroscopic identity of isotopes. The next great
advance was of an even more revolutionary character. This
consisted in the demonstration that the chemically indistin-
guishable products of the transformation of the radioactive
elements might also be spectroscopically identical. The idea
that elements of different atomic weight might yet have the
same spectrum originated in Sir Ernest Rutherford's laboratory
and appears to have been first entertained by A. S. Russell.
With Rossi * he undertook the comparison between the spec-
trum of pure thorium and that of a mixture of thorium and
ionium which radioactive evidence showed to contain a large
percentage of the latter element. No new lines attributable
to ionium were observed ; in fact the spectra obtained were
absolutely indistinguishable.
After giving in full the radioactive evidence as to the probable
percentage of ionium present, and showing that it was prac-
tically impossible for this to be too small for its spectrum to
appear, the writers go on as follows :
" There are, however, two other possible ways of explaining
our failure to obtain a distinct spectrum for ionium, besides the
one discussed above. It is possible that :
" (1) Ionium has no arc spectrum in the region investigated,
or
" (2) Ionium and thorium have identical spectra in the region
investigated.
" The first possibility is highly improbable, for all solids of
high atomic weights have arc spectra, and, further, all rare
earths have highly\ complicated spectra.
" The second possibility, though somewhat speculative
in nature, is suggested by some recent work on the chemical
properties of the radio -elements. There is no evidence
at present to disprove its truth. It is well known that there
are no less than four sets of longlived radio -elements, the
1 Russell and Rossi, Proc. Roy. Soc., 77A, 478, 1912.
10 ISOTOPES
members of each of which are chemically non-separable. These
elements do not all belong to the group of rare earths, many
non-radioactive members of which are known to be chemically
very similar. Mesothorium, for instance, which is chemically
non-separable from radium, belongs to the alkaline earth
group. Again the two non-separable a ray products which
are present in ordinary uranium, and which have been called
by Geiger and Nuttall uranium I and uranium II belong to
the chromium-molybdenum-tungsten group of elements. The
explanation of these striking chemical similarities is very
probably that the two very similar bodies are really different
members of the same group of elements, the difference in their
chemical properties being less pronounced than the difference
between other members of the same group, owing to the small
difference in their atomic weights. But the possibility that
they are identical in all physical and chemical properties, and
differ only in atomic weight and in radioactive properties, should
not be lost sight of. If this explanation should eventually provei
justified, the spectrum of ionium would be identical with that of
thorium."
It is not surprising that the idea was put forward with some
caution. Unlike that of chemical identity which had been
led up to by a gradual series of steps, it was entirely new and
contrary to all the preconceived ideas of the relations between
the spectrum of an element and the masses of its atoms. The
new departure was supported by Soddy x but received some
adverse criticism on the ground of insufficient evidence. The
later work bearing on this point will be described in Chapter X.
Already in 1911 the theory of the " Nucleus Atom " 2 had
been formulated. This gave the first hint as to the physical
meaning of chemical and spectroscopic identity, namely that
the nuclei of atoms might vary in their mass but yet, at the
same time, possess some property in common with each other,
namely nuclear charge, upon which the chemistry and spectra
depend.
In 1912 appeared the electrochemical work of Hevesy, 3
1 Soddy, Chem. News, Feb. 28, 1913. 2 V. p. 92.
3 G. Hevesy, Phil Mag., 23, 628, 1912 ; Physikal. Zeitsch. 15, 672,
715, 1912.
THE RADIOACTIVE ISOTOPES 11
which led to the discovery of the remarkable field of research
opened up by the use of radioactive bodies as indicators. 1 A
little later Paneth and Hevesy were able to show the complete
identity of the electrochemical properties of Radium D and
Lead. 2
In 1914 Rutherford and Andrade 3 examined the self -excited
X-ray spectrum of radium *jp They used a crystal of rock
salt for the analysis and got rid of the effect of the swift /?
rays by putting the source in a strong magnetic field. The wave
length of the L radiation proved to be exactly that expected
for lead from Moseley's experiment. 4 This was the first
proof that isotopes had identical X-ray spectra. The actual
values for ordinary lead were subsequently determined by
Siegbahn and found to be in excellent agreement with Ruther-
ford and Andrade's results.
7. The Chemical Law of Radioactive change. This
law, put in the briefest form, asserts : A radioactive element
when it loses an alpha particle goes back two places in the
Periodic Table ; when it loses a beta particle it goes forward
one place.
The law has been associated with the name of Soddy 5 who
was the first to suggest, in the form of a valency property,
that part of it relating to alpha rays. 6 But in its more com-
plete enunciation, which took place early in 1913, at least
four other investigators can claim a share.
Russell was the first to publish a law covering both kinds of
rays, 7 but owing to the fact that he failed to realise that the
sequence of elements in the periodic table is a continuous ex-
pression, his statement was not so simple and definite as it might
have been. Fajans, 8 using as foundation the electrochemical
1 V. p. 19.
2 Paneth and Hevesy, Sitzungber. K. Akad. Wiss. Wien, iiA, 123,
1037, 1913.
3 Rutherford and Andrade, Phil. Mag. 27, 854, 1914. * V. p. 93.
6 V. Stewart, Recent Advances in Physical and Inorganic Chemistry,
Longmans, 1919.
6 Soddy, The Chemistry of the Radio Elements, 29, First Edition,
Longmans, 1911.
7 Russell, Chem. News, Jan. 31, 1913.
8 Fajans, Physikal. Zeitsch. Feb. 15, 1913.
12 ISOTOPES
results of Hevesy, 1 and Soddy, working on the results of a
very full chemical investigation carried out at his request by
Fleck, 2 published the generalisation in its full and complete
form independently, and practically at the same time.
This law, which will be shown later to be [a, natural conse-
quence of the much wider generalisation discovered by Moseley, 3
has been of the greatest value in correlating the numerous
products of radioactive change, and predicting with accuracy
which of them will have identical properties. To the latter
the name Isotopes was applied by Soddy in the following
words : " The same algebraic sum of the positive and
negative charges in the nucleus when the arithmetical sum
is different gives what I call * isotopes ' or * isotopic elements '
because they occupy the same place in the periodic table.
They are chemically identical, and save only as regards the
relatively few physical properties which depend upon atomic
mass directly, physically identical also." Any element which
is the result of a series of changes involving the loss of twice
as many beta particles as alpha particles must clearly be the
isotope of the parent element, since it must inevitably, by the
above law, reach the same place in the periodic table at the end
of these operations.
8. Isobares. Just as we can have elements of the same
chemical properties but different atomic weight so we can also
have those with the same atomic weight but different chemical
properties. These Stewart 4 has called " Isobares." Any
product due to the loss of a beta ray (which has a negligible
mass) 5 must be an isobare of its parent substance, for, without
change of mass, it has moved in the periodic table and so
changed its chemical properties. It is interesting to note in
this connection that no isobare has actually been discovered
among the non-radioactive elements as yet, but they must
certainly exist. 6
1 Hevesy, PhysikaL Zeitsch. Jan. 15, 1913.
2 Fleck, Trans. Chem. Soc. 103, 381, 1052, 1913.
3 V. p. 93.
* Stewart, Phil. Mag. 36, 326, 1918.
5 V. p. 91.
6 V. p. 77.
THE RADIOACTIVE ISOTOPES 13
9. The Radioactive Transformations. The radioactive
elements are all formed from the two parent elements uranium
and thorium by a series of changes or transformations. These
changes can be classified according to their nature into two
types. In the first type of change called the a ray change
the atom loses a particle of mass 4 carrying two positive
charges (+ 2e) which has been identified with the nucleus
of the helium atom. 1 In the second or ($ ray change the
particle shot off has a negligible mass and carries a single
negative charge ( e). Hence in an a ray change the ele-
ment loses 4 units in atomic weight, while in a ft ray change
its weight is unaltered.
The rate of decay of an element is measured by the " half
value " period which may vary from 10 10 years to 10~ n of a
second. The velocity with which the rays are ejected als
varies and is apparently connected with the period of the
element by the very interesting relation of Geiger and Nuttall. 2
The intricate researches by which the complex series of trans-
formations have been explained belong to the subject of
Radioactivity and cannot be described here. From the point
of view of isotopes it will be enough to consider the final
results which are given in the two diagrams (Figs. 1 and 2).
In the first of these, which is due to Soddy, 3 the nuclear
charge or Atomic number, 4 upon which all the chemical
and spectroscopic properties of the elements depend, and
which expresses its position in the periodic table, is indicated
by a series of columns edged with thick lines sloping down-
wards to the right. The atomic weights are shown by fine
lines sloping in the opposite direction. The lines corre-
sponding to even atomic weights have been omitted to simplify
the scheme. All elements lying in the same column will
therefore be isotopes and all elements lying on the same line
sloping up to the right will be isobares. The a and /? ray
changes are shown by arrows and the period of decay of the
elements indicated by times expressed in suitable units.
1 V. Table p. 106.
8 Rutherford, Radioactive Substances and their Radiations, p. 607,
Cambridge, 1913.
3 Soddy, Trans. Chem, Soc., 115, 16, 1919. 4 V. p. 93.
14
ISOTOPES
FIG. 1. Diagram of the transformations of the radio-elements showing
atomic number, atomic weight and period of disintegration of each
product.
THE RADIOACTIVE ISOTOPES
15
The second diagram 1 is arranged in a simple manner to
show the general chains of transformation at a glance. In it
the a and f$ ray changes are plotted against atomic number
and the other information omitted. On this diagram all
elements lying on the same horizontal level will be isotopes.
To take an example, uranium I which has an atomic number
92 and an atomic weight 238 loses one a particle and becomes
uranium X, atomic number 90, atomic weight 234. This then
gives off two /? rays in succession, first becoming uranium X 2
92
URANIUM^, U.^
92
9!
90
U-Xjl >v Protoach'nium
U.X, Ionium U.Y \ Rad/'oacl: TnOMUM
Radioth.
91
90
89
Actinium
M
'th
89
88
Pad,
'UI77
Ai
JC Mesoi-h.j
I
X
88
87
87
86
Ra. Emanation
Ac. Em.
Th.Em.
86
85
85
84
fbfonium Ra.C'
Pa. A Ac.C 2
Ac. A Th.C
7h._A
84
\ A
/\
/ N.
83
Ra.E/ R<
/ Ac.C
/ ^
C
83
I/
\
/
\
/
\
82
PI). Ra.D Pb
Ra.B P& Pb.
AC.B Pt>. Pt>.
Th.B
82
81
Ra.C 2 Ac.D ~7h.D
81
FIG. 2. Diagram of the radioactive transformations in relation to atomic
numbers. In every case a step two downwards is accompanied by the
emission of an a particle and one downwards by a j8 particle.
and then Uranium II. Uranium II has an atomic number
92 so that it is an isotope of uranium I. It has an atomic
weight 234 so it is an isobare of uranium Xi and uranium X 2 .
Uranium II can disintegrate by shooting off an a particle in
two different ways ; about 8 per cent, of its atoms appear to
form uranium Y, which is probably the parent substance of
the actinium series. Disregarding this for the moment and
following the main chain, 92 per cent, of the atoms of uranium
1 Darwin* Nature, 106, 82, 1920.
16 ISOTOPES
II suffer an a ray change and are transformed into ionium,
atomic weight 230, atomic number 90. Ionium loses an a
particle and becomes radium, atomic weight 226. This by
the same process changes to radium emanation, then to radium
A, and then to radium B with atomic weight 214. We see
that uranium II has lost 5 a particles in succession, thereby
coming back 10 places, 92-82 in the periodic table, and its
atomic weight has been reduced 20 units in the process.
Radium B loses a /? particle, becoming "radium C which can
disintegrate in two different ways. An extremely small pro-
portion, 0-03 per cent., of its atoms undergo an a ray change to
radium C 2 which then loses a /? particle and may become inactive
lead of atomic weight 210. The vast majority of the atoms ofj
radium C lose a ft particle and form radium C'. This next
loses an a particle and becomes radium D, an active isotope
of lead of atomic weight 210. Radium D now loses two ft
particles in succession, becoming radium E and then radium F,
which is also called polonium. This finally undergoes its last
a ray change and becomes inactive uranium lead of atomic
weight 206.
The thorium and actinium chains can be followed on the dia-
grams in the same manner, but in the case of actinium the
parent elements are not satisfactorily settled so that the
atomic weights in this series are all doubtful.
10. The Atomic Weight of Lead. The theory of Isotopes
of which Professor Soddy had proved himself so prominent
an advocate and defender, received its most triumphant
vindication, as far as it concerned the products of radio-
activity, at the hands of the very chemists who had most
reason to doubt its general application, the specialists in the x
determination of atomic weights.
The charts of radioactive disintegration l show that the final j
product of every series is lead. If we take the main chain '
of the uranium -radium transformation this lead must have an
atomic weight 206, for it has lost 5 alpha particles each of
weight 4 since it was radium, and the atomic weight of radium
is 226. On the other hand if we take the main thorium chain
1 P. 14.
THE RADIO ACTIVE ISOTOPES 17
the lead end product must be 6 alpha particles lighter than
thorium (232-15) and so should have an atomic weight about
208.
Now ordinary lead, from non-radioactive sources has an
atomic weight 207 '20, so Soddy l suggested in 1913 that the
lead derived from minerals containing uranium but no thorium
might have a smaller atomic weight than ordinary lead, and
on the other hand the atomic weight of lead from minerals
containing thorium but no uranium might be greater.
The first experiments were made by Soddy and Hyman 2
with a very small quantity of lead from Ceylon Thorite. This
gave a perceptibly higher atomic weight than ordinary lead.
Later a large quantity of the same mineral was available.
The lead from this when carefully purified gave a density
0-26 per cent, higher than that of common lead. On the
assumption that the atomic volumes of isotopes are equal
this figure corresponds to an atomic weight of 207 '74. A chem-
ical atomic weight determination gave 207-694. A sample of
the same lead was sent to Vienna where Professor Honigschmid,
a well known expert in such matters, obtained from it a value
207-77 as a mean of eight determinations. These figures not
only showed that thorium lead had a higher atomic weight
than ordinary lead but also that their atomic volumes were
identical, as expected from theory. 3
At the same time as this work was in progress, the leading
American authority on atomic weights, T. W. Richards of
Harvard, started a series of investigations on lead derived from
various radioactive minerals. 4 The samples of lead from
uranium minerals all gave results lower than ordinary lead,
as was expected, and one particularly pure specimen of uranio-
lead from Norwegian cleveite gave 206-08, 5 a very striking
agreement with theory. The following table of properties
is taken from his Presidential address to the American
Association at Baltimore, December, 1918.
*
1 Soddy, Ann. Rep. Chem. Soc., 269, 1913.
2 Soddy and Hyman, Trans. Chem. Soc., 105, 1402, 1914.
3 Soddy, Roy. Ins., May 18, 1917.
4 Richards and Lembert, J. Amer. Chem. Soc., 36, 1329, 1914.
5 Richards and Wadsworth, J. Amer. Chem. Soc., 38, 2613, 1916.
C
18
ISOTOPES
Common
Lead.
A
Mixture
Australian.
B
Uranio-
Lead.
C
Perce
Diffei
A-B
ntage
"ence.
A-C
Atomic weight ....
20719
206-34
206-08
0-42
0-54
Density
1T337
H'280
1T273
0'42
0'56
Atomic volume ....
18-277
18-278
18-281
O'Ol
0-02
Melting point (absolute) .
600-53
600-59
o-oi
Solubility (of nitrate)
37-281
37-130
0-41
Refractive Index (nitrate) .
1-7815
1-7814
.
o-oi
Thermoelectric effect.
o-oo
Spectrum wave-length .
o-oo
o-oo
In further confirmation Maurice Curie in Paris * reported
206*36 for a lead from carnotite, and a still lower figure, 206-046,
was obtained by Honigschmid in Vienna for a lead from the
very pure crystallised pitchblende from Morogoro. This is
the lowest atomic weight found so far. The highest, 207-9,
was also determined by Honigschmid for lead from Norwegian
thorite. 2
11. Atomic weights of Thorium and Ionium. Although
the above results obtained with lead are far the most con-
clusive and important it is not the only element which affords
direct experimental evidence of the different atomic weights
of isotopes. The atomic weight of ionium, calculated by
adding the weight of one alpha particle to the atomic weight
of its product, radium, is 230, whereas that of thorium, its
isotope, is slightly above 232. Joachimsthal pitchblende con-
tains hardly any thorium so that an ionium -thorium prepara-
tion separated by Auer von Welsbach from 30 tons of this
mineral might be regarded as containing a maximum concen-
tration of ionium. On the other hand the period of thorium
is about 10 5 times longer than that of ionium so that it was
doubtful if even in this preparation there would be enough
ionium to show a difference in atomic weight. Honigschmid
and Mile. Horovitz have made a special examination of this
point, first redetermining as accurately as possible the atomic
weight of thorium and then that of the thorium -ionium prepar-
!M. Curie, Compt. Rend., 158, 1676, 1914.
2 Honigschmid, Zeit. fflektrochem. , 24, 163, 1918 ; 25, 91, 1919,
THE RADIOACTIVE ISOTOPES 19
ation from pitchblende. They found 232-12 for the atomic
weight of thorium, and by the same careful method 231-51
for that of the thorium -ionium.
12. Use of radioactive isotopes as indicators. Con-
sider an inactive element A which has a radioactive isotope B.
If these are mixed together in any proportions no chemical
or physical process known is capable of altering the ratio of
the proportions of this mixture to any measurable extent.
Now the radioactive methods of detecting and measuring B
are many millions of millions of times more delicate than
the chemical methods of detecting and measuring A, so that
by mixing with A a small quantity of B we can trace its presence
far beyond the limits of chemical analysis. We have, as it
were, marked the atoms of A with an indelible label so that
the minutest trace of the element can be measured with ease
and certainty.
By this powerful and novel device, which has been developed
by G. Hevesy l 10~ 9 gr. of lead can be determined quantita-
tively and solution concentrations can be dealt with down to
10" 14 of normal. By adding radium D to the lead salt and
estimating it electroscopically the solubility of lead sulphide
and chromate, and the amount of lead chloride carried down
in a silver chloride precipitate, may readily be determined.
Recently, by the same principle, it has been shown that a
free exchange of the metallic atom among the competing acid
radicles occurs for ionised, but not for non-ionised, compounds.
The general method was to mix solutions of two different
compounds of lead in equimolecular proportions, the one
compound only being " activated " by presence of thorium-B
(which is isotopic with lead), and to determine the activity of
the lead in the less soluble compound crystallising out. When
active lead nitrate and inactive lead chloride are dissolved
in molecular proportion in boiling pyridine, the lead in the
lead chloride crystallising out is half as active as the lead in
the original lead nitrate, but when such an active lead salt
is so mixed with an organic compound of lead, such as lead
tetraphenyl or diphenyl nitrate, in suitable solvents, no inter -
1 Hevesy, Brit. Assoc., 1913 ; Chem. News, Oct. 13, 166, 1913.
20 ISOTOPES
change of lead occurs, and the active lead salt retains its
original activity. This constitutes something like a direct
proof of the ionic dissociation theory and of the current views
as to the difference between the nature of chemical union in
electrolytes and non-electrolytes. When the acetates of
quadrivalent activated lead and of bivalent inactive lead are
mixed in glacial acetic acid, the activity of the first compound,
after crystallising out from the mixture, is reduced to one
half. This indicates, since the two lead ions differ only by
two electrons, a free interchange of electrons between them
and a dynamic equilibrium between ions and electrons
and between free electrons and the electrodes in electro-
lysis. 1
Isotopes can also be used to determine the velocity of
diffusion of molecules among themselves. 2 The rate of diffu-
sion is dependent on the molecular diameter, and not on the
mass, so that a radioactive element diffusing among the
inactive molecules of its isotope affords a means of investi-
gating this otherwise insoluble problem. The experiment has
been tried with molten lead. At the, bottom of a narrow
vertical tube was placed a layer of lead rendered active by
the presence of thorium-B, and above it a layer three times
the height of common lead. The whole was kept at 340
for several days. After cooling, the cylinder was cut into
four equal lengths, each melted and hammered into foil, and
the concentration of thorium-B in each determined by alpha
ray measurements. Values for the diffusion coefficient between
1*77 and 2-54 per sq. cm. per day, with a mean of 2-22 in
seventeen experiments, were obtained. On certain theories
of physical chemistry this corresponds with a diameter of the
lead molecule between 0-78 and 1-16 X 10~ 8 cm., according to
the formulae used to connect the two quantities. The value
found by similar theories when reduced to a temperature of
18 and for a fluid of the viscosity of water, becomes 2-13.
Since the value for lead ions diffusing in aqueous solutions is
0-68, this indicates that the molecular diameter in the case
1 G. Hevesy and L. Zechmeister, Ber., 53B, 410, 1920; Zeitsch.
Elektrochem. 26, 151, 1920.
2 J. Groh and Hevesy, Ann. Physik., iv., 63, 85, 1920.
THE RADIOACTIVE ISOTOPES 21
of metallic lead is only a third of that in the case of the ion,
and shows that the latter is probably hydrated. 1
13. Classification of the radioactive isotopes. It is
clear that the relations between isotopes formed by radioactive
disintegrations need not necessarily have the same simple form
as those subsisting between isotopes of the inactive elements.
Neuburger, 2 using the nucleus model of the radioelements
proposed by Lise Meitner, 3 suggests that the radioactive iso-
topes may be divided into three or even four classes.
(1) Isotopes of the first class are those which possess only
the same nuclear charge and the same arrangement of outer
electrons such as radium and mesothorium I.
(2) Isotopes of the second class have, in addition, the same
nuclear mass, that is to say the same atomic weight, and the
same total number of nuclear " building stones." Examples
of this class are ionium and uranium Y.
(3) Isotopes of the third class still possess the same number
of each nuclear building stone, but they have a different
arrangement of these in the atomic nucleus, and thus possess
different chances of disintegrating, such as Radium D and
Actinium B.
(4) Isotopes of the fourth class would be those possessing
the same arrangement of nuclear building stones in the atomic
nucleus, and thus the same probability of disintegrating.
Such isotopes actually exist, but we have no available means
of distinguishing between them. Hence we cannot at present
designate them definitely as isotopes. Examples of these are
radium C 2 and actinium D.
1 Soddy, Ann. Rep. Chem. Soc., 227, 1920.
a Neuburger, Nature, 108, 180, 1921.
3 Meitner, Die Naturwissenschaften, 9, 423, 1921.
CHAPTER III
POSITIVE RAYS
14. Nature of Positive Rays. Positive rays were dis-
covered by Goldstein in 1886 in electrical discharge at low
pressure. In some experiments with a perforated cathode he
noticed streamers of light behind the perforations. This
luminosity, he assumed, was due to rays of some sort which
travelled in the opposite direction to the cathode rays and
so passed through the apertures in the cathode, these he called
" canalstrahlen." 1 Subsequently Wien showed that they
could be deflected by a magnetic field. 2 They have been very
fully investigated in this country by Sir J. J. Thomson, 3 who
called them Positive Rays on account of the fact that they
normally carry a charge of positive electricity.
The conditions for the development of the rays are, briefly,
ionisation at low pressure in a strong electric field. lonisation,
which may be due to collisions or radiation, means in its
simplest case the detachment of one electron from a neutral
atom. The two resulting fragments carry charges of electricity
of equal quantity but of opposite sign. The negative^^harged
one is the electron, the atomic unit of negative^ MK?ity
itself, 4 and is the same whatever the atom ionisSBWft is
extremely light and therefore in the strong electric field rapidly
'attains a high velocity and becomes a cathode ray. The remain-
ing fragment is clearly dependent on the nature of the atom
ionised. It is immensely more massive than the electron, for
1 Goldstein, Berl. Ber., 39, 691, 1886.
2 Wien, Verh. d. Phys. GeselL, 17, 1898.
8 J. J. Thomson, Rays of Positive Electricity and their Application
to Chemical Analyses, Longmans, Green, 1913.
* B. A. Millikan, The Electron, University Chicago Press, 1918.
22
POSITIVE RAYS 23
the mass of the lightest atom, that of hydrogen, is about 1845
times that of the electron, and so will attain a much lower
velocity under the action of the electric field. However, if
the field is strong and the pressure so low that it does not
collide with other atoms too frequently it will ultimately attain
a high speed in a direction opposite to that of the detached
electron, and become a " positive ray." The simplest form
of positive ray is therefore an atom of matter carrying a
positive charge and endowed, as a result of falling through a
high potential, with sufficient energy to make its presence
detectable. Positive rays can be formed from molecules as
well as atoms, so that it will at once be seen that any measure-
ment of their mass will give us direct information as to the
masses of atoms of elements and molecules of compounds, and
that this information will refer t<i> the atones or molecules
individually, not, as in chemistry, to the mean of an immense
aggregate. It is on this account that the accurate analysis of
positive rays is of such importance.
In order to investigate and analyse them it is necessary
to obtain intense beams of the rays. This can be done in
several ways. The one most generally available is by the
use of the discharge in gases at low pressure.
15. Mechanism of the electric discharge in gases at
low pressure. It is a somewhat striking anomaly that while
the working of the very recently invented " Coolidge " X ray
bulb can be simply described and explained, this is far from
being the case with the much older ordinary " gas " tube.
Notwithstanding the immense amount of research work done
on the discharge at low pressure its most obvious phenomena
are well nigh entirely lacking explanation. Modern measure-
ments and other data have merely destroyed the older theories,
without, as yet, giving others to replace them.
For the purposes of describing positive rays it is not necessary
to consider such puzzles as the " striated discharge " or other
phenomena connected with the anode end of the tube, but
some ideas as to what is going on near the cathode will be a
considerable help in our interpretation of the results of positive
ray analysis, and vice versa.
24 ISOTOPES
16. The Crookes Dark Space. The comparatively dimly
lit space in front of the cathode, terminating at the bright
" negative glow " was first observed by Crookes. Its length
is roughly inversely proportional to the pressure of the gas in
the tube. Its boundary the edge of the negative glow is
remarkably sharp in most gases, quite amazingly so in pure
oxygen. If large plane cathodes are used so that the effect
of the glass walls up to now a complete mystery is minimised
very accurate and consistent measurements can be obtained.
Such measurements have been made under a great variety of
conditions by the writer. 1 The distribution of electric force
in the dark space has also been determined for large plane
electrodes 2 but no theory yet put forward can account for
the numerical relations obtained in these investigations, nor
for others obtained later with perforated electrodes. 3
One can, however, be fairly certain that ionisation is going
on at all points throughout the dark space, and that it reaches
a very high intensity in the negative glow. This ionisation
is probably caused for the most part by electrons liberated
from the surface of the cathode (Cathode Rays). These,
when they reach a speed sufficient to ionise by collision, liber-
ate more free electrons which, in their turn, become ionising
agents, so that the intensity of ionisation from this cause will
tend to increase as we move away from the cathode. The
liberation of the original electrons from the surface of the
cathode is generally regarded as due to the impact of positive
ions (Positive Rays) generated in the negative glow and the
dark space, but this idea, for which there is a fair amount of
definite evidence, is now called in question by some recent
experiments of Ratner. 4
In addition to cathode ray ionisation the positive rays
travelling towards the cathode themselves are capable of
ionising the gas, and radiation may also play an important
part in the same process. The surface of the cathode will
1 Aston, Proc. Roy. Soc. 79A, 80, 1907 ; Aston and Watson, ibid.
86A, 168, 1912 ; Aston, ibid. 87A, 428, 437, 1912.
2 Aston, Proc. Roy. Soc. 84A, 526, 1911.
8 Aston, Proc. Roy. Soc. 96A, 200, 1919.
* Ratner, Phil. Mag. 40, 795, 1920.
POSITIVE RAYS 25
therefore be under a continuous hail of positively charged
particles. Their masses may be expected to vary from that
of the lightest atom to that of the heaviest molecule capable
of existence in the discharge tube, and their energies from an
indefinitely small value to a maximum expressed by the
product of the charge they carry x the total potential applied
to the electrodes. The latter is practically the same as the
fall of potential across the dark space. If the cathode be
pierced the rays pass through the aperture and form a stream
heterogeneous both in mass and velocity which can be subjected
to examination and analysis.
17. Methods of detecting positive rays. The glow
caused by the passage of the rays through rarefied gas led to
their original discovery but is not made use of in accurate work.
For visual effects the rays are best detected by a screen made
of powdered willemite, which glows a faint green when bom-
barded by them. When permanent effects are required this
screen is replaced by a photographic plate. The sensitivity
of the plate to positive rays bears no particular relation to its
sensitivity to light, and so far the best results have been
obtained from comparatively slow " process " plates of the
type known as " Half-Tone." The real relative intensities of
rays of different mass cannot be compared by screens or
photographic plates, except in the possible case of isotopes of
the same element ; they can only be determined reliably by
collecting the rays in a Faraday cylinder and measuring their
total electric charge.
18. Sir J. J. Thomson's " Parabola " method of
analysis. The method by which Sir J. J. Thomson made
such a complete investigation into the properties of positive
rays, and which still remains pre-eminent in respect to the
variety of information it supplies, consists essentially in allow-
ing the rays to pass through a very narrow tube and then
analysing the fine beam so produced by electric and magnetic
fields.
The construction of one of the types of apparatus used is
indicated in Fig. 3. The discharge by which the rays are
made takes place in a large flask A similar to an ordinary X-ray
26 ISOTOPES
bulb of about 1J litres capacity. The cathode B is placed
in the neck of the bulb. Its face is made of aluminium, and
so shaped that it presents to" the bulb a hemispherical front
provided in the centre with a funnel-shaped depression. This
hole through which the rays pass is continued as an extremely
fine-bore tube, usually of brass, about 7 cms. long, mounted
in a thick iron tube forming the continuation of the cathode as
indicated. The finer the bore of this tube the more accurate
are the results obtained, and tubes have been made with success
as narrow as one-tenth of a millimetre, but as the intensity
of the beam of rays falls off with the inverse fourth power of
the diameter a practical limit is soon reached. The cathode
FIG. 3. Positive Ray Apparatus.
is kept cool during the discharge by means of the water-jacket
C.
The anode is an aluminium rod D, which is generally placed
for convenience in a side tube. In order to ensure a supply
of the gas under examination a steady stream is allowed to
leak in through an exceedingly fine glass capillary tube E, and
after circulating through the apparatus is pumped off at F by
a Gaede rotating mercury pump. By varying the speed of the
pump and the pressure in the gas-holder communicating with E,
the pressure in the discharge tube may be varied at will and
maintained at any desired value for considerable lengths of time.
The pressure is usually adjusted so that the discharge potential
is 30,000 to 50,000 volts. During the discharge all the conditions
necessary for the production of positive rays are present in A.
Under the influence of the enormous potentials they attain
POSITIVE RAYS 27
high speeds as they fly towards the cathode, and those falling
axially pass right through the fine tube, emerging as a narrow
beam.
This beam is subjected to analysis by causing it to pass
between the pieces of soft iron P, P' which are placed between
the poles M, M' of a powerful electromagnet. P and P' con-
stitute the pole pieces of the magnet, but are electrically
insulated from it by thin sheets of mica N, N', and so can be
raised to any desired potential difference by means of the
leads shown in the diagram. The rays then enter the highly
exhausted " camera " G, and finally impinge upon the fluores-
cent screen or photographic plate H. In order that the stray
magnetic field may not interfere with the main discharge in
A, shields of soft iron, I, I' are interposed between the magnet
and the bulb.
If there is no field between the plates P, P' the beam of rays
will strike the screen at a point in line with the fine tube called
the undeflected spot. If an electric field of strength X is
now applied between the plates a particle of mass m, charge e,
moving with velocity v, will be deflected in the plane of the
paper and will no longer strike the screen at the undeflected
spot, but at a distance x from it. Simple dynamics show
that if the angle of deflection is small x = k(Ke/mv 2 ). In the
same way, if the elecfric field is removed and a magnetic field
of strength H applied between P and P' the particle will be
deflected at right angles to the plane of the paper and strike
the screen at a distance y from the undeflected spot where
y = k'(TLe/mv), k and k' being constants depending solely on
the dimensions and form of the apparatus used. If now, with
the undeflected spot as origin, we take axes of co-ordinates
OX, OY along the lines of electric and magnetic deflection,
when both fields are applied simultaneously the particle will
strike the screen at the point (x, y) where y/x is a measure of
its velocity and y z /x is a measure of m/e its ratio of mass to
charge.
Now e can only exist as the electronic charge 4- 7 7 x 10~ 10
C.G.S. or a simple multiple of it. Thus if we have a beam of
positive rays of constant mass, but moving with velocities
varying over a considerable range, y 2 /x will be constant and
28
ISOTOPES
the locus of their impact with the screen will be a parabola
pp' (Fig. 4). When other rays having a larger mass m' but
the same charge are introduced into the beam, they will appear
as another parabola qq' having a smaller magnetic displacement.
If any straight line p, q, n be drawn parallel to the magnetic
axis OY cutting the two parabolas and the electric axis OX
in p, q, n it will be seen at once that m'/m =pn 2 /qn 2 . That
is to say, the masses of two or more particles can be compared
directly by merely measuring lengths the ratio of which is
entirely independent of the form of the apparatus and the
experimental conditions.
This is really the fundamental principle upon which the
method is based. A photographic record is obtained on which
v we can identify at least one
parabola as being associated
with atoms or molecules of
known mass ; all the other
parabolas can then be measured
and compared with this one and
their masses deduced. With
electric and magnetic fields
roughly known there is little
difficulty in- such an identifica-
tion, and to make quite sure
the absolute value of m/e for
the hydrogen atom was deter-
mined and found to agree with the values obtained by
other methods. In actual practice, since OX is an imagin-
ary line and has no existence on the photograph, in order
that the measurements may be made with greater conveni-
ence and accuracy the magnetic field is reversed during
the second half of the exposure, when in the case we are
considering two new parabolas will appear at rr', ss', due to
m and m' respectively ; the masses can now be compared by
the equation m'/m pr 2 /qs 2 : p, q, r, s being any straight
line cutting the curves approximately parallel to the magnetic
axis. The measurement of these lengths is independent of zero
determination, and if the curves are sharp can be carried out
with considerable accuracy.
FIG. 4. Positive Ray Parabolas.
PLATE I.
1. 2.
Photographs of Typical Positive Ray
Parabolas.
The Multiply Charged Parabolas
of Mercury.
The Parabolas of
Lithium.
4. The Parabolas of Neon.
POSITIVE RAYS 29
Some of the photographic results obtained by this method
of analysis are shown in Plate I. The fact that the streaks are
definite sharp parabolas, and not mere blurs, was the first
experimental proof that the atoms of the same element had
very approximately the same mass.
It has' been shown that the electrical displacement is in
inverse proportion to the energy of the particle. Since this
energy is simply dependent on and proportional to the electrical
potential through which the charged particle fell before it
reached the cathode and not upon its mass, the distribution
of intensity along the parabolas will be somewhat similar.
There will also be a definite maximum energy corresponding
to the whole drop of potential across the discharge tube, with
a corresponding minimum displacement on the plate ; so that
all normal parabolas will end fairly sharply at points p, q,
etc., equidistant from the magnetic axis OY. As the ionisation
is a maximum in the negative glow the parabolas are brightest
at or near these points. The extension of the curves in the
other direction indicates the formation of ions at points in the
discharge nearer the cathode which will so have fallen through
a smaller potential.
19. Secondary Rays. As the pressure in the camera,
though as low as possible, is never entirely negligible, the
particles may make collisions, and so gain and lose electrons,
while passing through the deflecting fields. This results in
what Sir J. J. Thomson calls " secondary rays," l which may
be of a great many types. Some appear on the plate as
general fog, others as straight beams seeming to radiate from
the undeflected spot, these will easily be recognised on the
photographs produced in Plate I. Secondary rays can pro-
duce parabolas which are very much like the genuine ones
caused by particles which have retained their charge through
both fields, and which may easily be mistaken for them unless
special precautions are taken.
20. Negatively Charged Rays. As there is intense
ionisation in the fine tube the charged particles may easily
collide with and capture electrons in passing through it. A
1 J. J. Thomson, Rays of Positive Electricity, p. 32.
30 ISOTOPES
singly charged particle capturing a single electron will, of
course, proceed as a neutral ray, and being unaffected by the
fields will strike the screen at the central spot. If, however,
it makes a second collision and capture it will become a nega-
tively charged ray. Rays of this kind will suffer deflection
in both fields in the opposite direction to the normal ones, and
will therefore give rise to parabolas of a similar nature but situa-
ated in the opposite quadrants, as indicated by the dotted
lines in the figure. Such negative parabolas are always less
intense than the corresponding normal ones, and are usually
associated with the atoms of electronegative elements such as
carbon, oxygen, chlorine, etc.
The negative parabolas of H, C and can be seen in the
photographs. Plate I (1) and (2).
21. Rays with Multiple Charges. If during ionisation
more than one electron is split off, the resulting positive ray
will have 'a double or multiple charge. Taking the case of a
doubly charged particle it may give rise to two distinct effects.
In the first place, if it retains its double charge while passing
through the analysing fields its behaviour will be quite indis-
tinguishable from that of a normal ray of half its mass. Thus
the effective mass of the doubly charged oxygen atom, written 1
Oj" + , will be 8. Parabolas due to C ++ and + + can be seen
in Plate I (2). In the second place, the particle may retain
its double charge through the whole potential fall of the dis-
charge but capture an electron in the fine tube. It will then
constitute a ray of normal ratio of mass to charge but with
double the normal energy, so that the normal end of the
parabolas will be extended towards the axis OY to a point half-
way between that axis and the line pq. Such extensions will
be seen on the bright parabolas due to carbon and oxygen in the
photographs reproduced in Plate I.
Most elements are capable of losing two electrons, some,
such as krypton, three or more, while mercury can lose ntTliis
rtdb- eight at a time. The results of the multiple charge on
atoms of mercury is beautifully illustrated in Plate I (3).
The parabola a corresponding to normal single charge will
1 In the normal singly-charged ray the plus sign is omitted for
convenience.
POSITIVE RAYS
31
be seen extended almost to the origin itself, while above a series
of parabolas of diminishing intensity /?, y, etc., indicate the
atoms which have retained two, three or more charges.
22. Dempster's method of positive ray analysis.
It is clear from the considerations on page 27 that if the posi-
tive particles all fell through the same potential and so possessed
the same energy, a magnetic field alone would suffice to perform
their analysis with regard to mass. A method of analysis
based on this idea has been devised by Dempster at the Ryer-
son Physical Laboratory, Chicago. 1
The method is essentially identical with that used by Classen
in his determination of e/m for electrons 2 The charged par-
FIG. 5. Dempster's Apparatus.
tides from some source fall through a definite potential differ-
ence. A narrow bundle is separated out by a slit and is bent
into a semicircle by a strong magnetic field ; the rays then
pass through a second slit and fall on a plate connected to an
electrometer. The potential difference P, magnetic field H,
and radius of curvature r determine the ratio of the charge to
6 2P
the mass of the particle by the formula - =ff^- 2
t7i/ -tl T
1 Dempster, Phys. Rev. 11, 316, 1918.
2 Classen, Jahrb. d. Hamburg Wiss. Anst., Beiheffc, 1907.
32 ISOTOPES
The apparatus consisted of a glass tube G, Fig. 5, where the
positive particles fell through a definite potential difference,
and the analysing chamber A, in which a strong magnetic field
was produced between two semicircular iron plates 2-8 cm.
thick and 13 cm. in diameter. The iron plates were soldered
into half of a heavy brass tube B, so as to leave a passage or
slot 4 mm. wide between the plates. A brass plate C closed
this slot except for three openings into which short brass tubes
were soldered. The glass tube G fitted into the first opening and
a tube for exhausting into the second. The electrometer
connection passed to a receiving plate through an ebonite plug
E which formed a ground conical joint with the third brass
tube. The two openings for the rays had adjustable slits S i, S 2 ,
and a screen D was introduced into the analysing chamber to
prevent reflected rays getting into the second slit. The whole
was placed between the poles of a powerful electromagnet.
The accelerating potential P was applied by means of a large
battery and was from 500 to 1750 volts or thereabouts. The
experimental procedure consisted in maintaining a constant
magnetic field and plotting the ionic current, measured by the
electrometer, against the potential. The peaks on the curve
corresponded to definite values of m/e, measured by the poten-
tial, and their heights to the relative quantities of the particles
present in the beam.
The method is limited in its application by the fact that the
ions must be generated with a velocity negligible compared
with that produced by the accelerating potential. The first
results were obtained from ions produced by heating salts on
platinum strips, or by bombarding them with electrons. It
was shown that the ions given off from heated aluminium
phosphate consisted for the most part of sodium and potassium
atoms, and that these had masses 23 and 39 respectively. The
resolution possible with the first apparatus was claimed to
be about 1 in 100. Dempster's recent successful application
of this method to the analysis of magnesium and lithium will
be described in a later chapter. 1
\V. p. 80.
CHAPTER IV
NEON
23. Positive Ray Analysis of Neon. It is a curious and
interesting point that while the first suggestion of the possi-
bility of the occurrence of isotopes was obtained from the
rarest of all substances on the earth's surface the radioactive
elements and their products ; so the first result indicating
the possibility of isotopes among the stable elements was
yielded by neon, a gas of which, in a purified state, there was
probably less than one gramme in existence.
Neon is one of the inactive constituents of the atmosphere,
in which it occurs to the extent of 0-00123 per cent, by volume.
It was first isolated by Ramsay and Travers in 1898, and was
accepted as an elementary monatomic element of the helium
group. Its density was measured with extreme care by
Watson l and found to correspond with an atomic weight
20-200 (0 = 16), so that it is the lightest element whose atomic
weight differs from a whole number in an unmistakeable
manner.
In the summer of 1912 there had been constructed in the
Cavendish Laboratory a Positive Ray apparatus which was a
considerable improvement on those made previously. 2 The
parabolas corresponding to masses differing by 10 per cent,
could be clearly resolved and distinguished by its means.
Many gases were submitted to analysis ; but no results were
obtained which could not be accounted for until in November
of that year a sample of the lighter constituents of air was
introduced. In describing the results obtained one cannot do
1 Watson, J.C.S. Trans. 1, 810, 1910.
2 J. J. Thomson, Rays of Positive Electricity, p. 20 g
33 D
34 ISOTOPES
better than quote Sir J. J. Thomson's own words from his
address to the Royal Institution on Friday, January 17,
1913.
" I now turn to the photograph of the lighter constituents ;
here we find the lines of helium, of neon (very strong), of
argon, and in addition there is a line corresponding to an
atomic weight 22, which cannot be identified with the line
due to any known gas. I thought at first that this line, since
its atomic weight is* one-half that of C0 2 , must be due to a
carbonic acid molecule with a double charge of electricity,
and on some of th plates a faint line at 44 could be detected.
On passing the gas slowly through tubes immersed in liquid
air the line at 44 completely disappeared, while the brightness
of the one at 22 was not affected.
" The origin of this line presents many points of interest ;
there are no known gaseous compounds of any of the recognised
elements which have this molecular weight. Again, if we
accept Mendeleef's Periodic Law, there is no room for a new
element with this atomic weight. The fact that this line is
bright in the sample when the neon line is extraordinarily
bright, and invisible in the other when the neon is compara-
tively feeble, suggests that it may possibly be a compound of
neon and hydrogen, NeH 2 , though no direct evidence of the
combination of these inert gases has hitherto been found. I
have two photographs of the discharge through helium in
which there is a strong line, 6, which could be explained by
the compound HeH 2 , but, as I have never again been able to
get these lines, I do not wish to lay much stress on this point.
There is, however, the possibility that we may be interpreting
Mendeleef's law too rigidly, and that in the neighbourhood of
the atomic weight of neon there may be a group of two or
more elements with similar properties, just as in another part
of the table we have the group iron, nickel, and cobalt. From
the relative intensities of the 22 line and the neon line we may
conclude that the quantity of the gas giving the 22 line is only
a small fraction of the quantity of neon."
Other samples of gas containing neon all gave similar results.
By good fortune some of the purest neon in existence was also
available ; this had been employed by the writer and Watson
NEON
35
in some investigations on the Crookes Dark Space l and was
actually a part of that by which the atomic weight had been
determined. This sample also yielded the two separate
parabolas with the same relative intensity as the others. One
of the photographs taken with neon is reproduced in Plate 1 (4) .
The last result proved that the most careful purification had
not appreciably altered the intensity ratio between the lines
and might at first sight appear a strong argument for the
NeH 2 explanation, but further study of the parabolas only
added more weight to the chemical objections against the
existence of such a compound. The only other alternative
was a novel and revolutionary one, namely that neon could
exist in two forms and that the relation between these was
precisely that which had been described by Soddy a short
time before as existing between the chemically inseparable
radio elements.
These considerations led the author to undertake a search-
ing investigation on the constitution of the gas by two distinct
lines of attack, first attempts at separation, secondly examina-
tion by positive rays. 2
24. Apparatus for the determination of density. As
neon is chemically inactive the most satisfactory proof of a
partial separation of its constituents is a change in density.
FIG. 6. Microbalance.
It was therefore necessary to devise some means of deter-
mining density accurately, quickly and with the minimum
1 Aston and Watson, Proc. Roy. Soc., 86A, 1912.
2 The neon necessary for this research was given by M. Georges
Claude of Paris.
36 ISOTOPES
quantity of gas. All these desiderata were obtained by the
construction of a simple quartz micro-balance shown in Fig. 6. 1
The principle upon which this works is that if a sealed
vacuous quartz bulb is equipoised against a solid piece of
quartz on a balance the system can only be exactly balanced,
at any predetermined position, when it is immersed in a fluid
of an absolutely definite density ; if the density is too high the
bulb will be buoyed up, if too low it will sink. We can there-
fore compare the densities of a known and an unknown gas
by introducing them successively into the balance case and
determining the pressures at which the system is exactly ^
balanced.
The moving part of the balance is made entirely of fused
quartz (shown black). It turns upon a single knife-edge cut
on a piece of quartz rod about 0-5 mm. thick. To this rod, a
few millimetres above the knife-edge, are fused two others
about the same thickness forming the arms of the beam. To
the end of one arm is fused a sealed vacuous quartz bulb
holding about 0-3 c.c. and to the other a counterpoise made of
a piece of rod about 2 mm. thick. The beam is supported by
its knife-edge on a horizontal quartz plate and housed in a
thick glass vacuum-tight case fitting as closely as possible so
that its volume is a minimum. The case is connected through
the capillary tube shown to a gas pipette and a pump for the
introduction and removal of gas and also to a simple form of
mercury manometer. The beam was adjusted during its
construction so that it balanced in air at about 85 mm.
pressure. In the process of adjustment the end of the counter-
poise was drawn out into a fine tail ending in a small knob ;
this was used as the pointer of the beam. The sensitivity of
the balance is about 10~ 6 mgrm., which enables the manometer
to be set to one-twentieth of a millimetre with ease.
25. Method of using the density balance. About the
right volume of gas, generally known from previous experience,
is admitted to the balance case and the mercury level in the/
manometer slowly raised (increasing the pressure in the balance
case) until the bulb rises and the knob at the extremity of the
1 Aston, Proc. Eoy. Soc., 89A, 440, 1914.
NEON
37
counterpoise appears on the field of a fixed reading microscope.
The pressure is then carefully adjusted until the knob reaches
some definite arbitrary zero point and shows no tendency to
move. The pressure is then read off. The gas is now pumped
off and the same operation repeated with a gas of known
density such as pure oxygen. The ratio of the densities is
clearly the inverse of the pressures read, and as the latter are
low the molecular weight is given direct without any correc-
tions being required.
.Difficulties connected with temperature, so serious in density
determinations on the usual scale, are eliminated, for so minute
is the quantity of gas (about 0-0005 grm.) used that when this
is compressed inside the massive walls of the balance case
thermal equilibrium is almost instantaneous. The whole
operation of determining the density of a gas to 0-1 per cent,
can be completed in ten minutes. Only about half a cubic
centimetre of the gas is required for the operation.
26. Experiments on separation by distillation. The
first attempt at separation was made by continual fractionation
Y'.Y Y
FIG. 7. Fractionation Apparatus.
over charcoal cooled in liquid air. The apparatus used is
illustrated in the accompanying figure ; the method of working
was as follows :
38 ISOTOPES
The gas was admitted in a, one of the small charcoal bulbs
a, b, c, d, all cooled in liquid air. After a reasonable time had
elapsed the first fraction was pumped off by lowering mercury
in gas-holder A and opening the connecting stop-cock between
it and a. After another interval the stopcock was turned, the
mercury raised in A and the gas forced into bulb b. The
mercury was next lowered in both A and B, the former receiving
the second fraction from a while the latter withdrew the first
fraction of the gas now in b. The fundamental assumption on
which this arrangement was made was that at this stage, if
the vapour-pressures of the gases are nearly the same, the gas
in A would have the same composition as that left in b, and
that they therefore might be mixed. This was done by raising
the mercury, which not only drove the gas from A into b but also
the lightest fraction from B into c, where it again fractionated,
each process driving the lower boiling gas forward and keeping
the higher back.
The apparatus may contain any number of units, the whole
system being made cyclical and continuous by joining the
charcoal bulb at one end with the gas-holder at the other.
Four such units were actually employed, and after four opera-
tions the liquid air was removed from a and the residue it
contained was pumped off completely with an Antropoff pump
as the first contribution to the heaviest fraction ; in the same
way that in D was also pumped off as that of the lightest. The
bulb a was then immersed again in liquid air and the process
continued.
After about two-thirds of the gas had been collected in this
way as light and heavy fractions, that remaining was all
pumped out as the middle fraction. The process was next
repeated with the light and heavy fractions in turn, the inter-
mediate ones being combined by a definite rule.
By this arrangement, which does many operations at once,
the small quantity of helium contained in the original gas was
removed in a remarkably short time, after which the neon was
subjected to continual fractionation for three weeks. The
gas had now been through about 3000 fractionations and was
divided into seven main fractions ; the densities of these were
determined in order by the quartz micro-balance starting with
NEON 39
the lightest, the figures for the pressures giving the same zero
as oxygen at 76-35 were as follows :
(1) (2) (3) (4) (5) (6) (7)
121-05 120-95 121-05 120-90 121-00 121-05 121-05
The mean of these, 121-00, gives a molecular weight of 2049,
which is identical within experimental error with the accepted
one of 20-200 determined by Watson. It was evident that no
appreciable separation had been achieved.
A positive ray photograph was taken of the two extreme
fractions and this showed no appreciable change in the relative
intensity of the two parabolas. It was however a very good
one for the purpose of measurement and a careful comparison
of their displacements with those of the known lines due to
CO and C0 2 showed, with a probability almost amounting to
certainty, that the atomic weight of the lighter was not as
great as 20-20.
Encouraged by this evidence it was decided to make a
further attempt at separation by the method of fractional
diffusion.
27. Experiments on separation by diffusion. The first
apparatus used was much the same as that described by
Kamsay and Collie in their work on the diffusion of argon and
helium. 1 The diffusion was carried out at a low pressure and
the plug was made of two short lengths of clay pipe in series.
The method of fractionation was that described by Travers. 2
About 100 c.c. of neon was divided first into seven and later
into eight fractions. The complete series of fractionations
was repeated fifteen times, after which the two extreme
fractions were roughly purified over charcoal and their densities
measured. These indicated a difference of about a half per
cent., a very hopeful result moreover the lighter fraction
showed no appreciable quantity of helium even when analysed
by the method of positive rays which is much more delicate
than the spectroscope for this purpose.
The extremely laborious process was again taken in hand
1 Ramsay and Collie, Proc. Roy. Soc. 60A, 206, 1896.
* Travers, A Study of Oases, p. 289.
40 ISOTOPES
and the fractionation repeated another twenty-one times, at
the end of which the whole of the lightest fraction was lost by
a most unfortunate accident. This was the more serious as
the two extreme fractions had been systematically enlarged
with a view to fractionating each separately.
Despite this setback the fractionation of the heaviest 20 c.c.
was proceeded with. This was divided into five fractions and
fractionated ten times. The next lightest fraction to the one
lost was taken, divided into five parts and fractionated twelve
times. These very tedious operations were now brought to a
close and the two extreme fractions of 2 to 3 c.c. each were
purified over charcoal with the greatest possible care.
The final densities which further purification failed to alter
were 20-15 and 20-28 (Oxygen = 32). This change in density
is small but it is much too marked to be ascribed to con-
tamination or to experimental error. Looked at in the light
of modern knowledge there can be no reasonable doubt that
partial separation had been actually achieved. The extent of
the separation is about that to be expected from the theoretical
considerations of separation by diffusion given on page 127.
A spectroscopic examination of these two fractions showed no
appreciable difference between them.
These results were announced at the meeting of the British
Association at Birmingham in 1913 and at the same time the
evidence afforded by the positive ray photographs discussed.
This is available from three distinct considerations : the
character of the lines, their position and their intensity.
A careful examination of the plates showed, when proper
allowance had been made for difference of intensity, that the
two parabolas had characteristics identical with one another.
Both were prolonged towards the vertical axis showing that
the particles causing them were equally capable of carrying
more than one charge. 1 Now up to that time no cases of
multiple charges had been found to occur on molecules, but
only on atoms. One was therefore led to infer that both
lines were due to elements.
Measurements of the position of the parabolas relative to
those of CO and other known bodies in the discharge tube gave
1 F. p. 30.
NEON 41
consistent results, indicating that the lighter of the two corre-
sponded with an atomic weight less than 20-2, but the accuracy
was not sufficient to make this certain. The relative intensity
of the parabolas was estimated by three independent observers
as about 10 to 1. Its apparent invariability was valuable
corroborative evidence against the possibility of the 22 line
being due to the presence of other gases in the discharge tube*
28. Second attempt at separation by diffusion. In
order to carry out further diffusion experiments an elaborate
automatic diffusion apparatus was devised so as to avoid the
excessive labour of working by hand. This worked on the
see-saw principle and dealt with 300 c.c. of neon at a time.
It was started in 1914, but as it had little success in its object
there is no need to describe it in detail. It will be enough to
say that although it performed the mechanical operations of
diffusion many thousands of times in a satisfactory manner the
separation achieved was exceedingly poor actually only
about half that attained previously. This disappointing result
was undoubtedly due to the mistake made in designing it to
carry out the diffusion at atmospheric pressure, for under these
conditions the " mixing " is very bad. 1
When the work was interrupted by the war it could be said
that although the presence of two isotopes in neon was indicated
by several lines of reasoning, none of these could be said to
carry absolute conviction.
29. The analysis of neon by the Mass-spectrograph.
By the time the work was resumed in 1919 the existence of
isotopes among the products of radioactivity had been put
beyond all reasonable doubt by the work on the atomic weight
of lead 2 and was accepted generally. This fact automatically
increased both the value of the evidence of the complex nature
of neon and the urgency of its definite confirmation. It was
realised that separation could only be very partial at the best
and that the most satisfactory proof would be afforded by
measurements of atomic weight by the method of positive
rays. These would have to be so accurate as to prove beyond
1 F. p. 127.
2 F. p. 16.
42 ISOTOPES
dispute that the accepted atomic weight lay between the real
atomic weights of the constituents, but corresponded with
neither of them.
A new method of positive ray analysis was therefore worked
out which will be described in the next chapter. This proved
amply accurate enough for the purpose and the results obtained
from neon, which are given in detail on page 64, show beyond
any doubt that this gas is a mixture of two isotopes of atomic
wei hts 20-00 and 22-00 respectively.
CHAPTER V
THE MASS-SPECTROGRAPH
30. Limitations of the parabola method. The parabola
method of analysis of positive rays described in Chapter III,
though almost ideal for a general survey of masses and velocities,
has objections as a method of precision, many rays are lost by
collision in the narrow canal-ray tube ; the mean pressure in
which must be at least half that in the discharge-bulb ; very
fine tubes silt up by disintegration under bombardment ; the
total energy available for photography falls off as the fourth
power of the diameter of the canal-ray tube.
The first two objections can be overcome, as will be described
below, by replacing the brass or copper tube by fine apertures
made in aluminium, a metal which appears to suffer little
disintegration, and by exhausting the space between these
apertures to the highest degree by means of a subsidiary
charcoal tube or pump. The falling off in intensity of the
parabolas as one attempts to make them finer is a very serious
difficulty, as the accuracy and resolving power depend on the
ratio of the thickness to the total magnetic deflexion ; and
if we increase the latter the electric deflexion must be increased
to correspond and the parabolas are drawn out, resulting again
in loss of intensity.
Also the nature of the patch thrown on the plate by the use
of a long circular tube will clearly be the same as that caused
by the light from an evenly illuminated disc passing through a
circular aperture of the same diameter, that is to say it will
have a penumbra. Similarly the parabolic streak produced
by an infinite series of such patches will not be particularly
suitable for accurate measurements as it has no definite edges.
43
44 ISOTOPES
31. Methods of increasing the intensity of the spot.
The concentration of the stream of positive rays down the axis of
the discharge-bulb is very marked, but there is good evidence
for assuming that the intense part of the stream occupies
a considerable solid angle. This suggests the possibility
of an increase of intensity by means of a device which
should select the rays aimed at a particular spot on the plate,
whatever direction they come from; For example, a thin gap
between two coaxial equiangular cones would allow the rays
to be concentrated at the vertex. The dimensions of the
patch formed would be roughly those of one given by a cylin-
drical canal-ray tube of diameter equal to the width of the gap.
The increase of intensity would therefore be considerable ;
but the method is not easy to put into practice, and, in the
case of deflexions through large angles, would necessitate a
curved photographic surface.
Clearly the simplest way of increasing the intensity of the \
spot without increasing its dimensions, at any rate in one!
direction, is to use two paraUel straight slits. In the case of'
the parabola method this device would only be of use in a
special case such as the resolution of a close double, as the
parabolas will only be sharp at points where they are parallel
to the slit.
Such a slit system eliminates the difficulty of the penumbra
mentioned above, at any rate so far as measurements at right
angles to the line image are concerned.
32. Possibilities of "focussing." Beams of charged
particles which are homogeneous electrically (constant mv 2 /e)
or magnetically (constant mv/e) can be focussed like rays of
light by special devices. l The method of Dempster, described
in the previous Chapter, makes use of a form of magnetic
focussing. But the rays generated by the ordinary discharge
bulb are heterogeneous both in mv 2 and mv so that what is
required is an arrangement which will focus all rays of constant
mass, even though their velocity may vary over an appreciable
range.
33. Principle of the Mass-spectrograph. This purpose
1 Aston, Phil. Mag., 38, 709, 1919.
THE MASS-SPECTROGRAPH
45
is achieved by the arrangement illustrated diagrammatically
in Fig. 8. The exact mathematical analysis has now been
worked out by R. H. Fowler, 1 but it is proposed to give only
the approximate theory here for the sake of simplicity.
The rays after arriving at the cathode face pass through
two very narrow parallel slits of special construction Si S 2 ,
and the resulting thin ribbon is spread out into an electric
spectrum by means of the parallel plates P ls P 2 . After
emerging from the electric field the rays may be taken, to a
first order of approximation, as radiating from a virtual
FIG. 8. Diagram of Mass-Spectograph.
source Z half way through the field on the line Si S 2 . A group
of these rays is now selected by means of the diaphragm D,
and allowed to pass between the parallel poles of a magnet.
For simplicity the poles are taken as circular, the field between
them uniform and of such sign as to bend the rays in the
opposite direction to the foregoing electric field.
If 6 and <p be the angles (taken algebraically) through which
the selected beam of rays is bent by passing through fields
of strength X and H, then
2 = ZX (1), and <pv LH (
1 Aston and Fowler, Phil. Mag., 1922.
46 ISOTOPES
where I, L are the lengths of the paths of the rays in the fields.
Equation (1) is only true for small angles, but exact enough
for practice. It follows that over the small range of 6 selected
by the diaphragm 6v and yv are constant for all rays of given
e/m, therefore
66 26v , 6<p 6v
+ -- = 0, and - + = 0,
6 v y v
so that 60 _ 26(p
T '' ~fjT y
when the velocity varies in a group of rays of given e/m.
In order to illustrate in the simplest possible way how this
relation may be used to obtain focussing, let us suppose the
angles (exaggerated in the diagram) small and the magnetic
field acting as if concentrated at the centre of the pole-
pieces. If the breadth ZO = b, the group selected will be
spread out to a breadth btid at 0, and at a further distance r
the breadth will be
bdd + r(66 + 6q>) or <50& + rl + . . (3)
Now as ,the electric and magnetic deflexions are in opposite
directions, 6 is a negative angle. Say 6 = 0'. Then if
9?>20', the quantity (3) will vanish at a value of r given by
r (<p _ 26') = b . 2(9',
This equation appears correct within practical limits for large
circular pole-pieces.
Referred to axes OX, OY the focus is at r cos ( 9? 26'),
r sin (<p 26'), or r, b.26' ; so that to a first-order approxima-
tion, whatever the fields, so long as the position of the diaphragm
is fixed, the foci will all lie on the straight line ZF drawn
through Z parallelHo OX. For purposes of construction G
the image of Z in OY is a convenient reference point, 9? being
here equal to 40'. It is clear that a photographic plate, indi-
cated by the thick line, will be in fair focus for values of e/m
over a range large enough for accurate comparison of masses.
34. Optical analogue. It may be a help to form an
understanding of the principle of the apparatus if we suppose
that the beam is one of white light and the electric and magnetic
PLATE II.
Photograph of the Original Mass-Spectrograph set up in the Cavendish
Laboratory in 1919.
/?, Discharge Tube. A, Anode connected to high potential terminal of induction coil below
table. C, Reservoir containing gas to be analysed. I 1; I 2 , Charcoal-liquid air tubes exhausting
slit-system and camera. S, Soft iron plates to shield discharge from stray magnetic field. L,
Leads from high tension battery to electric plates. M, Du Bois electromagnet. T, Pea lamp
for photographing fiducial spot. F, Vacuum-tight and light-tight control for moving photo-
graphic plate. W, Camera showing light-tight cap on the left. H, Magnet circuit ammeter.
O, Magnet circuit control resistances. G, Gaede rotating mercury pump connected to the camera
and the discharge tube by glass tubes and stopcocks.
THE MASS-SPECTROGRAPH
47
fields are glass prisms deflecting the light in opposite directions.
The slit system acts as a collimator. If the glass of the first
prism has a coefficient of dispersion double that of the second
the heterogeneity of the rays of light will cause a spreading
of the beam identical with that caused by heterogeneity (in
respect to velocity) in the case of the positive rays. It will be
clear that if we make the angle of refraction of the second prism
more than double that of the first an achromatic image will
appear at F.
Since it is a close analogue of the ordinary spectrograph and
gives a " spectrum " depending upon mass alone the instrument
is called a " mass-spectrograph " and the spectrum it produces
a " mass -spectrum." It possesses one notable advantage
over the optical spectrograph for, although we can never
change the ratio of the dispersions, we can make the refractions
whatever we will by the control of X and H, and so bring any
desired range of the spectrum on to the plate.
35. The Discharge Tube. Fig. 9 is a rough diagram of the
arrangement of the mass-spectrograph when used for analysing
positive rays generated by the ordinary discharge tube method.
The discharge-tube B is an ordinary X-ray bulb 20 cm. in
A
FIG. 9. Mass-Spectrograph.
diameter. The anode A is of aluminium wire 3 mm. thick
surrounded concentrically by an insulated aluminium tube 7
mm. wide to protect the glass walls, as in the Lodge valve.
The aluminium cathode C, 2-5 cm. wide, is concave, about
8 cm. radius of curvature, and is placed just in the neck of the
bulb this shape and position having been adopted after a
short preliminary research. 1 In order to protect the opposite
1 Aston, Proc. Camb. Phtt. Soc. t 19, 317, 1919.
48 ISOTOPES
end of the bulb, which would be immediately melted by the
very concentrated beam of cathode rays, a silica bulb D about
12 mm. diameter is mounted as indicated. The use of silica
as an anticathode has the great advantage of cutting down
the production of undesirable X-rays to a minimum. The
cathode is earthed.
The discharge is maintained by means of a large induction-
coil actuated by a mercury coal-gas break ; about 100 to 150
watts are passed through the primary, and the bulb is arranged
to take from 0-5 to 1 milliampere at potentials ranging from
20,000 to 50,000 volts. Owing to the particular shape and
position of the electrodes, especially those of the anode, the
bulb acts perfectly as its own rectifier.
The method of mounting the cathode will be readily seen
from Fig. 10, which shows part of the apparatus in greater
FIG. 10. Mounting of Cathode of Mass-Spectrograph.
detail. The neck of the bulb is ground off short and cemented
with wax to the flat brass collar E, which forms the mouth of
an annular space between a wide outer tube F and the inner
tube carrying the cathode. The concentric position of the
neck is assured by three small ears of brass not shown. The
wax joint is kept cool by circulating water through the copper
pipe shown in section at G.
The gas to be analysed is admitted from the fine leak into
the annular space and so to the discharge by means of the
side-tube attached to F shown in dotted section at Q. Ex-
haustion is performed by a Gaede mercury-pump through a
similar tube on the opposite side. The reason for this arrange-
ment is that the space behind the cathode is the only part of
the discharge bulb in which the gas is not raised to an extremely
high potential. If the inlet or outlet is anywhere in front of
THE MASS-SPECTROGRAPH 49
the cathode, failing special guards, the discharge is certain
to strike to the pump or the gas reservoir. Such special guards
have been made in the past by means of dummy cathodes in
the bore of the tubes, but, notwithstanding the fact that the
gas can only reach the bulb by diffusion, the present arrange-
ment is far more satisfactory and has the additional advantage
of enabling the bulb to be dismounted by breaking one joint
only.
36 . The Slit System. The very fine slits used in this appar-
atus were made with comparative ease as follows : A cylinder of
pure aluminium about 10 mm. long by 5 mm. wide is carefully
bored with a hole 1 mm. diameter. The resulting thick- walled
tube is then cleaned and crushed with a hammer on an anvil until
the circular hole becomes a slit about -3 mm. wide. Continuation
of this treatment would result in a slit as fine as required giving
the maximum resistance to the passage of gas, but its great
depth would make the lining up of a pair a matter of extreme
difficulty. The crushed tube is therefore now placed between
two V-shaped pieces of steel and further crushed between
the points of the V's at about its middle point until the required
fineness is attained. Practice shows that the best way of
doing this is to crush until the walls just touch, and then to
open the slit to the required width by judicious tapping at
right angles to that previously employed. With a little care
it is possible to make slits with beautifully parallel sides to
almost any degree of fineness, -01 mm. being easily attainable.
At this stage the irregularly shaped piece of aluminium is not
suited to accurate gas-tight fitting ; it is therefore filled
with hard paraffin to protect it from small particles of metal,
etc., which if entering cannot be dislodged owing to its
shape, and turned up taper to fit the standard mountings.
After turning, the paraffin is easily removed by heat and
solvents. The centre of the cathode is pierced with a 3 mm.
hole, the back of which is coned out to fit one of the stan-
dard slits Si. The back of the cathode is turned a gas-
tight fit in the brass tube 2 cm. diameter carrying it, the
other end of which bears the brass plug H which is also coned
and fitted with the second slit S 2 . The two slits, which are
E
50 ISOTOPES
roughly -05 mm. wide by 2 mm. long, can be accurately
adjusted parallel by means of their diffraction patterns. The
space between the slits, which are about 10 cm. apart, is kept
exhausted to the highest degree by the charcoal tube I t .
By this arrangement it will be seen that not only is loss of
rays by collision and neutralisation reduced to a minimum
but any serious leak of gas from the bulb to the camera is
eliminated altogether.
37. The Electric Field. The spreading of the hetero-
geneous ribbon of rays formed by the slits into an electric
spectrum takes place between two parallel flat brass surfaces,
J 1} J 2 , 5 cm. long, held 2-8 mm. apart by glass distance-pieces,
the whole .system being wedged immovably in the brass con-
taining- tube in the position shown. The lower surface is
cut from a solid cylinder fitting the tube and connected to it
and earth. The upper surface is a thick brass plate, which
can be raised to the desired potential, 200-500 volts, by means
of a set of small storage-cells. In order to have the plates
as near together as possible, they are sloped at 1 in 20 i.e.
half the angle of slope of the mean ray of the part of the
spectrum which is to be selected by the diaphragms. Of these
there are two : one, K 1? an oblong aperture in a clean brass
plate, is fixed just in front of the second movable one, K 2 ,
which is mounted in the bore of a carefully ground stopcock L.
The function of the first diaphragm is to prevent any possibility
of charged rays striking the greasy surface of the plug of the
stopcock when the latter is in any working position. The
variable diaphragm is in effect two square apertures sliding
past each other as the plug of the stopcock is turned, the fact
that they are not in the same plane being irrelevant. When
the stopcock is fully open as sketched in Fig. 10 the angle of
rays passing is a maximum, and it may be stopped down to
any desired extent by rotation of the plug, becoming zero before
any greasy surface is exposed to the rays. Incidentally the
stopcock serves another and very convenient use, which is to
cut off the camera from the discharge tube, so that the latter
need not be filled with air each time the former is opened to
change the plate.
THE MASS-SPECTROGRAPH
51
38. The Magnetic Field. After leaving the diaphragms
the rays pass between the pole-pieces M of a large Du Bois
magnet of 2500 turns. The faces of these are circular, 8 cm.
diameter, and held 3 mm. apart by brass distance-pieces.
The cylindrical pole-pieces themselves are soldered into a brass
tube 0, which forms part of the camera N. When the latter is
built into position, the pole-pieces are drawn by screwed
bolts into the arms of the magnet, and so form a structure
of great weight and rigidity and provide an admirable founda-
tion for the whole apparatus. Current for the magnet is
provided by a special set of large accumulators. With a
potential of 300 volts on the electric plates the hydrogen lines
are brought on to the scale at about 0-2 ampere, and an increase
to 5 amperes, which gives practical saturation, only just brings
the singly-charged mercury lines into view. The discharge
is protected from the stray field of the magnet by the usual
soft iron plates, not shown.
39. The Camera. The main body of the camera N is
made of stout brass tube 6-4 cm. diameter, shaped to fit on to
the transverse tube containing
the pole-pieces. The construc-
tion of the plate-holder is indi-
cated by the side view in Fig. 9
and an end-on view in Fig. 11.
The rays after being magnetically
deflected pass between two verti-
cal earthed brass plates Z, Z
about 3 mm. apart, and finally
reach the photographic plate
through a narrow slot 2 mm.
wide, 11-8 cm. long, cut in the
horizontal metal plate X, X.
The three brass plates forming
a T-shaped girder are adjusted and locked in position by a
set of three levelling-screws, at each end ; the right-hand
upper one is omitted in Fig. 11. The plates Z, Z serve to
protect the rays completely from any stray electric field,
even that caused by the photographic plate itself becoming
w
FIG. 11. The Plateholder of
the Camera.
52 ISOTOPES
charged until within a few millimetres of their point of
impact.
The photographic plate W, which is a 2 cm. strip cut length-
wise from a 5 x 4 plate, is supported at its ends on two narrow
transverse rails which raise it just clear of the plate X, X.
Normally it lies to the right of the slot as indicated, and to make
an exposure it is moved parallel to itself over the slot by means
of a sort of double lazy-tongs carrying wire claws which bracket
the ends of the plate as shown. This mechanism, which is not
shown in detail is operated by means of a torque rod V working
through a ground glass joint. Y is a small willemite screen.
The adjustment of the plate-holder so that the sensitised
surface should be at the best focal plane was done by taking
a series of exposures of the bright hydrogen lines with different
magnetic fields on a large plate placed in the empty camera
at a small inclination to the vertical. On developing this,
the actual track of the rays could be seen and the locus of points
of maximum concentration determined. The final adjustment
was made by trial and error and was exceedingly tedious, as
air had to be admitted and a new plate inserted after each
tentative small alteration of the le veiling-screws.
40. Experimental procedure. The plate having been
dried in a high vacuum overnight, the whole apparatus is
exhausted as completely as possible by the pump with the
stopcock L open. I x and I 2 are then cut off from the pump by
stopcocks and immersed in liquid air for an hour or so. The
electric field, which may range from 200 to 500 volts, is then
applied and a small current passed through the magnet sufficient
to bring the bright hydrogen molecule spot on to the willemite
screen Y, where it can be inspected through the plate-glass
back of the cap P. In the meantime the leak, pump, and coil,
have all been started to get the bulb into the desired state.
When this has become steady, Jj, is earthed to prevent any
rays reaching the camera when the plate is moved over the slot
to its first position, which is judged by inspection through P
with a non-actinic lamp. The magnet current having been set
to the particular value desired and the diaphragm adjusted,
the coil is momentarily interrupted while Jj is raised to the
THE MASS-SPECTROGRAPH
53
desired potential, after which the exposure starts. During this,
preferably both at the beginning and the end, light from a lamp
T is admitted for a few seconds down the tube R (Fig. 9) the
ends of which are pierced with two tiny circular holes. The
lower hole is very close to the plate, so that a circular dot or
fiducial spot is formed from which the measurements of the
lines may be made.
The exposures may range from 20 seconds in the case of
hydrogen lines to 30 minutes or more, 15 minutes being usually
enough. As soon as it is complete the above procedure is
repeated, and the plate moved into the second position. In this
way as many as six spectra can be taken on one plate, after
which L is shut, I 2 warmed up, and air admitted to the camera.
The cap P, which is on a ground joint, can now be removed,
and the exposed plate seized and taken out with a special pair
of forceps. A fresh plate is now immediately put in, P replaced
and the camera again exhausted, in which state it is left till
the next operation.
41. Form of the Spectrum Lines. Owing to the form of
the slits used, the shape of the spot formed when undeflected
rays from such a slit system strike a photograph surface
normally, is somewhat as indicated at a (Fig. 12). When they
strike the plate obliquely the image would be spread out in one
direction, as in b. This would be the actual form in the
a b c
FIQ. 12. Form of the Spectrum Lines.
apparatus, if the deflexions of the mean and extreme rays (i.e.,
the rays forming the centre and the tips) were identical. This is
true of the magnetic field since each cuts the same number
of lines of force ; but it is not so in the case of the electric
deflexion. Owing in part to the fact that the plates J 1} J 2 are
rectangular and in part to the stray field between the charged
plate Ji and the earthed tube in which it is mounted, the
54 ISOTOPES
extreme rays passing diagonally will be deflected more than
the mean rays and the spot bent into the form shown at c.
The convex side will be in the direction of the magnetic de-
flexion, as this is opposed to the deflexion causing the bend.
The image on the plate will therefore be the part of this figure
falling on the narrow slot in X, X ; and as the apparatus is not
exactly symmetrical, its shape in the spectra is the figure lying
between the lines X, X in Fig. 12, c.
42. The distribution of the mass -spectrum over the
photographic plate. In order to study the positions of the
focus F (Fig 8) on the plate corresponding to different values
of the effective mass m when X and H are constant, we may
assume perfect focussing and only consider a single median
ray. If R is the radius of curvature of the path of a ray of
effective mass m while in the magnetic field, and d the radius
of the field, clearly tan \ (p d/R. But X and 6 are constant,
hence mv z must be constant so that the radius of curvature
in the magnetic field varies as \/m. We may therefore write
tan J <p \/(m /m) . . (4)
where m is a constant and can be interpreted as that mass
which under the conditions of the experiment is bent through
a right angle in the magnetic field.
Again if ON the length of the perpendicular dropped from
the centre of the magnetic field upon ZF = p (a constant)
then
NF = p cot ((p 26) . . . (5)
By combining (4) and (5) we get an expression for NF/p in
terms of m and m. This is complicated, 1 but its differential
can be shown to vanish when tan J 99 = tan 26. Thus the
mass-scale is approximately linear near (p = 40.
This linear law was observed experimentally at the very
outset and though at the time it was unexplained it added
greatly to the ease and accuracy of the determinations of m.
The quantity actually measured is the distance between a
fixed point on the photographic place called the " fiducial
spot " 2 and the focussed image F. Let us call this distance
D. D and NF differ by a constant k about 5-4 cm. in the
1 Loc. cit. a V. p. 53.
THE MASS-SPECTROGRAPH 55
present apparatus so that the relation between D and m has
the form D = / (m/m ) where / is a function in which all the
coefficients p, k, and tan 26 are geometrical constants, the
fields only affect m . It follows directly that so long as the
apparatus is rigid : If Dj and D 2 are the distances from the
fiducial spot of any two points on the plate and m L and m a
the corresponding masses for given values of D! and D 2 , the
ratio mi/mz will be the same in every photograph.
43. Practical method of deducing the effective mass of
a particle from the position of its line on the photograph.
The mathematical investigation described above is of interest
as it explains the results obtained, but the actual determination
of masses from mass -spectra is a purely empirical process, and
consists in the comparison of the positions of the lines caused
by the masses in question with the positions of known reference
lines. The only assumption made was that given at the end
of the previous paragraph and even this was capable of verifica-
tion by experiment, using such methods as that described on
p. 57, or even more fundamentally, in the special case of the
ratio 2/1, by the known identity of the mass ratios 2 /0,
0/0 + + , and C/C + + .
The reference lines used at the outset of the work were
lines given by particles of elements and compounds the relative
masses of which were known to at least the order of accuracy
aimed for. The procedure was somewhat as follows. A series
of spectra were taken with say a mixture of C0 a and CH 4
in the discharge tube. Previous experience with the parabola
method of analysis led to the expectation that lines at 6-C + + ,
8-0 ++ , 12-C, 16-0, 28-CO, 32-0 2 , 44-C0 2 would certainly be
present, there would also be a series of hydrocarbon lines
between 12 and 16, CH, CH 2 , CH 3 which could be regarded as
known. A spectrum was selected containing as many as
possible of these known lines and their masses m t , m 2 , m 3 ,
were plotted against the distances of the lines from the fixed
fiducial spot and a curve drawn through the points so obtained.
This is our first calibration curve of necessity inaccurate owing
to the gaps between the points. A second spectrum was now
taken in which the same lines appeared in a different place,
56 ISOTOPES
for by altering the magnetic field we can place them wherever
we please, and the new set of distances from the fiducial spot
measured. These distances were now transformed into masses
(no longer integral) m'i,m'i,m' s , by means of the curve pre-
viously drawn. Supposing the curve to be accurate and the
.. -, i 1 1 9% i wi* m\ , , ,
ratio law to hold = - = - = r where r is clearly a
ra j m 2 m 3
measure of the change in ra in the mathematical discussion
above. In practice these ratios were found to be very nearly
the same, so that a mean value of r could be taken with
confidence. The known masses multiplied by that mean now
gave a new set of points on the original curve. By carrying
on this process all the serious gaps in the curve could be
bridged and its accuracy brought up to the required standard.
The calibration curve so formed renders the identification of
one line sufficient to deduce the masses corresponding to all
the other lines on the plate, and as in general many lines are
known on each spectrum, its accuracy is continually subject
to fresh test. In practice it was found perfectly reliable so
long as none of the geometrical constants of the apparatus
were altered.
Owing to the linear relation at q> = 40 the actual curve was
very nearly straight for a considerable portion of its length.
This allowed the following alternative procedure to be adopted
if desired. A linear relation was assumed and a table of
corrections made by means of reference lines, and these correc-
tions when subtracted from the observed displacements gave
an exactly linear relation with mass. A correction-curve
(apparently parabolic) was drawn, from which the appropriate
correction for any displacement could be written down and
the mass corresponding to this displacement obtained by
simple proportion.
In connection with the use of reference lines it might be
thought difficult to know which of the lines on a plate corre-
sponds to a known mass, since they are not labelled in any
way. A little consideration will show that the same difficulty
is raised in the case of the standard lines of the iron arc and
the stars in the sky, yet neither the spectroscopist nor the
astronomer have the least difficulty in recognising enough for
THE MASS-SPECTROGRAPH 57
their purpose, indeed a mistake in identity would lead in
most cases to an error so gross as to compel immediate atten-
tion. This comparison is perhaps a little flattering to the
lines on a mass-spectrum as these alter their relative intensity
to some extent, but in particular cases, such as those of the
hydrocarbons and mercury, identification is, after a little
experience, as easy as that of the Pole Star or of the D lines in
the spectrum of sodium.
44. Comparison of masses by the method of " coin-
cidence." The method of deducing the masses of particles
from the position of their lines described in the foregoing
paragraph is simple and straightforward. It also has the
great advantage of not requiring an accurate knowledge of
the numerical values of the electric and magnetic fields. The
only requisite is that these should be constant during the
exposure, and even if this constancy is not quite perfect the
shift in position will affect all the lines known and unknown
alike and therefore introduce no serious error into the results
obtained. There is, however, another method of comparing
masses which requires no knowledge, either theoretical or
empirical, of the relation between effective mass and measured
displacement. This is independent of the calibration curve
and therefore constitutes a valuable check on results obtained
by its use. It depends upon the following considerations :
Suppose we wish to compare an unknown mass m' with a
known mass m. A mass-spectrum is taken with fields X and
H such that the mass m gives a line at a certain position on
the plate. The fields are now altered until the line caused
by the unknown mass m' is brought to the identical position
on the plate previously occupied by the line due to m. The
paths of the rays in the two cases must be identical, hence if
X', H' are the new values of the fields it follows at once from
equations (1) and (2) * that ra'/ra = X/X' x (H'/H) 2 . Now
it is only necessary to measure one of the fields if we keep the
other constant and therefore H, which cannot be measured or
reproduced accurately, is kept constant, and X is measured.
For the latter purpose it is only necessary to measure the
1 V. p. 45.
58 ISOTOPES
potentials applied to the plates P 1} P 2 , which can be done with
the greatest ease and accuracy.
Thus, to take a numerical illustration, the position occupied
by the line due to carbon (12) with a potential on the plates
of 320 volts should be exactly coincident with that occupied
by the line due to Oxygen (16) with 240 volts when the mag-
netic field is kept constant. All such coincidences have so
far been found to occur within the error of experiment, what-
ever the position on the plate.
Methods depending on the measured variation of X with H
constant have some practical disadvantages. The first and
most obvious of these is that any small change in the value
of the magnetic field between the two exposures will lead to
a definite error, this error will be double the percentage change
in the field, since the square of the latter is involved. The
second objection is founded on considerations of intensity.
If the parabola method of analysis is compared with the
mass-spectrograph it will readily be observed that, in effect,
the latter focusses at a point all the rays which in the former
method form a short element of arc on a parabola. The
length of the element of arc is determined by the angle of
the electric spectrum allowed to pass, i.e. the width of the
diaphragm. Its position on the parabola is at our disposal,
for, referring to Fig. 4, p. 28, it will be seen that the higher we
make X, that is to say the higher the energy of the beam of
rays we select at constant 6, the nearer the element of arc will
approach the axis OY, in fact its distance from that axis will
simply be inversely proportional to X. Also, however many
parabolas we consider and however much we move them about
by changing H, so long as X is constant the elements of arc
selected will all lie on a line parallel to OY. Now it has
already been pointed out 1 that the intensity of normal para-
bolas is a maximum near the head p, where the energy corre-
sponds to the full fall of potential across the discharge tube,
and fades away rapidly, in some cases very rapidly indeed,
at points more distant from the origin. In order to get the
greatest intensity at the focussed spot we must therefore
choose X so that the element of arc selected will be near the
1 P. 29.
THE MASS-SPECTROGRAPH 59
head of the parabola. This is done in practice by observing
visually, by means of a willemite screen, the very bright line
given by the hydrogen molecule while different potentials are
applied to the plates. The best value of X so determined
must also be the best value for all the other normal lines, so
that in the ordinary calibration curve method, when X is kept
constant, it is possible to use conditions in which all the normal
lines on the mass-spectra will be at their brightest together,
whatever range we bring on to the plate by altering the
magnetic field.
In the coincidence method this very fortunate circumstance
cannot be taken advantage of, for with H constant the selected
elements of arc will now lie on a line parallel to OX. We
can only arrange matters for one, the lighter, of the two
masses to be compared, to be at its optimum. In the case
of the heavier the selected arc must lie at a greater distance
from the origin and therefore provide a much feebler intensity.
The disparity in brightness, due to this effect will be the greater
the greater the ratio of the masses considered ; it can be
corrected to some degree by softening the discharge tube
while the heavier mass is being photographed.
In spite of these drawbacks the principle underlying the
coincidence method is probably the most suitable for mass-
ratio measurements of the highest accuracy. The fact that
the paths of the rays is the same in the case of both masses
eliminates all errors due to non-uniformity of the fields and
the results are independent of any assumptions as regards
the ratios of the reference lines themselves. It is the only
method at present available in the case of elements far removed,
on the mass-scale, from the reference lines, and a modification
of it called the method of " bracketing " has been successfully
used to evaluate the masses of helium and hydrogen. 1
45. The measurement of the lines. The accurate
determination of the distance of the lines from the fiducial
spot is a physical problem of considerable interest. The
image itself is due to a caustic of rays, the edge of which will
be sharp on the side of maximum magnetic displacement, so
1 F. p. 69.
60 ISOTOPES
that this, the left side in the Plates, may be expected to main-
tain its sharpness when a large diaphragm is in use, while the
other will fade away gradually. Hence very bright lines will
be broadened to the right by this effect (which is analogous
to spherical astigmatism in ordinary lenses), but to the left
the only broadening will be that due to ordinary halation.
The relative importance of these two forms of spreading can
be gauged by taking photographs with a very small diaphragm,
for then the first will be eliminated and the second can be
estimated by comparing lines of different intensity. It is
found that for ordinary diaphragm apertures the halation
effect is much the smaller ; it can also be minimised by using
lines of approximately equal intensity so that the most reliable
measurements of lines for position are obtained from their
left-hand edges. This is well illustrated in the " bracketed "
lines of hydrogen a and c, Plate III. In (a) measurements
of the left hand side of the three lines shows this bracket to
be really symmetrical though it does not appear so to the eye,
on account of the astigmatic spreading of the middle line
caused by the use of an open diaphragm and rather too long
an exposure. In (c) the diaphragm was almost closed and
the exposures more carefully adjusted, so that both sides of
the lines are sharp and their breadths practically identical.
The most accurate measurements were made on a compara-
tor. The spectrum was set as closely as possible parallel to
the axis of the instrument, and the distances between the
left-hand edge of the lines and the fiducial spot read off on a
Zeiss standard scale. For faint lines it was necessary to use
a very low power eyepiece of the reading microscope, and in
the case of the faintest lines of all, the best results could be
obtained by laying a millimetre scale on the plate and estimat-
ing the distance from the fiducial spot to the optical centre of
the lines, by the unaided eye.
46. Resolving power and accuracy of mass deter-
mination. Taking the width of the slits as 1/25 mm. and
putting in the dimensions of the present apparatus the theory
shows that in the region <p = 40 lines differing by a little less
than 1 per cent, should be just separated. In actual practice
THE MASS-SPECTROGRAPH 61
a better result was obtained, for the instrument is capable of
separating the lines of xenon, which differ by 1 in 130 ; this
is probably because the part of the line which falls on the strip
of plate exposed is due to the narrower edges of the slits.
The numerical relation between mass and position in this
part of the spectrum corresponds to a shift of 1-39 mm. for
a change of mass of 1 per cent., so that even with the unaided
eye an accuracy of 1 part in 1 ,000 can be approached. Although
it is sufficient in theory to know the mass of one line only to
determine, with the calibration curve, the masses of all the
others, in practice every effort is made to bracket any unknown
line by reference lines and only to trust comparative measure-
ments when the lines are fairly close together. Under these
conditions an accuracy of 1 in 1,000 is claimed and there is
little doubt that in favourable cases it is exceeded.
47. Order of results and nomenclature. In the
descriptions of the results obtained with the mass-spectrograph
contained in the following chapters the order of the elements
given is, when possible, that in which the experiments were
made. There is a practical reason for this procedure, as in
most cases it was impossible to eliminate any element used
before the following one was introduced. Evacuation and
washing have little effect, as the gases appear to get embedded
in the surface of the discharge bulb and are only released very
gradually by subsequent discharge.
The problem of nomenclature of the isotopes became serious
when the very complex nature of the heavy elements was
apparent. It has been decided for the present to adopt the
rather clumsy but definite and elastic one of using the chemical
symbol of the complex element, with an index corresponding
to its mass : e.g. Ne 22 , Rb 87 . This system is made reasonable
by the fact that the constituents of complex elements have
all so far proved to have masses expressible in whole numbers.
48. Lines of the First, Second and higher Orders.
It was shown on page 30 that particles having two charges
gave a parabola corresponding to an effective mass of one
half the normal mass. In the same way a particle with three
charges will have an effective mass of one third, and so on.
62 ISOTOPES
These apparent masses will duly make their appearance on
mass-spectra as lines corresponding to simple fractions of the
real mass causing them. It is convenient in these cases to
borrow the nomenclature of optics and refer to the lines given
by singly, doubly, and multiply charged particles respectively
as lines of the first, second, and higher orders. Thus the
molecule of oxygen gives a first order line at 32, and its atom
first and second order lines at 16 and 8.
The empirical rule that molecules only give first order lines 1
is very useful in helping to differentiate between atoms and
compound molecules of the same apparent mass. Some
results given below, 2 however, show that in certain cases it
breaks down, so that inferences made from it must not be taken
as absolutely conclusive.
49. Negative mass -spectra. It has been mentioned
that positive rays could become negatively charged by the
capture of electrons by collisions in the narrow canal-ray tube
of the Thomson apparatus, and so produce parabolas in the
quadrant opposite to that containing the normal ones. The
slit system of the mass-spectrograph is specially designed to
eliminate such collisions as far as possible by exhausting the
space between the slits. If the means of exhaustion of this
space is deliberately cut off, and the normal electric and
magnetic fields both reversed in sign it is possible, at a small
cost in definition of the lines, to photograph the mass-spectra
of negatively charged particles. Such negatively charged
particles are only formed by elements or compounds having
marked electronegative properties. Very little work has been
done in this interesting field, but certain ambiguities in the
interpretation of the chlorine results have been satisfactorily
cleared up by its means.
1 J. J. Thomson, Rays of Positive Electricity, p. 54.
2 F. p. 75.
CHAPTER VI
ANALYSIS OF THE ELEMENTS
50. Arrangement of results. In this Chapter and the
one following it are given the experimental results obtained
from a large number of elements which have been subjected
to analysis with a view to determining their constitution.
This Chapter deals with those elements which, by reason of
their volatility or properties of forming volatile compounds,
can be treated by the ordinary discharge-tube method. The
analysis given in all these cases is that obtained by means of
the mass-spectrograph.
In Chapter VII will be found the results obtained by the
analysis of those elements, all metals, whose positive rays must
be generated by special devices. Here the analyses are
effected by several different methods.
The sequence of the elements in the two Chapters is that in
which the results were obtained ; with the exception of nickel,
which is included in the first group although its mass-spectrum
was not obtained until after the other metals had been under
observation.
51. Oxygen (At. Wt. 16 00) and Carbon (At. Wt. 12-00).
On a mass-spectrum all measurements are relative, and so
any known element could be taken as a standard. Oxygen
is naturally selected. Its molecule, singly-charged atom, and
doubly-charged atom give reference lines at 32, 16, and 8
respectively. The extremely exact integral relation between
the atomic weights of oxygen and carbon is itself strong
evidence that both are " simple " elements, and so far no
evidence appears to have arisen to throw any doubt on this
point. Direct comparison of the C line (12) and the CO line
(28) with the above standards shows that the whole number
relation and additive law hold to the limit of accuracy, i.e. one
63
64 ISOTOPES
part in a thousand; and this provides standards C ++ (6)
C (12), CO (28), and C0 2 (44).
Many of these lines will be recognised on the spectra repro-
duced on Plate III. The compounds of carbon and hydrogen
provide two valuable and easily distinguishable groups of
reference lines. The first, which may be called the Ci group,
contains five : 12-C, 13-CH, 14-CH 2 , 15-CH 3 , 16-CIL
(or 0). It is very well shown on Spectrum V, Plate III.
When water vapour is present, and particularly when a fresh
discharge-tube is used for the first time, it is followed by
17-OH, 18-OH 2 , and sometimes by 19 presumably OH 3
but always very faint. The second hydrocarbon or C 2 group
contains seven lines : 24, 25, 26, 27, 28, 29, 30, which include
the very strong and particularly valuable reference line 28 CO
or C 2 H 4 . This group is well illustrated in Spectra I and II,
Plate III. All the above lines may be expected on spectra
obtained by the ordinary discharge-tube method ; for an
addition of CO or C0 2 is usually made to the gases or vapours
under consideration and assists the smooth running of the
discharge. The hydrocarbons are derived from the wax and
grease used in the joints of the apparatus.
52. Neon (At. Wt. 20-20). As soon as the instrument was
found to work satisfactorily and enough mass-spectra contain-
ing reference lines had been obtained, neon was introduced
into the discharge tube. The best results were obtained with
a mixture of carbon monoxide and neon, containing about 20
per cent, of the latter gas.
The two first order and two second order lines due to neon
were all four available and well placed for measurement on
the mass spectra obtained. The following figures are taken
from the original paper j 1 they are the results of the measure-
ments made on two different plates, using six different spectra.
PLATE 1.
First order. Second order.
20-00 22-00 9-98 11-00
19-95 22-01 10-02 10-99
19-97(5) 22-00(5) 10-00 10-99(5)
1 Aston, Phil. Mag., 39, 454, 1920.
ANALYSIS OP THE ELEMENTS 65
PLATE 2.
20-00 21-90 10-01 11-06
19-98 22-10 9-98 10-98
20-00 22-03 9-98 11-01
19-90 21-98
19-97 22-00(5) 9-99 11-01
The method of measuring the position of the lines then in
use, combined with a photographic halation effect, 1 tended to
decrease the masses given by very bright lines. This is enough
to account for the reading of the intense 20 line giving a mass a
little too low. The above figures therefore can be accepted as
conclusive evidence that neon is a mixture of two isotopes of
atomic weights 20-00 and 22-00 (0 = 16) respectively, to an
accuracy of about one-~fcenth per cent. 2
The two first order lines of neon are shown in Spectrum I,
Plate III, but, of course, their relative intensities must not be
judged from such a half-tone reproduction. On the original
negatives the intensities are in good agreement with the
expected ratio 9 to 1 which is necessary to yield the accepted
atomic weight 20-20.
53. Possibility of a third isotope of neon. On some of
the clearest spectra obtained with neon present there is a
distinct indication of a line corresponding to a mass 21. This
is an exceedingly faint line and, at first, was thought to indicate
the presence of a third isotope. It is now considered more
probably due to an abnormal hydride of the kind discussed on
page 98.
54. Chlorine (At. Wt. 35-46). Spectra indicating that this
element was a mixture of isotopes were first obtained by the
use of hydrochloric acid gas, but as this was objectionable on
account of its action on mercury, phosgene (COC1 2 ) was sub-
stituted. Spectra II, III, and IV, Plate III, are reproduced
from one of the plates taken with this gas. Spectrum I is
reproduced for comparison, it shows the state of the tube
before chlorine compounds were introduced. It will be seen
1 F. p. 60.
2 Aston, Nature, Nov. 27, 1919 ; Phil. Mag. 39, 454, 1920.
F
66 ISOTOPES
that chlorine is characterised by the appearance of four very
definite lines in the previously unoccupied space to the right
of 2 (32) : measurement shows these lines to correspond
exactly to masses 35, 36, 37, and 38. On Spectrum II, Plate
III, taken with a small magnetic field, faint lines will be seen
at 17-5 and 18-5. These only appeared when chlorine was
introduced, and are certainly second order lines corresponding
to 35 and 37. Chlorine is therefore a mixture of isotopes, and
two of these have masses 35 and 37. Evidence that Cl 35 and
Cl 37 are the main if not the only constituents is given by the
strong lines 63 and 65 (Spectrum IV, Plate III), due to COC1 35
and COC1 37 . The lines 36 and 38 were naturally ascribed to
the hydrochloric acids corresponding to Cl 35 and Cl 37 . 1 That
this surmise is correct was definitely proved about a year later
when the mass spectra of negatively charged rays of chlorine
were successfully obtained in the manner described on p. 62.
On the negative mass spectra produced in this way only the
two chlorine lines 35 and 37 could be distinguished. The
property of forming negatively charged ions is a purely chemical
characteristic ; that isotopes of the same element should differ
radically in it is quite out of the question. It is therefore
perfectly certain that the lines 36 and 38 are not, to any
sensible extent, due to isotopes of chlorine.
On many of the spectra obtained from chlorine compounds
a very faint line is distinguishable at 39. This was regarded
as a possible third isotope (which would then be an isobare of
potassium. No decision on this point has been obtained from
the negative mass spectra, for these have, so far, been too
faint for the 39 line to be visible, even if it was present. A
careful comparison between the intensity of this line and those
at 35 and 37 on a large number of plates discloses an apparent
variation which tells rather decidedly against the idea that a
third isotope is present. More evidence, however, will be
necessary to clear this point.
55. Argon (At. Wt. 39-88 Ramsay, 39-91 Leduc). The
tube was run with a mixture of C0 2 and CH 4 , and then about
20 per cent, of argon added. The main constituent of the
1 Aston, Nature, Dec. 18, 1919 ; Phil. Mag., 39, 611, 1920.
PLATE III.
*
J^
- 80
-83
"" 84
13-3 *_
- 15"
ii6
! || 44
!
83
p
i-/6
40
._ ,
MASS SPECTRA.
permission of the Editors of The Philosophical Mayazine.)
\\
ANALYSIS OF THE ELEMENTS 67
element was at once evident from a very strong line at 40
(Spectrum VI, Plate III), reproduced in the second and third
orders at 20 and 13-33 (Spectrum V). The third order line is
exceedingly well placed for measurement, and from it the mass
of the singly-charged atom is found to be 40-00 02. At
first this was thought to be the only constituent, but later a
faint companion was seen at 36, which further spectra showed
to bear a very definite intensity relation to the 40 line. No
evidence drawn from multiple charges was available in this
case owing to the probable presence of OH 2 and C ; but the
above intensity relation and the absence of the line from
spectra taken just before argon was introduced, made it
extremely likely that it was a true isotope.
Any doubt on this point has been removed for all practical
purposes by results obtained during the later work on krypton
and xenon. Argon was always present to more or less extent
during these experiments and the invariable association of a
line at 36, of appropriate intensity, with the stronger jpne at
40 may be regarded as confirming the original conclusion in a
satisfactory manner. The presence of 3 per cent, of this
lighter isotope is sufficient to reduce the mean atomic weight
from 40 to 39-9.
56. Nitrogen (At. Wt. 14-01). This element shows no
abnormal characteristics : its atom cannot be distinguished,
on the present apparatus, from CH 2 nor its molecule from CO.
Its second order line on careful measurement appears to be
exactly 7, so it is evidently a simple element, as its chemical
combining weight would lead one to expect.
57. Hydrogen (At. Wt. 1-008) and Helium (At. Wt.
3-99). In connection with the analysis of positive rays the
element hydrogen is of peculiar interest in many ways. Its
invariable presence in rays generated by the ordinary dis-
charge-tube method, no matter what gas is being employed, is
itself a very striking phenomenon, even when due allowance
has been made for its abnormal power in affecting screens and
plates.
The ease with which its brilliant lines, the molecular one in
particular, can be generated and observed visually is of an
68 ISOTOPES
importance hardly to be exaggerated in the development and
technique of the mass-spectrograph. The advantage of the
visible presence of the H 2 line has already been referred to *
and was realised very keenly in the investigation of the alkali
metals when the method precluded the use of this line to
indicate when suitable conditions for exposure had been
obtained. 2
The hydrogen atom is the lightest particle ever observed to
carry a positive charge, which agrees very well with the
generally accepted idea that the true Moseley number of this
element is 1. This implies that the neutral atom of hydrogen
only contains one electron and therefore can only acquire a
single positive charge in losing it. The singly charged par-
ticle so formed is therefore the " proton " or ultimate atom
of positive electricity itself.
Helium, on the other hand, can lose two electrons and
acquire a double charge, indeed its atoms are invariably in
this state when ejected from the nuclei of radioactive elements
as alpha rays. Nevertheless, in spite of every effort to obtain
the second order line of helium for direct comparison with
the hydrogen molecule not the faintest indication of it has
yet been observed on a mass spectrum, although there is not
the least difficulty in obtaining its first order line to any
intensity required.
The explanation oi this is probably to be found in the very
high ionisation potential about 80 volts 3 associated with
the detachment of both electrons. If doubly charged helium
atoms are formed in the discharge tube and we have every
reason to consider this probable their chance of passing
through the slit system and the deflecting fields without
picking up a single electron may be practically nil. This is
made the more likely by the fact that helium is not absorbed
by charcoal and liquid air, so that when it is present the pressure
in the apparatus tends to become undesirably high.
1 F. p. 52.
8 F. p. 87.
3 Franck and Knipping, Phys. Zeit., 20, 481, 1919 ; Ver. Deut.
Phys. Ges. 20, 181, 1919 ; and Horton and Davies, Proc. Roy. Soc.
95A, 408, 1919; Phil. Mag. 39, 592, 1920.
ANALYSIS OP THE ELEMENTS 69
58. The determination of the masses of atoms of
hydrogen and helium by the method of " Bracketing." 1
The determination of masses so far removed as these from
the ordinary reference lines offers peculiar difficulties, but,
as the lines were expected to approximate to the terms of the
geometrical progression 1, 2, 4, 8, etc., the higher terms of
which are known, a special method was adopted by which a
two to one relation could be tested with some exactness. Two
sets of accumulators were selected, each giving very nearly
the same potential of about 250 volts. The potentials were
then made exactly equal by means of a subsidiary cell and a
current -divider, the equality being tested to well within 1 in
1000 by means of a null instrument. If exposures are made
with such potentials applied to the electric plates first in
parallel and then in series, the magnetic field being kept
constant, all masses having an exact two to one relation will
be brought into coincidence on the plate. 2 Such coincidences
cannot be detected on the same spectrum photographically ;
but if we first add and then subtract a small potential from
one of the large potentials, two lines will be obtained which
closely bracket the third. To take an actual instance using
a gas containing hydrogen and helium, with a constant current
in the magnet of 0*2 ampere, three exposures were made
with electric fields of 250, 500 + 12, and 500 12 volts
respectively. The hydrogen molecule line was found sym-
metrically bracketed by a pair of atomic lines (Plate III,
Spectrum VII, a and c), showing within experimental error
that the mass of the molecule is exactly double the mass of
the atom. When after a suitable increase of the magnetic
field the same procedure was applied to the helium line and
that of the hydrogen molecule, the bracket was no longer
symmetrical (Spectrum VII, 6), nor was it when the hydrogen
molecule was bracketed by two helium lines (d). Both results
show in an unmistakable manner that the mass of He is less
than twice that of H 2 . In the same way He was compared with
0+ + , and H 3 . 3 The method is discussed on p. 57. The values
obtained by its use can be checked in the ordinary way by
1 Aston, Phil. Mag. 39, 621, 1920.
2 F. p. 57. 8 F. p. 70.
70
ISOTOPES
comparing He with C ++ and H 3 with He, these pairs being
close enough together for the purpose. The following table
gives the range of values obtained from the most reliable
plates :
Line.
Method.
Mass assumed.
Mass deduced.
TT/-
/Bracket
0++
= 8
3-994-3-996
ie
{Direct
C+ +
= 6
4-005-4-010
f Bracket
C+ +
= 6
3-025-3-027
3 . ...
(Direct
He
= 4
3-021-3-030
H 2 . . . .
Bracket
He
= 4
2-012-2-018
From these figures it is safe to conclude that hydrogen is a
simple element and that its atomic weight, determined with
such consistency and accuracy by chemical methods, is the true
mass of its atom.
This result leads to theoretical consideration of the greatest
importance, which will be discussed later. 1
59. Triatomic Hydrogen H 3 . The occurrence of a
parabola corresponding to a mass 3 was first observed and
investigated by Sir J. J. Thomson. 2 He came to the conclusion
that it was probably due to triatomic hydrogen. The simplest
way of obtaining this substance is to bombard KOH with
cathode rays and pump off the gases so produced. The H 3
used for the above measurements was obtained in this way.
The mass deduced proves in a conclusive manner that the
particle causing it is a molecule of three hydrogen atoms, a
result independently established about the same time by the
chemical work of Wendt and Landauer. 3
60. Krypton (At. Wt. 82-92) and Xenon (At. Wt. 130-2).
The results with these elements were particularly interesting.
The first source available, was the remains of two small samples
of gas from evaporated liquid air. Both contained nitrogen,
oxygen, argon, and krypton, but xenon was only detected
spectroscopically in one and its percentage in that must have
been quite minute. Krypton is characterised by a remarkable
1 V. p. loo.
* J. J. Thomson, Rays of Positive Electricity, p. 116, 1913.
3 Wendt and Landauer, Jour. Am. Chem. Soc. 43, 930, 1920.
ANALYSIS OF THE ELEMENTS 71
group of five strong lines at 80, 82, 83, 84, 86, and a faint sixth
at 78. This cluster of isotopes is beautifully reproduced with
the same relative values of intensity in the second, and fainter
still in the third order. These multiply-charged clusters give
most reliable values of mass, as the second order can be com-
pared with A (40) and the third with CO or N 2 (28) with the
highest accuracy. It will be noted that one member of each
group is obliterated by the reference line, but not the same
one. The singly and doubly charged krypton clusters can be
seen to the right and left of Spectrum VIII, Plate III. It will
be noticed that krypton is the first element examined which
shows unmistakable isotopes differing by one unit only.
On the krypton plates taken with the greatest magnetic
field faint, but unmistakable indications of lines in the
region of 130 could just be detected. The richest sample was
therefore fractionated over liquid air, and the last fraction, a
few cubic millimetres, was just sufficient to produce the
xenon lines in an unmistakable manner. Five could be dis-
tinguished, but owing to difficulties in the way of accurate
measurement the provisional values first published were one
unit too low.
Later on in March, 1921, a sample of gas was obtained which
contained a large proportion of xenon, though it was by no
means free from krypton. This yielded some excellent mass
spectra, which not only served to correct the figures given for
the five isotopes discovered previously, but also indicated
the possibility of two additional ones.
The absolute position of the group on the mass scale was
satisfactorily fixed by means of the second order line of the
strongest member, which fortunately lies outside the third
order mercury group. This gave constant and accurate
values corresponding to 64-5. The five strong lines of xenon
are therefore 129, 131, 132, 134, 136. On the left of the first
there was to be seen on many of the plates distinct indications
of a faint component 128. Also the darkening between the
lines 129 and 131 appears decidedly greater than that to be
expected from ordinary halation and suggests the possibility
of a seventh isotope 130. The relative intensity of the lines
of krypton and xenon is best indicated in Fig 17, p. 109.
72 ISOTOPES
61. Mercury (At. Wt. 200-6). As this element is em-
ployed both in the apparatus for the admission of gas and
in the Gaede vacuum pump, it would be very difficult to
eliminate it entirely from the discharge. This is fortunately
neither necessary nor desirable in most cases, for it provides
a valuable reference scale and, for some reason unknown,
its presence is definitely beneficial to the smooth running
of the discharge tube.
Mercury is abnormal in its capacity for forming multiply-
charged particles. A study of its remarkable parabolas l
enabled Sir J. J. Thompson to show that the atom of mercury
can carry no less than eight charges, that is lose eight electrons.
He gives reasons for considering that it loses all eight at once
and then recaptures them one at a time, so giving rise to a
series of parabolas 200/1, 200/2, 200/3, etc. The brightest is
the first, which is due to atoms which have recaptured all but
one electron ; the others are progressively fainter.
Subjected to the greater resolving power of the mass spectro-
graph it was seen at once that mercury was a complex element.
Its first, second, third, and higher order lines appeared as a
series of characteristic groups around positions corresponding
to masses 200, 100, 66f, etc. Some of these will be easily
distinguished on the spectra reproduced. The second, third
and fourth order groups are well shown in Spectrum VIII,
Plate IV. Careful study of the group shows that it consists
of a strong line 202, a weak one 204 and a strong group 197-200
which cannot be resolved on the present instrument, but
which in all probability contains all the four integers in that
range.
62. Boron (At. W. 10 90). Fluorine (At. W. 1900).
Silicon (At. W. 28-3). It will be convenient to treat of
these three elements together. The atomic weights of boron
and fluorine have both been recently redetermined by Smith
and Van Haagen 2 with the above results. On the atomic
weight of silicon there is some divergence of opinion. The
1 V. Plate I (3).
2 Smith and Van Haagen, Carnegie Inst. Washington Puhl. No. 267,
1918.
I
' 75
-r
ll
PLATE IV.
a
- t f:-/ 6
24- : ^..44 - -5-s **-5*
32-* * B - /J-6
; "" -245
34-* !: * 40-
^ -7
-W'5 44-*|-;
11 ^j 47-* *
* -/6 48- g
49- H *
^ I > x ** -32
44-^ * "^ *
2O
-w
0, 5 to
MASS SPECTRA.
/rind permission of tJie Editors of The Philosophical Magazine.)
ANALYSIS OF THE ELEMENTS 73
international value is quoted above, but Baxter, Weatherell,
and Holmes make it nearer 28 -I. 1
After a failure to obtain the boron lines with some very
impure boron hydride, a sample of boron trifluoride was pre-
pared from boric acid and potassium borofluoride, and this
gave good results. Following the usual practice, it was mixed
with a considerable quantity of C0 2 before introduction into
the discharge-tube. Very complex and interesting spectra
were at once obtained, and it was remarked that this gas
possessed an extraordinary power of resurrecting the spectra
of gases previously used in the apparatus. Thus the char-
acteristic, first and second order lines of krypton were plainly
visible, although the tube had been washed out and run many
times since that gas had been used. This property of liberating
gases which have been driven into the surface of the discharge-
bulb is doubtless due to the chemical action of the fluorine,
liberated during the discharge, on the silica anticathode and
the glass walls. After running some time the corrosion of
the anticathode was indeed quite visible as a white frost over
the hottest part.
After several successful series of spectra had been secured,
the percentage of boron trifluoride in the gas admitted was
increased as far as possible, until the discharge became quite
unmanageable and the tube ceased to work. Just before it
did, however, it yielded two very valuable spectra which
confirmed the isotopic nature of boron. These are reproduced
side by side as they were taken (Spectra I & II) Plate IV. The
lines at 10 and 11 are undoubtedly both first-order lines of
boron. The hypothesis that these might be due to neon
liberated by the action mentioned is not tenable, both on
account of their relative intensities and the absence of strong
neon first-order lines. Even if it were, it could not explain
the presence of the well-defined lines at 5 and 5-5 which had
never been obtained before at all, and which must be second-
order lines of boron. This element therefore has at least two
isotopes 10 and 11. The relative photographic intensity of
1 Baxter, Weatherell and Holmes, Journ. Am. Chem. Soc., 42, 1194,
1920.
74 ISOTOPES
the lines 5 and 5-5 does not agree well with an atomic weight
as high as 10-9, and the discrepancy might be explained by
the presence of a third isotope at 12 ; which would be masked
by carbon, for it has not yet been found practicable to eliminate
carbon from the discharge. But Plate IV, Spectrum IV,
contradicts this suggestion for, as will be shown later, the
line at 49 is mainly if not wholly due to B 11 F 2 , so that there
should also be a line at 50 for B 12 F 2 . The line at 49 is very
strong, but at 50 any small effect there may be can safely
be ascribed to the fourth order of mercury. The evidence
is clearly against the presence of a third isotope of boron.
The exceedingly accurate whole-number value for the
atomic weight of fluorine suggests the probability of this
element being simple. This conclusion is borne out by the
strong line at 19-00 with second-order line at 9-50. The
accompanying line at 20, very faint in Spectrum II, Plate
IV, is no doubt HF. As there is no evidence whatever to
the contrary, fluorine is taken to be a simple element with
an atomic weight 19.
Having adopted these values for boron and fluorine, we
may now apply them to Spectra III and IV, Plate IV, taken
with boron trifluoride. Consider first the group of three very
strong lines 47, 48, and 49. The last two are to be expected
as being due to B 10 F 2 and B^Fa respectively, but since there
is no evidence of a boron 9 or a fluorine 18, line 47 cannot
be due to a compound of these elements. But line 47 only
appeared when BF 3 was introduced, and so must be due to
silicon fluoride formed by the action of the fluorine on the
glass walls and the silica anticathode.
To test this the BF 3 was washed out and replaced by SiF 4 ,
which had been made by the action of sulphuric acid on cal-
cium fluoride and silica in the usual way. This greatly reduced
the lines 48 and 49, and so they must be attributed to boron
compounds. At the same time line 47 remained very strong,
and was evidently due to a compound Si 28 F, so that silicon
has a predominant constituent 28. This conclusion is further
supported by the presence of very strong lines at 66, Si 28 F 2
and 85, Si^Fg.
The chemical atomic weight shows that this cannot be its
ANALYSIS OF THE ELEMENTS 75
only constituent. Lines at 29, 48, 67, and 86 all suggest a
silicon of atomic weight 29. Practically conclusive proof of
this is given in Spectrum V, Plate IV, which shows its second-
order line unmistakably at 14-50. The only other reasonable
origin of this line, namely second-order B 10 F, is eliminated
by the fact that there is no trace of a line at 10 in this spectrum.
The evidence of a silicon of atomic weight 30 is of a much
more doubtful character. Its presence is suggested by the
lines 30, 49, 68, and 87, but the possibility of hydrogen com-
pounds makes this evidence somewhat untrustworthy, and
no proof can be drawn from a second-order line 15, as this is
normally present and is due to CH 3 . On the other hand, if
we accept a mean atomic weight as high as 28-3, the relative
intensity of the lines due to compounds of Si 28 and Si 29
indicates the probable presence of an isotope of higher mass.
These considerations taken with the complete absence of any
definite evidence to the contrary make the possibility of
Si 30 worth taking into account.
63. Molecular lines of the Second Order. The work of
Sir J. J. Thomson on multiply -charged positive rays showed
very definitely that molecules carrying more than one charge
were at least exceedingly rare, 1 for not a single case was
observed which could not be explained on other grounds.
Up to the time of the experiments with the fluorine compounds
the same could be said of the results with the mass-spectrograph.
This absence of multiply-charged molecular lines, though
there is no particular theoretical reason for it, has been used
as confirmatory evidence on the elementary nature of doubtful
lines.
The spectra obtained with BF 3 show lines for which there
appears no possibility of explanation except that of doubly-
charged compound molecules. The two most obvious of these
may be seen on Plate IV, Spectrum III, and at the extreme
left-hand end of Spectrum IV. They correspond to masses
23-50 and 24-50, the first being quite a strong line. Were
there no lines of lower order corresponding to these, the whole-
number rule might be in question ; but all doubt is removed
1 J. J. Thomson, Rays of Positive Electricity, p. 54.
76 ISOTOPES
by the fact that the lines 47 and 49 are two of the strongest
on the plate. A comparison of several spectra upon which
these lines occur shows a definite intensity relation which
practically confirms the conclusion that the first pair of lines
are true second-order lines corresponding to the first order
lines of the second pair. Now lines 47 and 49 cannot by
any reasonable argument be elementary, they must in fact
be due to compounds of fluorine with boron B n F 2 or silicon
Si^F, or due to both. Further evidence of the capability
of fluorine compounds to carry two charges is offered by line
33-50, which is undoubtedly the second-order line corresponding
to 67, i.e. B 10 F 3 or Si 29 F 2 . So far as results go, fluorine
appears to be unique in its power of yielding doubly-charged
molecules in sufficient number to produce second-order lines
of considerable strength.
64. Bromine (At. Wt. 79-92). The results with this
element were definite and easy to interpret. Its chemical
combining weight is known with great certainty, and is very
nearly the whole number 80. It was rather a surprise, there-
fore, that it should give a mass-spectrum which showed it to
consist of a mixture of two isotopes in practically equal pro-
portions. Methyl bromide was used for the experiments, and
one of the results is reproduced in Plate IV, Spectrum VI.
The characteristic group consists of four lines at 79, 80, 81,
and 82. 79 and 81, apparently of equal intensity, are much
the stronger pair, and are obviously due to elementary bro-
mines. This result is practically confirmed by second-order
lines at 39-5 and 40-5 too faint to reproduce, but easily seen
and measured on the original negative. The fainter pair, 80
and 82, are the expected lines of the two corresponding hydro-
bromic acids.
65. Sulphur (At. Wt. 32-06). Spectra VII and VIII,
Plate IV, show the effect of the addition of sulphur dioxide
to the gas in the discharge-tube. Above each is a comparison
spectrum taken immediately before the gas was admitted, on
the same plate with approximately the same fields. The very
marked strengthening of lines 32 and 44 is no doubt due to S
and CS. New lines appear at 33 SH, 34 SH 2 , 60 COS, 64 S0 2
ANALYSIS OF THE ELEMENTS 77
or S 2 , and 76 CS 2 . It may be noticed that lines 32, 60 and 76
are accompanied by a faint line one unit higher and a rather
stronger line two units higher. In the first case it is certain
and in the others probable that these are, at least partly, due
to hydrogen addition compounds. If a higher isotope of
sulphur exists, as is suggested by the chemical atomic weight,
it seems unlikely that this should have mass 33, for this would
have to be present to the amount of 6 per cent., and should
give a line at 35 one-thirteenth the strength of 34 (normal
SH 2 ). No such line is visible. A sulphur of atomic weight 34
present to the extent of 3 per cent, is more likely, but there
is hardly enough evidence as yet to warrant its serious con-
sideration.
66. Phosphorus (At. Wt. 31-04). Arsenic (At. Wt.
74-96). The gases phosphine PH 3 and arsine AsH 3 were used
in the experiments on these elements, and the results were of
notable similarity. The mass-spectrum of each gas was
characterised by a group of four lines. The first and strongest
doubtless due to the element itself, the second rather weaker
due to the monohydride, the third very faint to the dihydride,
and the fourth fairly strong to the trihydride. The spectrum
of AsH 3 is shown in Spectrum IX, Plate IV ; that of phos-
phorus is similar but its lines are weak, and therefore unsuited
to reproduction. Both elements appear to have no isotopes,
and neither give visible second-order lines.
67. Selenium (At. Wt. 79-2). Tellurium (At. Wt.
127-5). The compounds used in the experiments on these
elements were selenium hydride, made by passing a stream of
hydrogen through boiling selenium, and tellurium methyl.
Complete failure resulted in both cases. There was, indeed,
on one spectrum an exceedingly faint line at 79, but no shred
of reliable evidence could be found to ascribe it to an isotope
of Se. In the case of tellurium no trace of any line near 127
could be discovered. The failure is unfortunate in the case
of Te on account of its well-known anomalous position in the
periodic table ; in the case of Se particularly so for the follow-
ing reasons : If the accepted atomic weight is even approxi-
mately correct this element must have one isotope, at least,
78 ISOTOPES
of atomic weight greater than 78. But the numbers 79, 80,
81, 82, 83, 84, are already filled by isotopes of Br and Kr,
so that it is extremely probable that one of the isotopes of Se
has an atomic weight identical with one of an element having
a different atomic number, i.e. is an Isobare. The latter are
known to exist among radioactive elements, but none have so
far been discovered during the work on mass spectra.
68. Iodine (At. Wt. 126-92). The results with this
element were fortunately both definite and conclusive. Methyl
iodide was employed, its vapour being introduced mixed with
CO 2 and CH 4 . It gave one strong line at 127 satisfactorily
confirmed by another single line at 142 due to CH 3 I.
This proves iodine to be a simple element in an unequivocal
manner, a rather unexpected result since all the speculative
theories of element evolution, by Van den Broek and others,
predict a complex iodine.
69. Antimony (At. Wt. 120-2). Antimony hydride SbH 8
was used. This was made by dissolving antimony magnesium
alloy in dilute acid. Unlike the corresponding arsenic com-
pound it gave an entirely negative result, no line whatever
being distinguishable in the region expected from the atomic
weight. This failure is probably to be ascribed to the exceed-
ingly unstable nature of the antimony compound.
70. Tin (At. Wt. 118-7). Tin tetrachloride was employed
in the investigation of this element. The vapour of this
compound attacks the tap grease used in the apparatus, which
makes it extremely difficult to deal with. The results were
entirely negative except in one case. On this occasion a
second attempt to get the selenium line from selenium hydride
was actually in progress, but a good deal of SnCl 4 vapour had
been introduced previously, and the chlorine lines were so
intense that some " resurrected " compound of chlorine must
have been the principal factor in the discharge. For some
unknown reason the discharge tube was working abnormally
well. On one of the spectra then obtained, Spectrum II, a
group of lines of even integral mass 116, 118, 120, 122, 124
(followed by iodine 127) could be distinguished and some of
ANALYSIS OF THE ELEMENTS 79
these may possibly have been due to isotopes of tin. This
supposition is slightly strengthened by the appearance of a
still fainter group of odd integral mass containing the lines
155, 157, etc., which might be isotopic tin monochlorides. It
has not been found possible to repeat this result, so that no
reliance is to be put upon it.
71. Nickel (At. Wt. 58-68). Nickel received attention
early in the history of positive rays as it is one of the elements
whose atomic weight is out of order in the periodic table ;
it should be heavier, not lighter than cobalt (58-97). It is
amenable to treatment in the ordinary discharge tube for it
forms an easily vaporisable carbonyl compound Ni(CO) 4 .
Unfortunately this is very rapidly decomposed by the electric
discharge, so that in the early experiments made by Sir J. J.
Thomson the walls of the discharge bulb became coated with
a black deposit of the metal, it was impossible to maintain a
steady discharge for a sufficient time, and no satisfactory
parabola corresponding to the element could be obtained.
Quite recently l by the use of abnormally high current
intensities in the discharge it has been found possible to over-
come these difficulties to some extent and to obtain a satis-
factory mass spectrum from a mixture of nickel carbonyl and
carbon dioxide. This consists of two lines, the stronger at 58
and the weaker at 60. They are most conveniently placed
between the mercury groups of the third and fourth order,
with which they can be compared with an accuracy of one-
tenth per cent. The results were also checked by comparison
with the CO 2 line at 44, and appear to be integral within the
above error. Nickel therefore consists of at least two isotopes.
The intensities of the lines are about in the ratio 2:1, and
this agrees with the accepted atomic weight. It may be
noticed that had the heavier isotope preponderated the atomic
weight of the element would have appeared normally placed
in the periodic table.
1 Nature, June 23, 1921, p. 520.
CHAPTER VII
ANALYSIS OF THE ELEMENTS (Continued)
72. Positive Rays of Metallic Elements. Positive rays
of most of the metallic elements cannot be obtained by the
ordinary discharge- tube method, since in general they have
extremely low vapour-pressures and are incapable of forming
stable volatile compounds. Mercury is a notable exception to
this rule, and its rays are exceedingly easy to produce.
Positively charged rays which appeared to be atoms of the
alkali metals were first observed by Gehrcke and Reichen-
heim. 1 They obtained them by two distinct methods : the
first, which may be conveniently called the " Hot Anode "
method, consisted in using as anode of the discharge -tube a
platinum strip coated with a salt of the metal and electrically
heated by an external battery. The second device, with which
they performed most of their pioneer work on Anode Rays,
was to use a composite anode of special construction which
worked without the need of external heating.
73. Dempster's analysis of Magnesium (At. Wt.
24-32). The experiments of Dempster with the " hot anode "
method of generating positive rays have already been noted. 2
Later, 3 he announced the very important discovery of the
three isotopes of magnesium, and subsequently published an
account of the experimental details. 4 The magnesium rays
were obtained from a piece of the metal which was heated
electrically by a coil of wire, and at the same time bombarded
by electrons from a Wehnelt cathode. The occluded gases
1 Gehrcke and Reichenheim, Ver. d. Phys. Gesett., 8, 559, 1906;
9, 76, 200, 376, 1907 ; 10, 217, 1908.
2 P. 31. 3 Dempster, Science, Dec. 10, 1920.
4 Dempster, Proc. Nat. Ac. Sci., 7, 45, 1921.
80
ANALYSIS OF THE ELEMENTS 81
were first driven off, and then the heating current was increased
till the metal was slightly vaporised and the magnesium lines
appeared. The following description of the analysis and the
curves obtained are taken direct from Dempster's paper :
22
\J
22 23 24 ^ 25 Z6 27
Atomic Weight.
FIG. 13. Curve for Magnesium.
28 29
27
24 25 26
Atomic Weight,
FIG. 14. Curve for Magnesium.
28
29
82 ISOTOPES
" The charged atoms of different atomic weights are succes-
sively brought on to the detecting electrode by keeping the
magnetic field constant and varying the potential which accele-
rates the rays, the potential required being inversely propor-
tional to the mass of the particles. Thus, if one atomic weight
is known the others may be found. Due to the finite width
of the slits, each element gives a curve, on the atomic weight
scale, which is theoretically a linear increase to a maximum
and then a linear decrease. The width half way to the maxi-
28
mum is given by m. - where ra is the atomic weight, S the
d
slit width and d the diameter of the circle in which the rays
travel. Under good vacuum conditions this theoretical
sharpness is practically obtained. For 1 mm. slits this width
of the curves should thus be one-half a unit on the atomic
weight scale. The former measurement with the apparatus
and the magnetic field determinations sufficed to locate elements
between 20 and 30 within one unit, and identified the strong
nitrogen rays (possible carbon monoxide) of molecular weight
28 which are given off when the metal is first heated.
" One series of experiments was as follows. After heating
the magnesium slightly and pumping, till a MacLeod gauge
gave no pressure indication, the nitrogen molecule was the
only particle present. The heating current was then increased
by steps to vaporise the magnesium. With 0*7 ampere, 28
alone was present, with 0'75 ampere an arc apparently struck
as the cathode-anode current jumped suddenly to five times
its value. The electron current was decreased to its former
value by cooling the cathode and the rays were measured.
It was found that three strong new lines had appeared. The
new lines which are undoubtedly due to magnesium were com-
pared with the nitrogen rays which were still faintly present
and found to have atomic weights, 24, 25 and 26. The obser-
vations are illustrated in Fig. 13, which gives the current or
number of particles for different atomic weights. The nitrogen
line had its maximum at 817 volts, and the atomic weight
abscissae are 28 x 817 divided by the volts applied. The
ordinates of the 28 line are multiplied by 10 in plotting to
make them comparable with the other three lines. The
ANALYSIS OF THE ELEMENTS 83
dotted continuation to the axis indicates the slight overlapping
of the lines. We conclude that magnesium consists of three
isotopes of atomic weights 24, 25 and 26.
" Later curves made with steadier discharge conditions are
more suitable than Fig. 13 for measuring the relative strengths
of the components. In Fig. 13 there appears to have been a
drop in intensity just before 24 was reached, in the measure-
ment from high to low atomic weights. The curve is of interest
as still containing 28 faintly and so serving accurately to
locate the weights which otherwise would have been uncertain
to a fraction of a unit.
" Fig. 14 is one of several later curves taken under steadier
conditions. These all have very closely the same appearance.
The components 25 and 26 are present very nearly in equal
amounts ; in some measurements 25 was found about nine-
tenths the intensity of 26. The component at 24 is approxi-
mately 6 times as strong as the one at 26. The ratio of 1:1:6
gives an average atomic weight 24-375, which is in as good
agreement with the accepted atomic weight for magnesium
as could be expected with the wide slits used in these first
experiments."
74. The mass-spectra of the alkali metals. In order
to analyse the metals of this group a modification of Gehrcke
and Reichenheim's hot anode method was employed by the
writer to generate the positive rays. After a certain amount
of initial difficulty in technique had been overcome this gave
satisfactory results. *
The apparatus for producing the rays was very simple, and will
be readily understood from the figure (Fig. 15). The hot
anode A is a strip of platinum foil *03 mm. thick, about 2 mm.
wide by 7 mm. long, welded to the two stout platinum leads
which are fused through the glass at C. It was raised to the
required temperature by current from one large storage-cell
connected through a rheostat as shown. As the anode is of
necessity the high-potential pole of a discharge-tube arranged
to give positive rays, this heating arrangement had to be very
carefully insulated. The anode was mounted on a ground
1 Aston, Phil. Mag., 42, 436, 1921,
84 ISOTOPES
joint as indicated so that it could be easily removed and
replaced. The discharge -tube was cylindrical, about 4 cm.
in diameter, mounted concentric to the axis of the perforated
cathode K. A side tube was fitted at B which could be cooled
in liquid air ; in some of the experiments this was charged
with charcoal.
1)
JT ^ -. H. ^
^x MHWW\
'#'
FIG. 15. Hot Anode Discharge Tube.
The anode was placed immediately opposite the perfora-
tion of the cathode and about 1 cm. away from it. The
platinum strip was bent at one end into a U-shaped channel
into which the salts could be melted. The discharge was
maintained by a large induction-coil used in the previous
work on mass spectra and rectified by means of a valve V.
75. Experiments with the Parabola method of
analysis. In the preliminary experiments the analysis of
the rays was performed by Sir J. J. Thomson's " parabola "
method, since this gives the maximum general information,
and it was only when suitable conditions and technique had
been ascertained that the mass spectrograph was applied.
The general procedure was to pump out the discharge-tube
to the lowest possible pressure, far lower than that necessary
to prevent all discharge with the anode cold, and then to
heat up the anode until the discharge started. This usually
happened at dull red heat, and by very careful adjustment of
the temperature and of the primary current in the coil it was
possible, under favourable conditions, to maintain a fairly
steady current of 1 to 2 miUiamperes at a potential of about
20,000 volts.
It will be seen that the arrangement resembles that of a
Coolidge X-ray tube reversed pole for pole, and it was hoped
that it might share the outstanding controllability of that
device ; but that expectation was only very partially realised.
ANALYSIS OF THE ELEMENTS 85
The mechanism of the discharge is extremely obscure, for the
current intensity is, of course, enormously in excess of that
to be expected from the ordinary thermionic release of positive
ions from the hot anode. 1 There was very little visible glow
in the tube, the X-radiation was small and, although a faint
cloud of sodium light nearly always appeared in front of the
red-hot anode, the pressure was too low for the anode rays
to be visible ; their point of impact with the cathode could,
however, be inferred from the scintillations on its surface.
Observations of this effect lead to the conjecture that the
bulk of the rays originate not from the surface of the salt
itself but from that of the heated platinum, and also that some
points on this are much more active than others, giving rise
to jets of rays. The direction of these jets seemed to depend
on the local configuration of the strip and was beyond prac-
tical control. The obvious device of moving the anode about
by means of the ground joint to get a radiant point in the
required place could not be applied, for the parabolas were
never bright enough to be visible on the willemite screen.
To add to these difficulties the salt disappeared very rapidly,
in some cases in a few minutes. Consequently exposures
were very limited in duration, and even in the most favourable
cases the results rarely had a satisfactory intensity.
The preliminary experiments were done with sodium phos-
phate, and before long encouraging results were obtained.
In all the successful exposures only a single parabola appeared,
and this showed that although the method on account of
the number of inevitable failures is an exasperating one to
use as a means of identifying isotopes it has the great merit
of producing the positive rays of the metals and no others.
This characteristic seems to be due to the very low pressure
employed and also possibly to the position of the anode itself,
which prevents any positive rays generated in more distant
parts of the tube from ever reaching the perforation in the
cathode in the necessary axial direction.
Such a selective action has two very important results.
In the first place, it eliminates the many ambiguities of the
1 Richardson, The Emission of Electricity from Hot Bodies, p. 234 et
seq., Longmans, 1916.
86 ISOTOPES
ordinary mass spectrum due to multiply-charged rays, or to
hydrogen and other addition products ; but, in the second, it
prevents the use of the oxygen line as a comparison standard.
As soon as it was demonstrated beyond any reasonable doubt
that sodium was a simple element (and its chemical atomic
weight is se exactly integral on the oxygen scale as to be
conclusive corroboration) it was taken as standard at 23.
76. Lithium (At. Wt. 6-94). The most successful experi-
ment done with the parabola method of analysis was one in
which a mixture of sodium and lithium phosphates was em-
ployed (this contained traces of potassium salts). By great
good fortune a very strong jet of rays must have been directed
along the axis and three satisfactory exposures were obtained
before the anode dried up. One of these is reproduced in
Plate I (5) A strong parabola at 7 and a weak one at 6 demon-
strate clearly that lithium is a complex element, as its chemical
atomic weight 6-94 leads one to expect. This result, which
was announced by the writer and G. P. Thomson in Nature,
February 24th, was confirmed independently by Dempster l
using the method described for magnesium. The several
photographs here considered all gave approximately the same
ratio of intensities, and they corresponded as well as was to
be expected with the accepted atomic weight. On the other
hand, G. P. Thomson's parabolas (which were obtained with
a composite anode) and Dempster's electrical measurements
suggest a more nearly equal intensity ratio and this ratio
appears to vary.
77. Sodium (At. Wt. 23-00). Sodium gave the brightest
effects, and its single line was obtained so intense that the
presence of another constituent to the extent of even less than
1 per cent, could probably have been detected. It may there-
fore be safely regarded as a simple element.
The parabola method of analysis is perfectly satisfactory
in the case of so light an element as lithium, but cannot be
used for the critical examination of the heavier members
of the group ; and so the apparatus for the production of the
1 Dempster, Science, April 15, 1921.
ANALYSIS OF THE ELEMENTS 87
rays was fitted to the mass spectrograph already described. 1
The experimental difficulties became now very serious indeed,
for, in addition to those already indicated, there was no means
of finding the most suitable voltage to apply to the electro-
static plates. In normal cases this is done by visual inspection
of the hydrogen lines, but here it could only be guessed at.
Under these conditions it is not a matter for surprise that
the photographs, though sufficient for the purpose of detecting
isotopes, only gave very faint lines and so cannot be reproduced
as illustrations.
78. Potassium (At. Wt. 39-10). A mixture of potassium
sulphate, potassium bromide, and a little sodium phosphate
was now used on the anode, and after several unsuccessful
attempts some fairly satisfactory spectra were obtained which
contained both sodium and potassium lines. Using the
former as standard the latter consisted of a bright component
at 39, and a very faint component at 41.
79. Rubidium (At. Wt. 85-45). Rubidium chloride was
now added to a little of the mixture used in the potassium
experiments and spectra containing the potassium and rubi-
dium lines were obtained. Rubidium is very definitely double.
Its components are more nearly equal in intensity than those
of lithium or potassium. Measured against the potassium
line 39 its stronger component is 85 and the weaker 87. The
intensity ratio agrees reasonably well with the accepted
atomic weight 85-45.
80. Caesium (At. Wt. 132-81). When a mixture of
rubidium chloride and caesium chloride was used evidence of
a line at 133, measured against the two rubidium lines, was
soon obtained. Pure caesium chloride was then substituted
and the utmost possible exposure given to search for a lighter
component, which was to be expected from the fractional
chemical atomic weight 132-81. Although by this means the
intensity of the line 133 was increased to a satisfactory pitch
no other neighbouring line was found. If, therefore, a lighter
isotope of caesium exists it must differ from 133 by many
1 V. Chap. V.
88 ISOTOPES
units which seems very unlikely or it cannot be present in
proportion sufficient to account for the fractional atomic
weight obtained by chemical means.
81. Thompson's work on Beryllium (At. Wt. 9-1).
G. P. Thomson l has recently investigated the Anode rays
obtained from a composite anode similar to that devised by
Gehrcke and Reichenheim 2 and has subjected them to analysis
by the parabola method. After the parabolas of the isotopes
of lithium had been successfully obtained 3 he went on to
investigate the element beryllium. The best results were
obtained from a mixture of sodium bromide and beryllium
fluoride. This gave a single strong parabola corresponding
to an atomic weight 9 (Na = 23). The accepted chemical
atomic weight is rather higher, so a careful examination was
made to discern any possible faint companions at 10 or 11.
He concludes that neither of these can be present to any sen-
sible extent, and therefore that beryllium is probably a simple
element.
82. Calcium (At. Wt. 40-07) and Strontium (At. Wt.
87-63). Thomson also obtained by the same method para-
bolas due to these elements, the latter very faint, but the
resolution at his disposal was too low to decide their constitution.
From the position of the strong parabola of calcium he con-
cludes that one or more of the atomic weights 39, 40, 41 were
present ; and as all these are already known to exist as isotopes
of either potassium or argon, it follows that calcium must be
an isobare of one or other of these elements. 4
*
83. Table of Elements and Isotopes. The following list
tabulates the results contained in this and the previous Chapter.
The isotopes of complex elements are given in the order of the
proportions present. Brackets indicate that the figures are
provisional only.
1 G. P. Thomson, Phil. Mag., 42, 857, 1921.
2 V. p. 80.
3 V. p. 86. -
4 V. p. 148.
ANALYSIS OF THE ELEMENTS
Table of Elements and Isotopes
89
Element.
Atomic
number.
Atomic
weight.
Minimum
number of
isotopes.
Masses of isotopes in order of
intensity.
H . .
1
1-008
1
1-008
He . .
2
4-00
1
4
Li . .
3
6-94
2
7, 6
Be . .
4
9-1
1
9
B . .
5
10-9
2
11, 10
C . .
6
12-00
1
12
N . .
7
14-01
1
14
. .
8
16-00
1
16
F . .
9
19-00
1
19
Ne . .
10
20-20
2
20, 22, (21)
Na . .
11
23-00
1
23
Mg . .
12
24-32
3
24, 25, 26
Si . .
14
28-3
2
28, 29, (30)
P . .
15
31-04
1
31
S. . .
16
32-06
1
32
Cl . .
17
35-46
2
35, 37, (39)
A . .
18
39-88
2
40, 36
K . .
19
39-10
2
39, 41
Ni . .
28
58-68
2
58/60
As . .
33
74-96
1
75
Br . .
35
79-92
2
79, 81
Kr . .
36
82-92
6
84, 86, 82, 83, 80, 78
Rb . .
37
85-45
2
85, 87
I . .
53
126-92
1
127
X . .
54
130-2
5, (7
129, 132, 131, 134, 136,
(128, 130 ?)
Cs . .
55
132-81
1
133
Hg . .
80
200-6
(6)
(197-200), 202, 204
Ca
Zn
Dempster's later results (V. p. 148)
20 40-07 (2) (40, 44 ?)
30 65-37 (4) (64, 66, 68, 70)
CHAPTER VIII
THE ELECTRICAL THEORY OF MATTER
84. The Whole Number rule. By far the most important
result of the measurements detailed in the foregoing chapters is
that, with the exception of hydrogen, the weights of the atoms
of all the elements measured, and therefore almost certainly
of all elements, are whole numbers to the accuracy of experi-
ment, in most cases about one part in a thousand. Of course,
the error expressed in fractions of a unit increases with the
weight measured, but with the lighter elements the divergence
from the whole number rule is extremely small.
This enables the most sweeping simplifications to be made in
our ideas of mass, and removes the only serious objection to a
unitary theory of matter.
85. The Unitary Theory of the constitution of matter.
From the very earliest times it has been a favourite hypothesis
that all matter is really composed of one primordial substance,
Air, Fire, Earth and Water have all been suggested in the past.
The first definite theory of the constitution of the atoms of
the elements out of atoms of a primordial element (Protyle,
Urstoff, etc.) was made by Prout in 1815. Front's Hypothesis
was that the atoms of the elements were different aggregations
of atoms of hydrogen. On this view it is obvious that the atomic
weights should all be expressed by whole numbers when the
atomic weight of hydrogen itself is taken as unity. Owing to
the roughness of the methods available and the considerable
inaccuracies of the atomic weight determinations made at that
time there was little to disprove the hypothesis, and its marked
simplicity gained it many adherents. But as time went on
chemical methods grew more precise and it became more and
more impossible to reconcile experimental results with integral
90
THE ELECTRICAL THEORY OF MATTER 91
combining weights until the evidence against it was strong
enough to cause J. S. Stas (1860-1865) to state: " I have
arrived at the absolute conviction, the complete certainty, so
far as it is possible for a human being to attain to certainty in
such matters, that the law of Prout is nothing but an illusion, a
mere speculation definitely contradicted by experience."
Nevertheless, though abandoned temporarily by the chemist
as impracticable, the idea of primordial atoms appealed strongly
to the mind of the philosopher and the physicist. Herbert
Spencer, in his hypothesis of the constitution of matter, says :
" All material substances are divisible into so-called elementary
substances composed of molecular particles of the same nature
as themselves ; but these molecular particles are complicated
structures consisting of congregations of truly elementary
atoms, identical in nature and differing only in position,
arrangement, motion, etc, and the molecules or chemical atoms
are produced from the true or physical atoms by processes of
evolution under conditions which chemistry has not been able
to reproduce."
The discovery of the electron, the proof that it was the same
whatever the atom from which it was detached and, most
important of all, the demonstration by Sir J. J. Thomson and
others that electricity could simulate the known properties
of matter, gave us the key to the riddle of what these primordial
atoms really are. The only serious obstacle, the fractional
atomic weights, has now been removed so that there is nothing
to prevent us accepting the simple and fundamental conclusion :
The atoms of the elements are aggregations of atoms of
positive and negative electricity.
86. The Atom of Negative Electricity, or Electron.
The fundamental unit of negative electricity makes its appear-
ance in physical phenomena in many guises, such as the cathode
ray of electrical discharge, the beta ray of radioactive change,
the thermion of the wireless valve. A very complete account of
it has recently been published by Millikan 1 so that it is not
proposed to describe its history and properties at any length
1 The Electron, by R. A. Millikan, University of Chicago Press, 1917.
92 ISOTOPES
here. It will be sufficient to note a few of its more important
constants.
Its charge e is given by Millikan as 4-774 x 10 ~ 10 E.S.U.
The most reliable measurements of e/m for the electron, at low
velocity, give the value 5-30 x 10 17 E.S.U. Hence its mass is
almost exactly 9'00 x 10" 28 grs., 1845 times less than the mass
of the hydrogen atom, or 0-00054 on the ordinary scale of atomic
weights (Oxygen =16).
87. The atom of Positive Electricity, or Proton. Our
physical knowledge of this body is not nearly so complete as
that of its counterpart the electron. It is very significant that
in no analysis of positive rays so far performed have we been
able to discover a particle of mass less than that of the hydrogen
atom. This direct result, supported as it is by many less direct
lines of evidence, leads logically to the conclusion that the
hydrogen positive ray, i.e. the positively charged part remain-
ing when an electron is detached from a neutral hydrogen atom,
is the atom of positive electricity itself. The name " proton " l
was suggested for it by Sir Ernest Rutherford at the Cardiff
meeting of the British Association in 1920. The charge on a
proton is, of course, equal and of opposite sign to that on the
electron. Its mass in the free state has been measured directly 2
and is practically identical with that of the neutral atom of
hydrogen 1-66 x 10~ 24 grs., or 1-007 on the oxygen scale. 3
88. The Nucleus Atom. Certain experimental results,
notably the scattering of alpha rays, led Sir Ernest Rutherford
in 19 II 4 to formulate an atom model which has resulted in the
most remarkable advances in both physics and chemistry, and
is now almost universally accepted as correct in fundamental
principle. This is that an atom of matter consists of a central
massive nucleus carrying a positive charge which is surrounded,
at distances relatively great compared with its diameter, by
" planetary " electrons. The central nucleus contains all the
positive electricity in the atom, and therefore practically all its
mass. The weight of the atom and its radioactive properties
are associated with the nucleus ; its chemical properties and
1 From Greek irpurov first the primary substance.
2 P. 67. 8 V. p. 105.
Rutherford, Phil. Mag. 41, C69, 1911.
THE ELECTRICAL THEORY OF MATTER 93
spectrum, on the other hand, are properties of its planetary
electrons. It is clear that in a neutral atom the positive charge
on the nucleus must be equal to the sum of the negative charges
on the planetary electrons.
89. Moseley 's Atomic Numbers. The scattering experi-
ments mentioned above indicated that the net positive charge
on the nucleus (expressed in terms of the natural unit e) was
roughly equal to half the atomic weight. Now if we arrange the
elements in order of atomic weight, starting with hydrogen, each
element will have a position the number of which will be about
half its atomic weight. It was suggested by Van den Broek
that this atomic number might be equal to the charge on the
nucleus. Two years after the formulation of the nucleus atom
theory Moseley undertook an investigation of the changes which
took place in the wave-length of the X-rays given off when
various elements were used in turn as anticathodes. The result
of this piece of work, now classical, 1 was the establishment of the
most important generalisation in the history of chemistry since
Mendeleef 's Periodic Law. Discussing a quantity Q related to
wave-length Moseley writes :
" It is at once evident that Q increases by a constant amount
as we pass from one element to the next, using the chemical
order of the elements in the periodic system. Except in the
case of Nickel and Cobalt, 2 this is also the order of the atomic
weights. While, however, Q increases uniformly the atomic
weights vary in an apparently arbitrary manner, so that an
exception in their order does not come as a surprise. We have
here a proof that there is in the atom a fundamental quantity,
which increases by regular steps as we pass from one element to
the next. This quantity can only be the charge on the central
positive nucleus, of the existence of which we already have
definite proof. Rutherford has shown, from the magnitude of
the scattering of a particles by matter, that the nucleus carries a
positive charge approximately equal to that of A/2 electrons
when A is the atomic weight. Barkla, from the scattering of
X-rays by matter, has shown that the number of electrons in an
1 Moseley, Phil. Mag., 26, 1031, 1913.
2 Cf. Barkla, Phil. Mag., 14, 408, 1907.
94 ISOTOPES
atom is roughly A/2, which for an electrically neutral atom is
the same thing. Now the atomic weights increase on the
average by about 2 units at a time, and strongly suggest the
view that N increases from atom to atom always by a single
electronic unit. We are therefore led by experiment to the
view that N is the same as the number of the place occupied by
the element in the periodic system. This atomic number is
then for H 1, for He 2, for Li 3 for Ca 20 for Zn 30, etc.
This theory was originated by Broek l and since used by Bohr.
We can confidently predict that in the few cases in which the
order of the atomic weights A clashes with the chemical order of
the periodic system the chemical properties are governed by N,
while A itself is probably some complicated function of N."
Subsequent work has supported in an unquestionable manner
the ideas so expressed by Moseley. That the number of the
element in the order of the periodic table is actually the same as
the positive charge on the nuclei of its atoms, expressed of
course in terms of the natural unit of electric charge e, has been
proved by direct experiment for some of the heavier elements.
The recent work of Chad wick 2 leaves little room for doubt on
that point. At the other end of the scale all the known pro-
perties of hydrogen point to the conclusion that its atomic
number is 1 ; its exceptional atomic weight, as will be seen later,
is itself strong corroborative evidence of this.
A complete table of the elements with their Atomic Numbers,
Atomic Weights and isotopes (where these are known) is given
on page 142.
From Moseley's law of atomic numbers the explanation of the
empirical rule of radioactive transformation given on page 11
follows at once. An alpha particle carries two positive charges,
a beta particle one negative one. If therefore the atom of
a radioactive substance emits one alpha particle from its
nucleus it naturally descends two units in atomic number, that
is moves back two places in the periodic table. If on the other
hand it emits one beta particle it clearly moves forward one
place, for by the operation the nucleus has acquired one
additional charge.
1 Van den Brock, Phys. Zeit. 14, 33, 1913.
2 Chadwick, Phil Mag., 40, 734, 1920,
THE ELECTRICAL THEORY OF MATTER 95
90. The Bohr Atom. In this atom model the electrons
outside the nucleus are supposed to be in a state of continual
revolution about it, like planets round the sun. This rotation
is considered to take place in orbits defined in a very special
manner by means of a " quantum relation." This postulates
that when, and only when, an electron changes its orbit,
radiation is given out and the energy acquired by the change of
orbit is entirely given off as radiation of frequency v where the
change of energy equals hv where h is Planck's quantum or
element of action (6-55 x 10~ 22 C.G.S. ). l This theory lends itself
to exact mathematical analysis but unfortunately it can only be
worked out adequately for the two simplest cases, the neutral
hydrogen atom and the singly charged helium atom. Here*
however, its success is most remarkable; for not only is it possible
to calculate by its means the wave length of the chief series lines
of the hydrogen spectrum, to an accuracy almost unprecedented
in physics, but, by applying the relativity correction for change
of mass with velocity to the rotating electron, the fine structure
of the lines and the effects of electric and magnetic fields have
been predicted with the most astonishing exactness by Sommer-
f eld, Epstein and others. 2 Bohr has recently expressed the hope
of extending his theory to heavier atoms by means of a new
device which he terms the principle of " correspondence." 3
91. The Lewis Langmuir Atom. This form of atom
model was primarily designed to afford some theoretical basis
for the numerous general qualitative properties of elements and
their compounds. In it the electrons outside the nucleus are
supposed to be at rest at, or vibrating about, definite points.
The first two electrons will form a pair, the next eight will tend
to set in positions corresponding to the eight corners of a cube,
or some other solid figure, and so on. In this way we shall get a
series of shells or sheaths one outside the other. Langmuir has
recently 4 reduced his postulates to the following three :
(1) The electrons in atoms tend to surround the nucleus in
, Phil Mag. 36, 1, 476, 857, 1913.
2 Sommerfeld, Atombau and Spektrallinien, Brunschweig, 1921.
3 Bohr, Nature, 107, 104, 1921.
4 Langmuir, Brit. Assoc. Edinburgh meeting, 1921,
96 ISOTOPES
successive layers containing 2, 8, 8, 18, 18, 32 electrons re-
spectively.
(2) The atoms may be coupled together by one or more
" duplets " held in common by the complete sheaths of the
atoms.
(3) The residual charge on the atom and on each group of
atoms tends to a minimum.
This atom model is not amenable to mathematical treatment,
but it has been exceedingly successful in accounting for the
general chemical qualitative properties of many of the elements
and in predicting those of their compounds.
92. Diagrammatical representation of atoms of
Isotopes and Isobares. The accompanying diagrams (Fig.
16) are intended to indicate the sort of arrangements which
may take place in atoms. The small dark circle is the nucleus,
the number of protons and electrons comprising it being indi-
cated by the numerals. The electrons outside the nucleus are
indicated by small light circles.
(1) is an atom of atomic weight 6. Its nucleus contains 6
protons and 3 electrons, hence its atomic number is 3. It is in
fact the atom of the lighter isotope of lithium of atomic weight
6, Li 6 . To be electrically neutral it must have 3 electrons out-
side the nucleus. Now the principles underlying Langmuir 's first
postulate are derived from the Periodic Table and are certainly
correct. Langmuir explains this by saying that the first two of
these electrons will form an innermost ring or shell of two-
This shell being now complete, any more electrons will go out-
side and start the next shell of eight, so we indicate this by
putting the third electron in a circle of greater diameter.
Now suppose we add one electron and one proton to this atom.
If both enter the nucleus we shall get the configuration repre-
sented by (2) The nuclear charge is unaltered, so that the
arrangement of the exterior electrons will be precisely the same.
It follows that all properties depending on these electrons such
as atomic volume, spectrum, chemical properties, etc., will be
quite unaltered. But the weight of this atom is now 7, so it
is an isotope of lithium ; it is actually the atom of the heavier
constituent Li 7 . (1) AND (2) ARE ISOTOPES,
THE ELECTRICAL THEORY OF MATTER 97
But now suppose we add a proton and an electron to (1) so
that the proton only enters the nucleus and the electron remains
outside as shown at (3) We shall now have a charge 4 on the
nucleus and two electrons in the outer ring. The chemical
(V) Atom of Li 6
(2) Atom of Li 7
(3) Atom of hypothetical isotope of Beryllium
(4-) Atom of F
(S) Atom of Ne 20
f6) Atom of We 22
(7) Atom of Net O = Electron
FIG. 16. Diagrammatic Representation of Nucleus Atoms. The planetary
electrons are shown as lying on plane circles, the first containing 2, the
second 8 and so on. The dark circle is the nucleus and the rf- and
charges within it are indicated by figures. (1) and (2) are Isotopes. (2)
and (3) are Isobares and (5) and (6) are Isotopes.
properties of such an atom, if it could exist, would be completely
different from those of lithium, but would be identical with those
of beryllium, of which it would be an isotope. But its mass is
clearly identically the same as that of (2) so that (2) AND (3)
AEE ISOBABES.
H
98 ISOTOPES
In the same way (4) will be recognised as the atom of fluorine
(5) and (6) as the atoms of the two isotopes of neon and (7) as
the atom of sodium.
93. The relation between Isotopes and Elements in the
same Group. As far as can be seen the chemical properties by
which the elements are divided into groups depend practically
entirely on the outermost shell of electrons, which are therefore
called valency electrons. Now consider all that part called by
Langmuir the kernel of the atom lying within the shell of these
valency electrons. The movements or configuration of the
outermost electrons will depend in the first degree on the charge
on the kernel, which may be looked upon as a virtual nucleus.
The kernels of atoms (1) and (7) both have the same net charge
1, so that the elements they represent should have many chemi-
cal similarities. These they certainly have as both are alkali
metals. In general the atoms of elements belonging to the
same group chemically have the same number of electrons in
the outer shell and the same net charge on the kernel. On this
view it will be seen that the similarity of isotopes may be
regarded as the extreme limiting case of the similarities long
observed . between elements of the same chemical group.
94. Abnormal compounds formed by charged atoms.
The tendency of elements to form compounds with each other,
that is the property with which the idea of valency is associated,
is ascribed to the tendency of the atom to complete its outer
shell. This it can do either by parting with the electrons in
this shell and so promoting the next inner completed shell to
outer position, or by sharing the electrons in the atom of
another element so that they fill the gaps in its own outer shell.
We have already alluded to the success which has attended
this idea in explaining valency and the properties of chemical
compounds. For the present argument it will be enough if
it is understood that lithium and sodium (1) (2) (7) will very
readily part with their solitary valency electron and become
positively charged, i.e., will be strongly electropositive elements
with valency + 1 ; whereas a fluorine (4) will have an equally
powerful tendency to take up an electron and become nega-
tively charged and so will be a strongly electronegative
THE ELECTRICAL THEORY OF MATTER 99
element with valency 1. Both of these tendencies will be
satisfied if (4) and (7) combine forming the compound molecule
NaF, for the outer electron of (7) will enter the outer shell of
(4) thus forming two complete shells of eight (Langmuir's
octets). We trace the tendency of the atom of fluorine, or any
other halogen, to form compounds with the atom of an electro-
positive element, or with the atom of hydrogen, to the fact
that it has one too few electrons in its outer shell.
Now the only way we can give a positive charge to an atom
of neon (5) or (6) is by knocking one or more electrons out of
its outer shell. Suppose we remove one from (5) as indicated
by the dotted line. (5) now will have a similar outer shell to
(4) and a valency 1, so we may expect that atoms of the
inert gases carrying a single positive charge will behave
chemically in a similar manner to neutral halogen atoms and
will therefore be capable so long as they are charged of forming
hydrides. This very important idea was first suggested by
Sir J. J. Thomson in connection with the charged atoms of
chlorine 1 and certainly supplies a very satisfactory explana-
tion of the very abnormal hydrides of inert gases and com-
pounds such as OH 3 discovered in positive rays. The line
at 41 (Spectrum VI, Plate III) is probably to be put down
to a charged hydride of argon of this type. Exceedingly
faint lines at 5 in the case of helium, and 21 in the case of
neon, are probably to be ascribed to similar abnormal compound
(HeH) and (NeH) respectively. In the case of atoms carrying
more than one charge it can be generally stated that each
positive charge given to an atom will increase its negative
valency by one.
95. The failure of the additive law in respect to mass.
We have seen that, for velocities small compared with that of
light, the masses of the proton and the electron may be
regarded as universal constants. If the additive law were
strictly true as regards the summation of their masses it is
clear that any mass whatever, whether it were that of an atom
or a molecule, a planet or a star, or even the universe itself
could be expressed in the form NM where N is a pure integer
1 J. J. Thomson, Proc. Roy. Soc. 99A, 90, 1921.
100 ISOTOPES
and M the mass of the neutral system 1 proton + 1 electron
( the atom of hydrogen). The simplicity of this idea, which
is Prout's theory in the language of modern physics, is extremely
attractive ; but we know it to be false, for although the
discovery of isotopes has removed the difficulty of the grosser
fractions associated with such elements as neon and chlorine,
we are still left with the more minute but none the less real
one associated with hydrogen itself. To explain this the
additive law must be qualified by some such reasoning as is
contained in the following paragraph.
96. The explanation of the fractional mass of the
hydrogen atom by the hypothesis of " packing."
According to generally accepted views the proton and the
electron possess mass, or what on the relativity theory is
regarded as the same thing, weight, by virtue of the energy
in the electromagnetical field which surrounds them. It
can easily be shown on classical lines that if we give it a
spherical form a charge e spread uniformly over the surface
of the sphere will have a mass m when its radius a is such that
2 e 2
m - ; hence to give the electron its proper mass its charge
o CL
must be compressed to a sphere of diameter about 3-8 x 10~ 13
cm. By the same argument the proton will be nearly two
thousand times smaller and have a diameter 2-06 x 10~ 16 cm.
The extreme range of the diameter of atoms themselves is
1 5 x 10~ 8 cm., so that it will at once be realised that the
structure of an atom is an exceedingly open one, even more
so than that of our solar system.
Now it can be shown that if we bring two charges of opposite
sign sufficiently close together their fields will affect each
other in such a way that the mass of the system will be reduced.
This effect is quite inappreciable for distances comparable
with the diameter of an atom, but begins to make itself felt
when the distance apart is of the order of the size of the electron
itself as given above. The nucleus of the atom of an ordinary
element (not hydrogen) contains both protons and electrons
and is very small compared with the atom itself. Its dimen-
sions can be roughly determined by actual experiment in the
THE ELECTRICAL THEORY OF MATTER 101
case of the heavy elements and are found to be 'so* sinall thai
even to get in the electrons alone these would have to be
packed very closely together. Such a nucleus will contain
more protons than electrons, roughly twice as many, so that
it may be regarded as practically certain that : In the nuclei
of normal atoms the packing of the electrons and protons is so
close that the additive law of mass will not hold and the mass of
the nucleus will be less than the sum of the masses of its con-
stituent charges.
The nucleus of a hydrogen atom consists of one single free
proton, its planetary electron is too far away to cause any
effect so that it is clear that we shall find the mass associated
with the atom of hydrogen greater than one-fourth the mass
of a helium atom or one-sixteenth the mass of an oxygen
atom. The mass lost when four free protons and two free
electrons are packed close together to form a helium nucleus
(No. 9, p. 106) is roughly 0-7 per cent, of the whole and it can
be calculated that, if we take the value of the diameter of
the electron given above, the protons must approach nearer
than half of this to give so large a reduction. This means that
the charges must be so closely packed that the electrons are
actually deformed.
The whole number rule may now be simply translated into
a statement that the mean packing effect in all atoms is
approximately constant, and the unit of mass 1 when 0=16
will be (mass of a packed proton) + J (mass of free electron)
+ | (mass of packed electron). The whole number rule is not,
and never was supposed to be, mathematically exact, for this
would imply an identical packing effect in the case of all
atoms, an exceedingly improbable supposition. It is almost
certain that atoms of some elements, such as nitrogen, weigh
slightly more than a whole number (looser packing) while
those of others such as caesium or iodine may weigh
slightly less (closer packing). The limit of accuracy so far
attained in mass-spectrum measurement is not sufficient to
detect a change of the order expected, except in the case of
hydrogen, where the variation in mass is exceptionally high.
97. The structure of the nucleus. The manner in which
102 ISOTOPES
t/he units of electricity are arranged in the nucleus of an atom
has received a good deal of attention from theorists but ideas
on this subject are almost entirely of a conjectural character.
Thus Harkins x has proposed a constitutional formula for the
nuclei of all the elements. In this, besides electrons and
protons, he uses as building units a particles (4 protons -+- 2
electrons) of mass 4, and hypothetical units of mass 3 with a
single positive charge (3 protons -f 2 electrons). The matter
has been more recently discussed and nucleus models sug-
gested by Rutherford, 2 E. Gehrcke 3 and others.
The fact that the helium nucleus is almost exactly an
integer on the oxygen scale that is to say helium has approxi-
mately normal packing gives a distinct balance of probability
that helium nuclei actually exist as such in the nuclei of normal
elements. In support of this idea it has been stated that the
presence of helium nuclei inside the nuclei of radioactive
atoms is definitely proved by the ejection of a particles by
the latter. In the writer's opinion this is much the same as
saying that a pistol contains smoke, for it is quite possible
that the a particle, like the smoke of the pistol, is only formed
at the moment of its ejection. Brosslera 4 defends this view
and points out that if the alteration from looser to closer
packing of the charges forming the particle is at all large
energy will be liberated amply sufficient for the purpose of
detaching it and giving it the energy of an a ray. The
reason to expect that this energy will be set free will be
described in the next section. Brosslera's suggestion that
in the nuclei of radioactive atoms there are loosely bound
protons and electrons and that these, given something of the
nature of a certain exact and instantaneous correlation,
might combine to form an a particle is in good accordance
with the most reasonable theory of radioactive disintegration,
which was first put forward by Lindemann. 5
There are therefore two different ideas which we may regard
1 Harkins, Phys. Rev., 15, 73, 1920.
2 Rutherford, Proc. Roy. Soc., 97A, 374, 1920.
3 Gehrcke, Phys. Zeit., 22, 151, 1921.
4 Brosslera, Rev. Chim., 1, 42, 74, 1921.
5 Lindemann, Phtt. Mag., 30, 560, 1915.
THE ELECTRICAL THEORY OF MATTER 103
as working hypotheses. According to the first the nuclei of
atoms consist of helium nuclei, or a particles, held together
in some way so that their packing effect upon each other is
small ; and, in the case of atoms not having a mass of the
type 4n, additional protons and electrons. According to
the second we only have to suppose that the mean packing
of all the charges in the nucleus is such as will account for the
whole number rule with sufficient exactness, but that the
actual arrangement of the protons and electrons need not
necessarily be at all similar to that in a helium nucleus.
The experimental evidence is, so far, definitely in favour
of the first of these views. In their remarkable work on the
disintegration of light atoms by the collision of swift a rays
Rutherford and Chadwick * show that as the result of such
collisions swift hydrogen rays, i.e. free protons, are liberated
from the atoms of boron, nitrogen, fluorine, sodium, aluminium
and phosphorus. They point out that the masses of the
atoms of all these elements are of the types 4w + 2 and
4n + 3. The effect is not obtained from atoms of the type
4n so that this result suggests that in these the protons are
already all bound together to form helium nuclei.
98. Cosmical effects due to change of mass. It has
long been known that the chemical atomic weight of hydrogen
was greater than one quarter of that of helium, but so long as
fractional weights were general there was no particular need
to explain this fact, nor could any definite conclusions be
drawn from it. The results obtained by means of the mass-
spectrograph 2 remove all doubt on this point, and no matter
whether the explanation is to be ascribed to packing or not,
we may consider it absolutely certain that if hydrogen is
transformed into helium a certain quantity of mass must be
annihilated in the process. The cosmical importance of this
conclusion is profound and the possibilities it opens for the
future very remarkable, greater in fact than any suggested
before by science in the whole history of the human race.
We know from Einstein's Theory of Relativity that mass
1 Rutherford and Chadwick, Phil Mag., 42, 809, 1921.
2 V. p. 70.
104 ISOTOPES
and energy are interchangeable l and that in C.G.S. units a
mass m at rest may be expressed as a quantity of energy
me 2 , where c is the velocity of light. Even in the case of the
smallest mass this energy is enormous. The loss of mass when
a single helium nucleus is formed from free protons and
electrons amounts in energy to that acquired by a charge e
falling through a potential of nearly thirty million volts. A
swift a ray has an energy of three to four million volts so that
the change of packing suggested by Brosslera need not be
nearly so great to provide the energy needed. If instead of
considering single atoms we deal with quantities of matter
in ordinary experience the figures for the energy become
prodigious.
Take the case of one gramme atom of hydrogen, that is to
say the quantity of hydrogen in 9 c.c. of water. If this is
entirely transformed into helium the energy liberated will be
0077 X 9 X 10 20 = 6-93 x 10 18 ergs.
Expressed in terms of heat this is 1-66 x 10 ll calories or in
terms of work 200,000 kilowatt hours. We have here at last
a source of energy sufficient to account for the heat of the
Sun. 2 In this connection Eddington remarks that if only
10 per cent, of the total hydrogen on the Sun were trans-
formed into helium enough energy would be liberated to main-
tain its present radiation for a thousand million years.
Should the research worker of the future discover some
means of releasing this energy in a form which could be
employed, the human race will have at its command powers
beyond the dreams of scientific fiction ; but the remote
possibility must always be considered that the energy once
liberated will be completely uncontrollable and by its intense
violence detonate all neighbouring substances. In this event
the whole of the hydrogen on the earth might be transformed
at once and the success of the experiment published at large
to the universe as a new star.
1 Eddington, Time, Space and Gravitation, p. 146, Cambridge, 1920.
* Eddington, Brit. Assoc. address, 1920 ; Perrin, Scientia, Nov., 1921.
THE ELECTRICAL THEORY OF MATTER 105
99. The stable systems of protons and electrons
known to occur. Starting with our standard bricks, the
protons and electrons, we may make, theoretically at least,
an infinity of systems by the combination of any number of
each. It is interesting to consider the systems actually occur-
ring in practice, that is to say those which are sufficiently stable
to give definite evidence of their existence. The follow-
ing table gives, in order of mass, the first twenty -four known.
Where the circles representing the charges touch each other,
to form nuclei, the packing is extremely close, where they do
not touch they are to be taken as distant thousands of times
further from each other. The masses of the first twelve are
deduced as follows : The most accurate value for the chemical
atomic weight of hydrogen is 1-0077 (0 = 16), and as it is very
improbable that it consists of isotopes we take this as the
mass of a neutral hydrogen atom. The mass of the electron
is 0-00054 and as the packing effect is nil we arrive at the
figure 1-0072 for the mass of the proton, and this agrees within
the experimental error with that directly determined by the
mass-spectrograph. The most probable value of the mass of
a neutral helium atom is 4-00(0) we will assume the last figure
for the sake of simplicity. The masses of (13) to (24) are less
accurately known.
The stability, where known, is expressed in volts and
represents the potential through which a charge e must fall
in order to acquire sufficient energy to disrupt the particular
configuration concerned. This is the ionisation potential in
the case of atoms.
106
ISOTOPES
Number 1
of diagram.
Atomic
Number.
Nuclear
Constitution.
43
O
i
Stability.
Description.
1
-1
0-00054
_
Electron
f
2
1
1+
+ 1
1-0072
__
Proton or
positively
charged H
atom
3
1
1+
1-0077
14
Neutral H
atom
4
1
1+
_1
1-0082
Negatively
-
charged H
atom
5
1
1+
+ 1
2-0149
Positively
charged H
molecule
6
o o
1
1+
2-0154
4-3
Neutral H a
molecule
7
1
1+
+1
3-0226
small
Positively
charged
8
o o o
1
1+
3-0231
small
Neutral H 3
9
8
2
4+2-
+2
3-999
>3x 10 6
Doubly
charged
helium
atom or
alpha ray
10
c 88
2
4+2-
+ 1
3-999
55
Singly
charged
helium
atom
11
2
4+2-
4-000
25
Neutral
helium
atom.
THE ELECTRICAL THEORY OF MATTER 107
N
of d
Atomic
Number.
Nuclear
Constituti
Description.
12
13
14
15
16
17
18
19
20
21
22
23
24
o
o
O O
o o
o o
o o
o
o
o
o o
o
o
4+2-
6+3-
6+3-
7+4-
7+4-
6+3-
7+4
9+5-
9+5-
+1
+1
+1
+1
10+5-
10 + 5-
10+5-
11+6-
+2
+2
5-007
6-0
6-0
7-0
7-0
6-0(07)
7-0(07)
9-0
9-0
10-00
10-00
10-00
11-00
4.9*
4.9
3-3*
Positively
charged
HeH
Positively
charged
Li 6 atom
Neutral Li 6
atom
Positively
charged
Li 7 atom
Neutral Li 7
atom
Neutral
Li 6 H
molecule
Neutral
Li 7 H
molecule
Positively
charged
Be atom
Neutral Be
atom
Doubly
charged
B 10 atom
Positively
charged
B 10 atom
Neutral B 10
atom
Doubly
charged
B 11 atom
* Calculated from frequency of radiation.
CHAPTER IX
ISOTOPES AND ATOMIC NUMBERS
100. The relation between chemical atomic weight
and atomic number. Inasmuch as it is now recognised
to be in general merely a statistical mean value the importance
of the chemical atomic weight has been greatly reduced by
the discovery of isotopes. Its position as the natural numerical
constant associated with an element has been taken by the
atomic number, though from the point of view of chemical
analysis the chemical atomic weight is just as important as
it ever was.
The possibility of anomalies in the order of the elements
in the periodic table when their chemical atomic weights are
considered, is now obvious enough. The true weights of the
atoms as directly determined, are so intermingled in the order
of the natural numbers and the proportions present in complex
elements so varied that such anomalies are bound to occur,
indeed it is rather surprising there are not more.
The following table (Fig. 17) shows the masses of the isotopes
of three groups of elements now completely investigated. The
approximate proportions present are indicated by the heights
of the columns ; plain for the alkali metals, black for the inert
gases, and hatched for the halogens. The anomalous order
of argon and potassium is at once seen to be due to the fact
that whereas the heavier constituent of argon is present in
much the greater proportion, in potassium the reverse is the
case. Had the proportions of heavier and lighter isotopes
been similar in each case the atomic weight of potassium would
have been greater instead of less than that of argon.
108
ISOTOPES AND ATOMIC NUMBERS
109
19
20
~T
21
22 3 23
35
Fluorine (9) Neon (10)
Sodium* (11)
36 37 38 39 40
(Chlorine 17) Argon (18)
Potassium (19)
41
I
78 79 80 81 82 83 84 85 86 87
Bromine (35) Krypton (36) Rubidium (37).
1
FIG. 17
. 1
, f I
B
27 128 129
Iodine
Isotopes of the
130 131 132 133 134 135 136
(53) Xenon (54) Caesium (55)
Halogens, the inert gases and the alkali metals.
101. Statistical relations exhibited by elements and
their isotopes. Although our knowledge of true atomic
weights is far from complete, for out of eighty-seven existing
elements only twenty-seven have been analysed, of which
thirteen are simple, interesting relations have already become
clear which are stated in the form of rules as follows :
In the nucleus of an atom there is never less than one electron to
every two protons. There is no known exception to this law.
It is the expression of the fact that if an element has an atomic
number N the atomic weight of its lightest isotope cannot be
less than 2N. Worded as above, the exception in the case of
hydrogen is avoided. True atomic weights corresponding
exactly to 2N are known in the majority of the lighter elements
up to A 33 . Among the heavier elements the difference between
110 ISOTOPES
the weight of the lightest isotope and the value 2N tends to
increase with the atomic weight ; in the cases of mercury it
amounts to 37 units. The corresponding divergence of the
mean atomic weights from the value 2N has of course been
noticed from the beginning of the idea of atomic number.
The number of isotopes of an element and their range of
atomic weight appear to have definite limits. Since the
atomic number only depends on the net positive charge in
the nucleus there is no arithmetical reason why an element
should not have any number of possible isotopes. An
examination of the tables of results given on p. 89 and
at the end of the book show that so far the largest number
determined with certainty is 6 in the case of krypton. It
is possible that xenon has even more, but the majority of
complex elements have only two each. The maximum differ-
ence between the lightest and heaviest isotope of the same
element so far determined is 8 units in the cases of krypton
and xenon. The greatest proportional difference, calculated
on the lighter weight, is recorded in the case of lithium, where
it amounts to one-sixth. It is about one-tenth in the case of
boron, neon, argon and krypton.
The number of electrons in the nucleus tends to be even. This
rule expresses the fact that in the majority of cases even
atomic number is associated with even atomic weight
and odd with odd. If we consider the three groups of
elements, the halogens, the inert gases and the alkali metals,
this tendency is very strongly marked. Of the halogens odd
atomic numbers all 6 ( + 1 ?) atomic weights are odd. Of
the inert gases even atomic numbers 13 (+ 2 ?) are even and
3 odd. Of the alkali metals odd atomic numbers 7 are
odd and 1 even. In the few known cases of elements of the
other groups the preponderance, though not so large, is still
very marked and nitrogen is the only element yet discovered
to consist entirely of atoms whose nuclei contain an odd number
of electrons.
A further interesting result is the absence of isobares. So
far none have been definitely identified, but it is quite obvious
that in the cases of elements such as calcium and selenium they
must exist, for the supply of integers in the region of their
ISOTOPES AND ATOMIC NUMBEKS 111
atomic weights have been exhausted by the needs of other
elements.
A table of the first 40 natural numbers and the true atomic
weights corresponding to them is given in Fig. 18. The gaps
Li
LJ] [B C |B|B|C| |N| |o| |F|Ne
12 3 * 5 67 8 9 1O 11 12 13 14; 15 16 17 18 19 20
|Ne |Na|Mg|Mg |Mg|Al?[si
Cl A Cl
21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
FIG. 18. The first 40 natural numbers, showing those occupied
by atomic weights of known elements.
are particularly interesting and seem to show no semblance
of regularity. It is very clear that many more experimental
results will have to be obtained before any satisfactory theory
for the occurrence of these, or of the other laws, is to be formu-
lated.
102. The preponderance of elements of even atomic
number. In discussing the nuclear structure of elements
the question of their relative abundance in nature is one of
great interest. This may be estimated by direct chemical
analysis of the Earth's crust, and such extra-terrestrial sources
as are available in the form of meteorites. The spectroscope
will tell us what elements are present in the stars, but unfortu-
nately it does not give much direct information as to their
relative quantities.
On this question we can classify to use biological terms
either by individuals or by species. We may examine the
percentage composition, which will give a measure of the total
number of individual atoms of each element present, or we
may inquire into the number of different nuclear species which
occur and classify them without respect to their individual
abundance.
A very valuable discussion from the first point of view has
been published by Harkins, 1 who considers the percentage
composition of meteorites and of parts of the Earth's crust.
He demonstrates in a most convincing manner that there are
1 Harkins, Jour. Amer. Chem. Soc., 39, 856, 1917.
112 ISOTOPES
immensely more atoms of elements of even atomic number.
This interesting preponderance can, with a reasonable amount
of probability, now be extended to even atomic weight, by
the statistics given in the preceding paragraphs, but it will
not be certain until the constitution of certain abundant
elements such as iron has been actually determined.
The second point of view can be examined by means of the
atomic weights of the radioactive isotopes and also by the true
atomic weights given by the mass-spectra. In both cases
nuclear systems of even atomic number are found to predomi-
nate. The mass-spectra of 13 elements of even, and 14 ele-
ments of odd atomic number indicate 32 isotopes of even
atomic number and 20 of odd. The average element of even
atomic number has therefore 2-5 isotopes to 1-4 for each element
of odd atomic number.
The table on p. 15 shows that among the radioactive isotopes
the preponderance is greater 32 as against 10 but it is
possible that the former figure may include some atomic
systems absolutely identical though of different origin.
103. The constancy of chemical atomic weights.
One of the first difficulties in the way of accepting the idea of
the complex constitution of an element such as chlorine was
the constancy of its atomic weight. This had been determined
by many different observers using different methods and the
results were always the same within a very small experimental
error. This difficulty may be met, in the first place, by noting
that the vast majority, if not all, of the really accurate values
were obtained from chlorine which must have been originally
derived from the sea. The sea has been mixed for so long
that it would be absurd to expect to find chlorines of different
chemical atomic weights in it. Had ordinary galena been
the only source of lead used in the atomic weight determina-
tions given on page 16 no difference would have been found.
It was only by examining the lead from extraordinary radio-
active sources that the results were obtained which gave such
definite and valuable support to the theory of isotopes.
The atomic weight of chlorine from sources other than the
sea is now receiving the attention of chemists, though it is
ISOTOPES AND ATOMIC NUMBERS 113
naturally very difficult to be at all sure that any known source
of chlorine is not of marine origin. Mile. Irene Curie 1 has
examined the atomic weight of chlorine from three minerals
whose marine origin seems unlikely. The values obtained
from a sample of sodalite (sodium aluminium chlorosilicate)
from Canada, and from a sample of calcium chlorophosphate
from Norway agree with the value for chlorine from sea- water.
The value 35-60, for chlorine from a sample of sodium chloride
from a desert region in Central Africa was slightly high.
The comparison of the atomic weights of terrestrial and
meteoric nickel made by Baxter and Parsons 2 is interesting
in this connection. As a mean of nine determinations with
the terrestrial material the figure 58-70 was found, whilst
three experiments with meteoric nickel gave 58-68. The
standard value found by Richards and Cushman was 58-68
(Ag = 107-88). The difference found between terrestrial and
meteoric nickel is considered to be within the limits of experi-
mental error, but further comparisons are to be made.
The writer regards these negative results as having a cause
probably much more fundamental than the mere mechanical
mixing of the different constituent isotopes during the history
of the body containing them, namely a constancy of proportion
during the evolution of the elements themselves. This will
be considered later. The case of the radioactive leads is
entirely exceptional. These substances have been produced
continuously during the history of the earth's crust and are
being so produced to-day. Although ordinary lead may con-
sist of isotopes which is practically certain and these isotopes
may be identical in every respect with those produced in the
last stage of radioactive disintegration, yet there is no reason
whatever to assume that ordinary lead is itself the accumulated
result of these processes. It takes its place among the other
ordinary elements and would doubtless have done so had
thorium and uranium never existed.
104. The agreement between the chemical atomic
weight and the mean atomic weight deduced from the
mass spectrum. The mean atomic weight of the isotopes
1 I. Curie, Compt. Rend. 172, 1025, 1921.
2 Baxter and Parsons, Jour. Amer. Chem. Soc., 43, 507, 1921.
I
114
ISOTOPES
of a complex element can be calculated if the relative intensities
of their lines in the mass-spectrum is known. This has been
directly measured by Dempster. 1 The charged particles of
isotopes of the same element are practically certain to affect
the photographic plate to the same extent as each other,
hence we can obtain a rough estimate of their relative pro-
portion by comparing the intensities of the lines. If this is
done it is found that the great majority of the elements so
far tested give mean results in good agreement with the
accepted chemical values. The following table gives the data
concerning four in which the difference is noteworthy :
Element.
Atomic
Weight.
Mean from
Mass-spectrum.
Difference.
Per cent.
Difference.
Boron ....
10-90
10-750-07
0-15
1-37
Krypton .
82-92
83-5 0-3
0-6
0-72
Xenon
130-2
131-3 0-3
1-1
0-85
Caesium
132-81
133 0-3
0-2
0-05
The case of boron is the most difficult to account for. The
masses of its isotopes 10 and 11 certainly do not differ from
integers by more than one or two parts in a thousand. The
ratio of the intensities of their second order lines 5 and 5-5
(and there were no other substances present which could
possibly give such lines) is equally certainly not as high as
9:1. It was for this reason that a third isotope 12 was
suspected, but as no evidence of this has been found it seems
most probable that the chemical atomic weight is still slightly
too high.
The atomic weights of krypton and xenon are not of course
chemical in the ordinary sense, as they are deduced direct
from density determinations. Any trace of the impurity most
likely to be present, argon in the first case, krypton in the
second, would tend to make the densities too low, and this
appears the most likely explanation.
In the case of caesium the chemical result may be correct,
for the probable error in the determination of mass is at least
as large as the discrepancy. On the other hand caesium
1 F. p. 81.
ISOTOPES AND ATOMIC NUMBERS 115
appears to be a simple element, in which case its chemical
atomic weight must represent the true weight of its atoms.
Any error in this figure would probably be of the sign suggested,
for it is the heaviest member of its chemical group. If, how-
ever, as is possible, the true mass of its atom differs from an
integer by as much as 0-2 it is a fact of the greatest interest.
105. The meaning of the word " element." The
exact idea conveyed by the word " element " in chemistry
and physics has given rise to endless difficulties in the past.
In this connection Crookes in 1886 sums up the matter as
follows : " Of the attempts hitherto made to define or explain
an element, none satisfy the demands of the human intellect.
The textbooks tell us that an element is ' a body which has
not been decomposed ' ; that it is ' a something to which we
can add, but from which we can take away nothing,' or ' a
body which increases in weight with every chemical change.'
Such definitions are doubly unsatisfactory : they are provi-
sional, and may cease to-morrow to be applicable to any
given case. They take their stand, not on any attribute of
things to be defined, but on the limitations of human power ;
they are confessions of intellectual impotence."
There was good reason for this dissatisfaction. The dis-
covery ten years later of the electron, and the subsequent
electrical theory of matter robbed the word of any pretence
to its original meaning ; for although Ramsay attempted to
introduce into chemistry electricity itself as an element, it
soon became obvious that this extension was unsuitable. The
discovery of isotopes brings us face to face with two possible
alternatives. The first is to call each isotope, as it is dis-
covered, a new element. The second is to fix the word pre-
cisely, now and for the future, as meaning a substance with
definite chemical and spectroscopic properties which may or
may not be a mixture of isotopes in other words to associate
it exclusively with the conception of atomic number. On
this view there would be, corresponding to Moseley's numbers,
92 possible elements, of which 87 are known.
If we adopt the first of these alternatives a new word will
be necessary to express such substances as chlorine or mag-
116 ISOTOPES
nesium, hitherto called elements, and also the word element
would mean something entirely different from what it has
meant in all the chemical and physical literature of the past
century. It would moreover be still subject to alterations in
the f Jotire.
In the opinion of the writer the second alternative the
association of element with atomic number is much the
more preferable. The difficulties arising from it are practi-
cally confined to the radioactive substances which can differ
from one another even when their atomic numbers and atomic
weights are identical. This is not very serious, for the radio-
active elements are in a class by themselves and the special
nomenclature already applied to them could be retained or re-
vised as convenient without affecting that of general chemistry.
106. Disintegration theory of the evolution of the
elements. A theory has been put forward by some writers
that all the elements occurring in nature are the result of
radioactive disintegrations of the ordinary type, but continued
far beyond the ordinary limit observed at present. For
instance, if we continue the a ray changes of the thorium series
far enough we shall ultimately reach helium. The emission
of an a particle is the only change known to occur which alters
the atomic weight and it always does so by 4 units at a time.
Hence from thorium we shall get a series of elements or iso-
topes of atomic weights from 232 to 4 of the general type 4w.
Uranium in the same way will yield a similar series of the type
4ft + 2. In order to obtain isotopes of odd atomic weight
it is necessary to postulate parent elements of the type 4ft + 1
and 4ft + 3.
Using hypotheses based on this general idea Van den Broek, 1
Harkins, 2 Kohlweiler, 3 Kirchoff 4 and others have built up the
most elaborate systems of isotopes.
1 Van den Broek, Phys. Zeit., 17, 260, 579, 1 916 ; 22, 164, 1921.
2 Harkins and Wilson, Jour. Am. Chem. Soc., 37, 1367, 1915 ; Har-
kins and Hall, ibid., 38, 169, 1916 ; Harkins, Phys. Rev., 15, 73,
1920 ; Nature, 105, 230, 1920 ; Jour. Amer. Chem. Soc., 42, 1956,
1920 ; Phil. Mag., 42, 305, 1921.
3 Kohlweiler, Zeit. fur physical. Chem. , 94, 5 1 3, 1 920 ; Phys. Zeit. , 21,
311, 543 ; 22, 243, 1921. 4 Kirchoff, ibid., 21, 711, 1920.
ISOTOPES AND ATOMIC NUMBERS 117
The writer regards this view as unlikely and misleading.
In the first place it does not appear to succeed in its objects.
As an explanation of how the elements may have been evolved
it starts with at least four elements as complicated as any
known to exist, which does not advance the inquiry very
much. On the other hand it may be used to predict the
atomic weights of the isotopes composing known elements,
and a great many predictions of this kind have been made.
Here, though the measure of its success has varied to some
extent with the particular modification of the theory employed,
it has never been worthy of serious consideration. In cases
where two or three isotopes of a given element were pre-
dicted they proved as often wrong as right, and when the
number of isotopes of integral atomic weights was so large
that some agreements were inevitable the argument obviously
loses all its force.
Another objection is that radioactive transformations do
not continue, as far as we can see, beyond the stage (lead)
indicated in the diagrams on p. 15. The lighter elements are
definitely not radioactive. The radioactivity of potassium
and rubidium is exceedingly small and its nature doubtful ;
in any case it is best ascribed to minute vestiges of radioactive
isotopes, not to feeble radioactivity of the main constituents.
It seems therefore more reasonable, for the present, to regard
the property of radioactivity as absent entirely from the
inactive elements than to suppose it present but too weak to
be detected. It must not be gathered from these remarks
that it is considered impossible to imagine physical conditions
violent enough to disrupt the nuclei of light atoms, but rather
that the mechanism causing such disruption need not be
similar in any way to that causing normal radioactivity.
107. Crookes' theory of the evolution of the elements.
A more attractive theory than the one given above is that
the complex atoms of matter have been evolved by the aggre-
gation of simpler atoms. This idea has received a good deal of
attention in the past. Crookes 1 remarks on it as follows :
" Let us picture the very beginnings of time, before geological
1 Crookes, Brit. Assoc. address, 1886.
118 ISOTOPES
ages, before the earth was thrown off from the central nucleus
of molten fluid, before even the sun himself had consolidated
from the original protyle. Let us still imagine that at this
primal stage all was in an ultra-gaseous state, at a temperature
inconceivably hotter than anything now existing in the visible
universe ; so high indeed that the chemical atoms could not
yet have been formed, being still far above their dissociation
point. In so far as protyle is capable of radiating or reflecting
light, this vast sea of incandescent mist, to an astronomer in a
distant star, might have appeared as a nebula, showing in
the spectroscope a few isolated lines, forecasts of hydrogen,
carbon and nitrogen spectra.
"But in due course of time some process akin to cooling,
probably internal, reduces the temperature of the cosmic
protyle to a point at which the first step in granulation takes
place ; matter as we know it comes into existence, and atoms
are formed."
This vivid picture may be brought up to date by the sub-
stitution of free protons and electrons for the hypothetical
protyle. We can imagine regions containing matter where
the temperature is so high that not only is the dissociation of
atoms from atoms and nuclei from planetary electrons com-
plete but also protons and electrons are in a state of agitation
so violent that even the most stable nuclei cannot be formed.
We should have here matter of the simplest form we can
imagine, or rather of no form at all, simply a more or less
neutral electric gas. Such a condition is by no means impos-
sible in our universe and may actually occur during one of those
excessively violent catastrophes occurring in far distant space
and observed by us as new stars.
By some such cooling process as that suggested by Crookes
we easily imagine the free charges combining to form the
nuclei of elements. Whether those of heavier elements are
formed direct by the charges getting into particular geometrical
relations with each other, or whether helium nuclei are formed
first and then subsequently coalesce depends on which theory
of nuclear structure is adopted. In any case vast quantities of
energy will have to be radiated off and this radiation may be
of such extremely high frequency that it is capable of dis-
ISOTOPES AND ATOMIC NUMBERS 119
rupting nuclei themselves, so that there might be at this stage
rapid and continuous transformations from heavier to lighter
nuclei and vice versa.
For the present we are interested in the number of each
type of atom which survives. It is obvious that if the con-
ditions of cooling are practically identical throughout the whole
mass there is no reason why the composition of the matter
produced should vary. If 3 atoms of Cl 35 are formed to every
1 of Cl 37 at any one point the same ratio must hold at every
point so that a complex element of constant atomic weight
will be formed. But it is much more likely that different parts
of this primordial mass will undergo their transformations
under different rates of cooling, etc., so it is worth while
inquiring if variation in the mean atomic weight of a complex
element is to be expected.
The quantity of one particular atomic nucleus formed will
probably depend (a) on the probability of a certain configura-
tion of charges happening as a chance event ; (6) the stability of
the particular nucleus formed as the result of that event.
Again to take the case of chlorine each isotope may be regarded
as completely stable and the relative quantities formed will
simply depend on condition (a). Now it is not unreasonable
to suppose that this is not seriously affected by different rates
of cooling, and in this case the isotopes will be evolved in
constant proportion. As we know of no natural process by
which the proportion of isotopes can be altered appreciably
the complex elements will have to-day the same chemical
atomic weight as when they were first formed.
The above argument is of course purely a speculative one,
and the conclusion drawn from it would fall to the ground at
once if noteworthy differences of atomic weight in a single
complex element were found supposing that element was not
the product of a radioactive change at different points on the
earth's surface. It may be worth noting that condition (a)
suggests that, in general, the lighter atoms will outnumber
the heavier ones. In all matter available in nature this
preponderance is actually enormous.
If the matter forming the earth ever went through a prim-
ordial stage such as that suggested above it certainly did so
120 ISOTOPES
more than 10 9 years ago. It follows that of the radioactive
elements then formed only two, thorium and uranium, will
now be found on the earth, for the other radioactive elements
existing to-day are of such short period that they must have
been formed since. Hence we may divide the original elements
very simply and definitely into two groups : (1) All the
inactive elements, whose nuclei are sufficiently simple to be
stable ; (2) Thorium and Uranium, whose nuclei are so
complex that they are only partially stable.
Other less stable elements may have been formed then but
there can be no proof of this for they would, in any case, have
disappeared long ago, and it is clear that the other radioactive
elements now found can all be regarded as formed from the
two parent elements in comparatively recent times.
CHAPTER X
THE SPECTRA OF ISOTOPES
108. The Spectra of isotopes. As has already been
stated 1 the first experimental work on the spectra of isotopes
was that of Russell and Rossi in 1912 who failed to distinguish
any difference between the spectrum of thorium and that of a
mixture of thorium and ionium containing a considerable
percentage of the latter. The same negative result was
obtained by Exner and Haschek. 2 During the fractional
diffusion of neon 3 no spectroscopic difference was detected
between the heaviest and the lightest fraction, though as the
separation was small this negative evidence was not very
strong. In 1914 Soddy and Hyman showed that the spectrum
of lead derived from thorium was identical with that of ordinary
lead. 4 Furthermore in the same year the experiments of
Richards and Lembert, 5 Honigschmidt and Horowitz, 6 and
Merton 7 proved the same result. Merton concluded from his
1914 experiments that the difference in wave-length for the
A 4058 line must be less than 0-003 A. Before going on to
consider the more recent results it will be as well to discuss the
magnitude of the difference to be expected from theory.
109. The magnitude of the Gravitational effect. In
the Bohr theory of spectra the planetary electrons of the atom
rotate round the central positively charged nucleus in various
1 V. p. 9.
2 Exner and Haschek, Sitz. Akad. Wiss. Wien, iia, 121, 175, 1912.
3 V. p. 39.
4 Soddy and Hyman, Jour. Chem. Soc., 105, 1402, 1914.
5 Richards and Lembert, Jour. Amer. Chem. Soc., 36, 1329, 1914.
6 Honigschmidt and Horowitz, Sitz, Akad. Wiss. Wien, iia, 123,
1914.
7 Merton, Proc. Roy. /Soc., 91A, 198, 1914.
121
122 ISOTOPES
stable orbits. The frequencies of the spectral lines emitted
by the element are associated in an absolutely definite manner
with the rotational frequencies of these orbits which are
calculated by what is known as a " quantum " relation.
Without going further into the theory it will be seen at once
that if we alter the force acting between the central nucleus
and its planetary electrons these orbits will change and with
them the frequency of the light emitted. It is therefore of
interest to examine the magnitude of the change, to be expected
from this theory, when we alter the mass of the nucleus without
changing its charge, and so pass from one isotope to another.
The difference in the system which will first occur to one is
that although the electrical force remains the same the gravi-
tational force must be altered. The order of magnitude of
the change expected in the total force will clearly be given by
considering the ratio between the electrical and gravitational
forces acting, to take the simplest case, between the proton
and the electron in a neutral hydrogen atom.
Assuming the law of force to be the same in both cases, this
ratio is simply e 2 /GMm ; where e is the electronic charge
4-77 x 10" 10 , G the universal gravitational constant 6-6 x 10~ 8 ,
M the mass of the proton 1-66 x 10~ 24 , and m the mass of the
electron 9-0 x 10" 28 . Putting in these numerical values we
obtain the prodigious ratio 2-3 X 10 39 . In other words the
effect of doubling the mass of the nucleus without altering its
charge would give the same percentage increase in the total
pull on the planetary electron, as would be produced in the
pull between the earth and the moon by a quantity of meteoric
dust weighing less than one million millionth of a gramme
falling upon the surface of the former body. The gravitational
effect may therefore be dismissed as entirely negligible.
110. Deviation of the Bohr orbits due to change in
the position of the centre of gravity of the rotating
system. Although we may neglect the gravitational effect
there is another, of quite a different order, which arises in the
following manner. The mass of the electron compared with
that of the nucleus is small but not absolutely negligible, hence
it will not rotate about the nucleus as though that were a
THE SPECTRA OF ISOTOPES 123
fixed point, but both will rotate about their common centre
of gravity. The position of this centre of gravity will be
shifted by any alteration in the mass of the nucleus. If E, M
and e, m are the respective charge and mass of the nucleus and
the rotating electron, the equation of motion is
rM Ee
m x ^ w 2 =
M + m r 2
where r is the distance between the two charges and w the
angular velocity. Bohr J introduced this effect of the mass of
the nucleus in order to account for the results obtained by
Fowler. 2 The Bohr expression for the frequency then becomes
__ 2jr 2 e 2 E 2 mM
" h* (M + m)
where e, E and m, M are the charges and masses of the electron
and nucleus respectively. If we suppose that the atomic
weight of lead from radium to be one unit less than that of
ordinary lead, this theory predicts a difference in wave-length,
for the principle line, of 0-00005 A between the two, a quantity
beyond the reach of the most delicate methods of spectrum
analysis used up to the present.
111. Later experiments of Aronberg and Merton.
In 1917 Aronberg, 3 applying the extremely high dispersion
derived from the spectrum of the sixth order of a Michelson
10-inch grating to the line A 4058 emitted from a specimen of
radio-lead of atomic weight 206-318, observed a difference of
0-0044 A between this and ordinary lead, of atomic weight
207-20. This remarkable result has been since confirmed by
Merton of Oxford 4 who gives the difference of wave-length
between radio-lead from pitchblende and ordinary lead as
0-0050^0-0007, Merton made use of a totally different optical
system, namely a Fabry and Perot etalon, so that the agreement
is very striking.
It is to be noticed that the effect observed was not a mere
1 Bohr, Nature, 92, 231, 1913.
2 Fowler, Nature, 92, 95, 1913.
3 Aronberg, Proc. Nat. Acad. Sci., 3, 710, 1917, and Astrophys,
Jour., 47, 96, 1918.
4 Merton, Proc. Roy. Soc., 96A, 388, 1920.
124
ISOTOPES
broadening of the line but a definite shift, and that, though
of the same sign, it is about one hundred times greater than
that predicted by the Bohr theory. Merton also found a shift
of 0-0022^0-0008 A between the wave-length of thorite-lead
and ordinary lead, differing in atomic weight by about 0-6.
The heavier atom shows the higher frequency in all cases.
This remarkable discrepancy between the shift predicted by
theory and that actually observed has been discussed by
Harkins and Aronberg. 1
At a recent discussion on isotopes at the Royal Society 2
Merton commented upon the line 6708 A emitted by the
element lithium, which consists of two components 0-151 A
apart. If lithium is accepted as a mixture of isotopes 6 and 7, 3
he calculated that each of these components should be accom-
panied^ by a satellite, some sixteen times as faint, displaced by
0-087 A. So far he had not been able to observe such satellites.
Previous experiments of Merton and Lindemann 4 on the
expected doubling in the case of neon had given no conclusive
results on account of the physical width of the lines. It was
hoped that this difficulty could be overcome by the use of
liquid hydrogen temperatures.
Still more recently Merton 5 has repeated his experiments on
lead, using a very pure sample of uranium lead from Australian
Carnotite. His final results are indicated in the following
table :
A
r A( Carnotite lead)"!
L A ( ordinary lead) J
[Wave number (ordinary lead)"]
Wave-number (Carnotite lead )J
4058
0-011 0-0008
0-0650-005
3740
0-0074^0-0011
0-0530-008
3684
0-00480-0007
0-0350-005
3640
0-0070 0-0003
0-052^0-002
3573
0-00480-0005
0-0370-004
1 Harkins and Aronberg, Jour. Am. Chem. Soc., 42, 1328, 1920.
2 Merton, Proc. Roy. Soc., 99A, 87, 1921.
8 V. p. 86.
4 Lindemann, ibid.
5 Merton, Roy. Soc. Proc., 100A, 84, 1921.
THE SPECTRA OF ISOTOPES 125
It will be noticed that the shift for the line A 4058 is rather
more than twice that obtained before. Merton suggests that
the most probable explanation of this difference is evidently
that the Carnotite lead used is a purer sample of uranium lead
than that obtained from the pitchblende residues. It is also
apparent that the differences are not the same for different
lines, an interesting and somewhat surprising result.
112. " Isotope " effect on the Infra-red spectrum of
molecules. The extreme smallness of the isotope " shift "
described above in the case of line spectra emitted by atoms is
due to the fact that one of the particles concerned in the
vibration is the electron itself, whose mass is minute compared
with that of the nucleus. Very much larger effects should be
expected for any vibration in which two atoms or nuclei are
concerned, instead of one atom and an electron. Such a
vibration would be in the infra-red region of the spectrum.
This effect was first observed by Imes 1 when mapping the
fine structure of the infra-red absorption bands of the halogen
acids. In the case of the HC1 "Harmonic " band at 1-76^,
mapped with a 20,000 line grating, the maxima were noticed
to be attended by satellites. Imes remarks : " The apparent
tendency of some of the maxima to resolve into doublets in the
case of the HC1 harmonic may be due to errors of observation,
but it seems significant that the small secondary maxima are
all on the long-wave side of the principal maxima they accom-
pany. It is, of course, possible that still higher dispersion
applied to the problem may show even the present curves to
be composite."
Loomis 2 pointed out that these satellites could be attributed
to the recently discovered isotopes of chlorine. In a later
paper 3 he has shown that, if m-,. is the mass of the hydrogen
nucleus, and ra 2 the mass of the charged halogen atom, the
difference should be expressed by the quanity * , 2 the
1 m l j r m 2
square root of which occurs in the denominator of the expression
1 Imes, Astrophysical Journal, 50, 251, 1919.
2 Loomis, Nature, Oct. 7, 179, 1920.
3 Loomis, Astrophysical Journal, 52, 248, 1920.
126 ISOTOPES
for frequency. " Consequently the net difference between
the spectra of isotopes will be that the wave-lengths of lines
in the spectrum of the heavier isotope will be longer than the
corresponding lines for the lighter isotope in the ratio
1 + 1/1330 : 1 for chlorine and 1 + 1/6478 : 1 for bromine.
Since the average atomic weight of chlorine is 35*46 the amounts
of Cl 35 and Cl 37 present in ordinary chlorine must be as
1-54 : 0-46 or as 3-35 : 1 and, if the lines were absolutely sharp
and perfectly resolved, the absorption spectrum of ordinary
HC1 should consist of pairs of lines separated by 1/1330 of
their frequency and the one of shorter wave-length should have
about 3-35 the intensity of the other. The average atomic
weight of bromine is 79-92, hence the two isotopes are present
in nearly equal proportions and the absorption spectrum of
HBr should consist of lines of nearly equal intensity separated
by 1/6478 of their frequency."
The latter will be too close to be observed with the dispersion
employed. In the case of the HC1 band at 1-76 ^ the difference
of wave number on this view should be 4-3. The mean differ-
ence of wave number given by Loomis' measurements of 13
lines on Imes' original curves for this band is 4-5 ^ 0-4 corre-
sponding to 14 A in wave-length.
The spectroscopic confirmation of the isotopes of chlorine
has also been discussed by Kratzer, 1 who considers that the
oscillation-rotation bands of hydrogen chloride due to Imes 2
are in complete accordance with the theory.
1 H. Kratzer, Zeit. Physik., 3, 60, 1920.
2 LOG. cit.
CHAPTER XI
THE SEPARATION OF ISOTOPES
113. The Separation of Isotopes. The importance,
from purely practical and technical points of view, of the
theory of isotopes would have been insignificant had its
application been confined to the radioactive elements and their
products, which are only present in infinitesimal quantities
on the Earth. But now that the isotopic nature of many
elements in everyday use has been demonstrated, the possi-
bility of their separation, to any reasonable extent, raises
questions of the most profound importance to applied science.
In physics all constants involving, e.g., the density of mercury
or the atomic weight of silver may have to be redefined, while
in chemistry the most wholesale reconstruction may be
necessary for that part of the science the numerical founda-
tions of which have hitherto rested securely upon the constancy
of atomic weights.
It is therefore of great interest to consider in turn the
various methods of separation proposed and examine how
far they have been successful in practice.
114. Separation by Diffusion. The subject of the
separation of a mixture of two gases by the method of Atmoly-
sis or Diffusion has been thoroughly investigated by the late
Lord Rayleigh. 1 The diffusion is supposed to take place
through porous material. The conditions under which
maximum separation is to be obtained are that " mixing "
is perfect, so that there can be no accumulation of the less
diffusible gas at the surface of the porous material, and that
the apertures in the material through which the gases must
i Rayleigh, Phil. Mag., 42, 493, 1896,
127
128 ISOTOPES
pass are very small compared with the mean free path of the
molecules. If these conditions are satisfied he obtains as an
expression for the effect of a single operation :
* + y =_^ r^- + Y r
X + Y X + Y "-* X + Y *-*
where (X Y) (x, y) are the initial and final volumes of the
gases, ,M, v, the velocities of diffusion, and r the enrichment
of the residue as regards the second constituent.
The velocity of diffusion of a gas is proportional to the
square root of the mass of its molecules, so that if a mixture
of two isotopes is allowed to diffuse a change in composition
must be brought about. Now no known isotopes differ from
each other much in mass, so the difference between their
rates of diffusion will also be small, hence the above equation
may be written in the approximate form
/y I /\j 1 * ..
|_ = r* where k = " a small quantity and,
2L + x fj,
and, finally, the enrichment by diffusion of the residue as
regards the heavier constituent may be expressed with sufficient
accuracy by the expression
r = mz-m /Initial volume
v Final volume
where m it m 2 are the molecular masses of the lighter and
heavier isotope respectively. In the most favourable case
known at present, that of the isotopes of neon, the number
over the root is 21 so that the change in composition obtain-
able in a single operation will in practice be very small.
If we take the density of the original mixture as unity, the
increase in density of the residual gas to be expected from the
operation of diffusion will be approximately
w a + mi
Now neon consists of monatomic molecules differing between
each other in mass by 10 per cent, and the heavier is present
to the extent of 10 per cent. In the diffusion experiments
described on p. 39 the effective ratio of the initial volume to
THE SEPARATION OF ISOTOPES 129
the final volume was estimated as certainly greater than 500
and probably less than 10,000, so that r lies between 1-3
and 1-5. Hence the increase of density of the heavier residue
should have been between -003 and -005. It was actually
004.
115. The separation of the isotopes of chlorine by the
diffusion of HC1. In the case of other isotopic gaseous
mixtures the numerical obstacles in the way of practical
separation will be correspondingly greater. Thus in the case
of HC1 the 36th root is involved, and in that of HBr the 80th
root. The only way by which measurable increase in density
may be hoped for will clearly be by increasing the effective
ratio of the initial to final volumes to an heroic degree. This
can be done by experiments on a huge scale or by a vast
number of mechanical repetitions.
Harkins started to attack the HC1 problem in 1916 * using
the first of these two alternatives. In 1920 he mentions a
quantity of 19,000 litres of HC1 as having been dealt with in
these experiments. 2 In the following year 3 he published
numerical results indicating that a change in atomic weight
of 0-055 of a unit had been achieved.
At the recent discussion on isotopes 4 Sir J. J. Thomson
pointed out that a change in the molecular weight of HC1
should be caused by allowing a stream of the gas to flow over
the surface of a material which absorbed it. The higher
diffusion coefficient of the lighter isotope would result in it
being absorbed more rapidly than the heavier one, so that the
residue of unabsorbed gas should give a higher molecular
weight. This " free diffusion " without the interposition of
porous material has been recently tried in the Cavendish
Laboratory by E. B. Ludlam, but no measurable difference
has so far been detected.
116. Separation by Thermal Diffusion. It has been
1 Harkins, Jour. Amer. Chem. Soc., Feb., 1916.
2 Harkins, Science, Mar. 19, 1920 ; Nature, Apl. 22, 1920 ; see
also Phys. Rev., 15, 74, 1920 ; Science, 51, 289, 1920 ; Jour. Amer.
Chem. /Soc., 42, 1328, 1920.
3 Harkins, Science, Oct. 14, 1921 ; Nature, Oct. 3, 1921.
4 J. J. Thomson, Proc. Roy. Soc., 99A, 98, 1921.
K
130 ISOTOPES
shown on theoretical grounds independently by Enskog 1
and Chapman 2 that if a mixture of two gases of different
molecular weights is allowed to diffuse freely, in a vessel of
which the ends are maintained at two different temperatures
T,T', until equilibrium conditions are reached, there will be
a slight excess of the heavier gas at the cold end, and of the
lighter gas at the hot end. The separation attained depends
on the law of force between the molecules and is a maximum
if they behave as elastic spheres. The effect was experi-
mentally verified for a mixture of C0 2 and H 2 by Chapman
and Dootson, 3 and recently Ibbs 4 has demonstrated that the
separation can be carried out continuously and that the time
for equilibrium to be established is quite short.
Chapman has suggested 5 that thermal diffusion might be
used to separate isotopes. He shows that the separating
power depends on a constant Jc T . And when the difference
between the molecular masses m ly m 2 is small the value of
this is approximately given by
, _
~
17 ra 2
3 m z + m, 9-15 8-25 A
where Ai,A 2 denote the proportions by volume of each gas in
the mixture ; thus Ai + A 2 = 1. The actual separation is
given by
Ax - A', = - (A, - A' 2 ) = fc T log T'/T.
He gives the following numerical example : " Suppose that it is
desired to separate a mixture of equal parts of Ne 20 and Ne 22 ,
then, writing Wi = 20, m 2 = 22, Ai = A 2 = J, we find that
lc T = 0-0095. Suppose that the mixture is placed in a vessel
consisting of two bulbs joined by a tube, and one bulb is
maintained at 80 absolute by liquid air, while the other is
heated to 800 absolute (or 527 C.). When the steady state
has been attained the difference of relative concentration
between the two bulbs is given by the equation
1 Enskog, Phys. Zeit., 12, 538, 1911 ; Ann. d. Phys., 38, 750, 1912.
2 Chapman, Phil. Trans., 217A, 115, 1916; Phil. Mag., 34, 146,
1917.
3 Chapman and Dootson, Phil. Mag., 34, 248, 1917.
4 Ibbs, Proc. Roy. Soc., 99A, 385, 1921.
5 Chapman, Phil. Mag., 38, 182, 1919.
THE SEPARATION OF ISOTOPES 131
A! A r j = (A, A',) = 0-0095 log, 800/80
= 0-022
or 2-2 per cent. Thus the cold bulb would contain 48-9 per
cent. Ne 20 to 51*1 per cent. Ne 22 , and vice versa in the hot
bulb. By drawing off the contents of each bulb separately,
and by repeating the process with each portion of the gas, the
difference of relative concentrations can be much increased.
But as the proportions of the two gases become more unequal,
the separation effected at each operation slowly decreases.
For instance, when the proportions are as 3 : 1, the variation
at each operation falls to 1-8 per cent. ; while if they are as
10 : 1 the value is 1-2 per cent. This assumes that the mole-
cules behave like elastic spheres : if they behave like point
centres of force varying as the inverse nth power of the distance,
the separation is rather less; e.g., if n= 9, it is just over
half the above quantities."
Chapman points out that for equal values of log p/p and
log T/T pressure diffusion (centrifuging) is about three times
as powerful as thermal diffusion but suggests that it may be
more convenient to maintain large differences of temperature
than of pressure.
117. Separation by Gravitation or "Pressure Dif-
fusion." -When a heterogeneous fluid is subjected to a
gravitational field its heavier particles tend to concentrate
in the direction of the field, and if there is no mixing to counter-
act this a certain amount of separation must take place. If
therefore we have a mixture of isotopes in a gaseous or liquid
state partial separation should be possible by gravity or
centrifuging.
The simplest case to consider is that of the isotopes of neon
in the atmosphere and, before the matter had been settled by
the mass-spectrograph,' analysis of the neon in the air at very
great heights was suggested as a possible means of proving
its isotopic constitution. 1 The reasoning is as follows :
If M be the atomic weight, g the gravitational constant,
p the pressure, and p the density, then if no mixing takes
place dp .== gpdh, h being the height. In the isothermal
1 Lindemann and Aston, Phil. Mag., 37, 530, 1919.
132 ISOTOPES
layer convection is small. If it is small compared with
diffusion the gases will separate to a certain extent. Since
T is constant
RT/3 , dp M/>
* = TT and / =RT^'
whence p = p e~^ ,
p Q being the density at the height h at "which mixing by
convection ceases, about 10 kilometres, and A^ the height
above this level. If two isotopes are present in the ratio 1
to KO, so that the density of one is p Q and of the other KO/> O
at height h , then their relative density at height h -j- A/& is
given by
K =
Putting T = 220 as is approximately true in England,
A& being measured in kilometres. If M t M 2 = 2, therefore
" . _ e 1-075 x 10 *Afc
K."
It might be possible to design a balloon which would rise to
100,000 feet and there fill itself with air. In this case the
relative quantity of the heavier constituent would be reduced
from 10 per cent, to about 8-15, so that the atomic weight of
neon from this height should be 20-163 instead of 20-2. If
one could get air from 200,000 feet, e.g. by means of a long-
range gun firing vertically upwards, the atomic weight of the
neon should be 20-12.
A more practicable method is to make use of the enormous
gravitational fields produced by a high speed centrifuge.
In this case the same equation holds as above except that
g varies from the centre to the edge. In a gas therefore
dp _ Mv 2 dr _ Mco 2 ,
__
whence p = p e SET,
v being the peripheral velocity. Here again, if K is the
THE SEPARATION OF ISOTOPES 133
ratio of the quantities present at the centre, the ratio at the
edge will be
A peripheral velocity of 10 5 cm./s. or perhaps even 1-3 X 10 5
cm./s. might probably be attained in a specially designed
TC
centrifuge, so that ^ might be made as great as e-o -2 o5(M l -M 2 ) or
even e
If MI M 2 is taken as 2 a single operation would therefore
give fractions with a change of K of 0-65. In the case of neon
the apparent atomic weight of gas from the edge would be
about 0-65 per cent, greater than that of gas from the centre,
i.e. a separation as great as the best yet achieved in practice
by any method could be achieved in one operation. By
centrifuging several times or by operating at a lower tempera-
ture the enrichment might be increased exponentially.
Centrifuging a liquid, e.g. liquid lead, would not appear so
favourable, though it is difficult to form an accurate idea of
the quantities without a knowledge of the equation of state.
If compression is neglected and the one lead treated as a
solution in the other, a similar formula to that given above
holds. On assumptions similar to these Poole 1 has calculated
that a centrifuge working with a peripheral velocity of about
10* cm. /sec should separate the isotopes of mercury to an
extent corresponding to a change of density of 0-000015.
The only experiments on the separation of isotopes by the
use of a centrifuge, so far described, are those of Joly and
Poole 2 who attempted to separate the hypothetical isotopic
constituents of ordinary lead by this means. No positive
results were obtained and the check experiments made with
definite alloys of lighter metals with lead were by no means
encouraging.
118. Separation by Chemical Action or Ordinary
Fractional Distillation. The possibility of separating iso-
topes by means of the difference between their chemical
affinities or vapour pressures has been investigated very fully
1 Poole, Phil. Mag., 41, 818, 1921.
2 Joly and Poole, Phil. Mag., 39, 372, 1920.
134 ISOTOPES
from the theoretical standpoint by Lindemann. The thermo-
dynamical considerations involved are the same in both cases.
The reader is referred to the original papers l for the details
of the reasoning by which the following conclusion is reached :
" Isotopes must in principle be separable both by fractiona-
tion and by chemical means. The amount of separation to
be expected depends upon the way the chemical constant is
calculated and upon whether ' Nullpunktsenergie ' is assumed.
At temperatures large compared with f$v? which are the only
practicable temperatures as far as lead is concerned, the
difference of the vapour pressure and the constant of the
6v
law of mass action may be expanded in powers of ?=. The
most important term of the type log ~ is cancelled by the
chemical constant if this is calculated by what seems the only
reasonable way. The next term in ~ is cancelled by the
* Nullpunktsenergie ' if this exists. All that remains are
terms containing the higher powers of ^. In practice there-
fore fractionation does not appear to hold out prospects of
success unless one of the above assumptions is wrong. If the
first is wrong a difference of as much as 3 per cent, should
occur at 1200 and a difference of electromotive force of one
millivolt might be expected. Negative results would seem
to indicate that both assumptions are right."
As regards experimental evidence it has already been pointed
out that the most careful chemical analysis, assisted by radio-
active methods of extraordinary delicacy, was unable to achieve
the slightest separation of the radioactive isotopes. The
laborious efforts to separate the isotopes of neon by a differ-
ence of vapour pressure over charcoal cooled in liquid air also
gave a completely negative result.
119. Separation by evaporation at very low pressure.
If a liquid consisting of isotopes of different mass is allowed
1 Lindemann, Phil. Mag., 37, 523, 1919; 38,173, 1919.
* fiv is the " characteristic " and T the " Absolute " temperature.
THE SEPARATION OF ISOTOPES 135
to evaporate it can be shown that the number of light atoms
escaping from the surface in a given time will be greater than
the number of heavier atoms in inverse proportion to the
square roots of their weights. If the pressure above the
surface is kept so low that none of these atoms return the
concentration of the heavier atoms in the residue will steadily
increase. This method has been used for the separation of
isotopes by Bronsted and Hevesy, who applied it first to the
element mercury.
The mercury was allowed to evaporate at temperatures from
40 to 60 C. in the highest vacuum attainable. The evaporat-
ing and condensing surfaces were only 1 to 2 cms. apart, the
latter was cooled in liquid air so that all atoms escaping
reached it without collision and there condensed in the solid
form.
It will be seen that the liquid surface acts exactly like the
porous diaphragm in the diffusion of gases. 1 The diffusion
rate of mercury can be obtained approximately from the
diffusion rate of lead in mercury 2 and is such that the mean
displacement of the mercury molecule in liquid mercury is
about 5 x 10" 3 cm. sec." 1 . It follows that if not more than
5 x 10~ 3 c.cm. per cm. 2 surface evaporate during one second
no disturbing accumulation of the heavier isotope in the
surface layer takes place.
The separation was measured by density determination.
Mercury is particularly well suited for this and a notable
feature of this work was the amazing delicacy with which it
could be performed. With a 5 c.cm. pyknometer an accuracy
of one part in two millions is claimed. The first figures
published 3 were :
Condensed mercury. . . . 0-999981
Residual mercury .... 1-000031
The densities being referred to ordinary mercury as unity.
The later work was on a larger scale. 4 2700 c.cm. of mercury
were employed and fractionated systematically to about
1 V. p. 127.
2 Groh and Hevesy, Ann. der Phys., 63, 92, 1920.
3 Bronsted and Hevesy, Nature, Sept. 30, 1920.
4 Bronsted and Hevesy, Phil. Mag., 43, 31, 1922.
136 ISOTOPES
1/100,000 of its original volume in each direction. The final
figures were :
Lightest fraction vol. 0-2 c.c. . . 0-99974
Heaviest fraction vol. 0-3 c.c. . . 1-00023
Mercury behaves as though it was a mixture of equal parts
of two isotopes with atomic weights 202-0, 199-2 in equal
parts or of isotopes 201-3, 199-8 when the former is four times
as strong as the latter, and so on.
120. Separation of the isotopes of chlorine by free
evaporation. The same two investigators were able to
announce the first separation of the isotopes of chlorine 1
by applying the above method to a solution of HC1 in water.
This was allowed to evaporate at a temperature of 50 C.
and condense on a surface cooled in liquid air. Starting with
1 litre 8-6 mol. solution of HC1 100 c.c. each of the lightest
and heaviest fraction were obtained.
The degree of separation achieved was tested by two different
methods. In the first the density of a saturated solution of
Nad made from the distillate and the residue respectively
was determined with the following results :
Density (salt from distillate) = 1-20222
Density (salt from residue) = 1-20235
These figures correspond to a change in atomic weight of 0-024
of a unit.
In the second method exactly equal weights of the isotopic
NaCls were taken and each precipitated with accurately the
same volume of AgN0 3 solution, in slight excess. After pre-
cipitation and dilution to 2,000 c.c. the approximate concen-
tration of the filtrate was determined by titration, also the
ratio of Ag concentration of the two solutions was measured
in a concentration cell. Calculation showed that the difference
in atomic weight of the two samples was 0-021 in good agree-
ment with the density result.
121. Separation by Positive Rays. The only method
which seems to offer any hope of separating isotopes completely,
and so obtaining pure specimens of the constituents of a com-
1 Bronsted and Hevesy, Nature, July 14, 1921.
THE SEPARATION OF ISOTOPES 137
plex element, is by analysing a beam of positive rays and
trapping the particles so sorted out in different vessels. It is
therefore worth while inquiring into the quantities obtainable
by this means.
Taking the case of neon and using the parabola method of
analysis with long parabolic slits as collecting vessels we find
that the maximum separation of the parabolas corresponding
to masses 20 and 22 (obtained when electric deflexion 6 is
half the magnetic) is approximately
1 M,-M 2 B_
V2 M! 28'
Taking a reasonable value of 6 as -3 the maximum angular
width of the beam for complete separation = 0-01. If the
canal-ray tube is made in the form of a slit at 45 to axes,
i.e. parallel to the curves, the maximum angular length of
the beam might be say 5 times as great, which would collect
the positive rays contained in a solid angle of -0005 sq. radian.
The concentration of the discharge at the axis of the positive
ray bulb is considerable, and may be roughly estimated to
correspond to a uniform distribution of the entire current
over a J sq. radian. One may probably assume that half the
current is carried by the positive rays, and that at least half
the positive rays consist of the gases desired. If neon is
analysed by this method therefore the total current carried
by the positive rays of mass 20 is
0005 x4xjxjxi= -0005 i.
If i is as large as 5 milliamperes this = 1-5 X 10 4 E.S.U.
2-7 x lOx X 10-"
i.e. one might obtain about one-tenth of a cubic millimetre of
Ne 20 and 1/100 cubic millimetre of Ne 22 per 100 seconds run.
It is obvious that even if the difficulties of trapping the rays
were overcome, the quantities produced, under the most
favourable estimates, are hopelessly small.
122. Separation by photochemical methods. A re-
markably beautiful method of separating the isotopes of
138 ISOTOPES
chlorine has been suggested by Merton and Hartley which
depends upon the following photochemical considerations.
Light falling on a mixture of chlorine and hydrogen causes
these gases to combine to form hydrochloric acid. This must
be due to the activation of the atoms of hydrogen or those of
chlorine. Supposing it to be the latter it is conceivable that
the radiation frequency necessary to activate the atoms of
Cl 35 will not be quite the same as that necessary to activate
those of Cl 37 . Calling these frequencies v^ and v& respectively
it would seem possible, by excluding one of these frequencies
entirely from the activating beam, to cause only one type of
chlorine to combine and so to produce pure HCI 35 or HCI 37 .
Now ordinary chlorine contains about three times as much
Cl 35 as Cl 37 and these isotopes must absorb their own activat-
ing radiation selectively. In this gas therefore light of
frequency v& will be absorbed much more rapidly than that
of frequency *> 87 , so that if we allow the activating beam to
pass through the right amount of chlorine gas i> 35 might be
completely absorbed but sufficient v 31 radiation transmitted
to cause reaction. On certain theories of photo-chemistry
light containing v 37 but no v 35 would cause only atoms of
Cl 37 to combine so that a pure preparation of HCI 37 would
result. Pure Cl 37 made from this product could now
be used as a filter for the preparation of pure HCI 35 , and
this in its turn would yield pure Cl 35 which could then be
used as a more efficient filter for the formation of more
HCI 37 .
Had this very elegant scheme been possible in practice it
would have resulted in a separation of a very different order
to those previously described and the preparation of un-
limited quantities of pure isotopes of at least one complex
element. There is however little hope of this, for so far the
results of experiments on this method have been entirely
negative.
123. Other methods of separation and general con-
clusions. The following methods have also been suggested.
By the electron impact in a discharge tube, in the case of the
inert gases, the lighter atoms being more strongly urged towards
THE SEPARATION OF ISOTOPES 139
the anode ;* by the migration velocity of ions in gelatine ; a
by the action of light on metallic chlorides. 3
A survey of the separations actually achieved so far shows
that from the practical point of view they are very small.
In cases where the method can deal with fair quantities of
the substance the order of separation is small, while in the
case of complete separation (positive rays) the quantities
produced are quite insignificant. We can form some idea by
considering the quantity
Q = (difference in atomic weight achieved) X (average
quantity of two fractions produced in grammes). As regards
the first of these factors the highest figure so far was 0-13
obtained by the writer in the original diffusion experiments on
neon, but as the quantities produced were only a few milli-
grams Q is negligibly small. The highest values of Q have
been obtained by Bronsted and Hevesy by their evaporation
method. 4 It is 0-5 in the case of Hydrochloric Acid, 0-34 in
that of Mercury.
When we consider the enormous labour and difficulty of
obtaining this result it appears that unless new methods are
discovered the constants of chemical combination are not
likely to be seriously upset for some considerable time to come.
1 Skaupy, Zeitsch. Phys. y 3, 289, 460, 1920.
2 Lindemann, Proc. Roy. Soc., 99A, 104, 1921.
3 Renz, Zeit. Anorg. Chem., 116, 62, 1921.
4 V. p. 134.
APPENDIX I
Table of atomic weights and isotopes of the elements.
The elements are given in order of their atomic numbers. The
different periods are indicated by gaps after the inert gases.
A curious relation, pointed out by Rydberg, is that the
atomic numbers of all the inert gases are given by taking the
series 2 (I 2 + 2 2 + 2 2 -f 3 2 + 3 2 + 4 2 + . . .) and stopping the
summation at any term. This gives the numbers used by Langmuir
(p. 95).
The atomic weights given are the International ones except in
the cases marked with an asterisk, where the figures are taken from
some of the recent determinations given below.
The isotopes where known are given in order of their atomic
masses. The proportion of an isotope in a complex element is
indicated by the index letters a, 6, c ... in descending order.
In the case of isotopes of the radioactive elements 81-92 the roman
numeral gives the number of them believed to exist. The nomen-
clature of some of the rare earths 69-72 is not yet standardised.
The names here are those used by Moseley. Some of these elements,
though detected by their X-ray spectra, have never been isolated.
The elements corresponding to atomic numbers 43, 61, 75, 85, 87
(all odd) have not yet been discovered.
Recent atomic weight determinations. The following is a
list of some of the elements whose atomic weights have been re-
determined quite recently, together with references to the papers
in which they were published. Where more than one value is
given different methods were used :
Fluorine 19-001. Moles and Batuecas, Jour. Chim. Phys., 18, 353,
1920.
Aluminium 26-963. Richards and Krepelka, Journ. Am. Chem. Soc.,
42, 2221, 1920.
Silicon 28-111. Baxter, Weatherell and Holmes, ibid. , 42, 1 1 94, 1 920.
Scandium 45-10. Honigschmid, Zeit. Electrochem., 25, 93, 1919.
Tin 118-703. Baxter and Starkweather, Journ. Am. Chem. Soc., 42,
905, 1920.
118-699. Braun and Krepelka, ibid., 42, 917, 1920.
141
142
APPENDIX I
Tellurium 127-73, 127-79. Bruylants and Michielsen, Bull. Acad.
Belg., 119, 1919.
Samarium 150*43. Owens, Balke and Kremers, Journ. Am. Chem.
800., 42, 515, 1920.
Thulium 169-44, 169-66. James and Stewart, ibid., 42, 2022, 1920.
Bismuth 209-02. Honigschmid, Zeit. Electrochem., 26, 403, 1920.
208-9967. Classen and Wey, Ber., 53, 2267, 1920.
Antimony 121-773. Willard and McAlpine, Journ. Am. Chem. Soc., 43,
797, 1921.
Lanthanum 138-912. Baxter, Tani and Chapin, Journ. Am. Chem.
Soc., 43, 1085, 1921.
Germanium 72-418. Miller, Journ. Am. Chem. Soc., 43, 1085, 1921.
Zinc 65-38. Baxter and Hodges, ibid., 43, 1242, 1921.
Cadmium 112-411. Baxter and Wilson, ibid., 43, 1230, 1921.
Element.
1
I
If
It
Number of
Isotopes.
Masses of isotopes.
. 'C ** Hydrogen
< *o Helium .
H
He
1
2
1-008
4-00
1
1
1-008
4
oo Lithium .
Li
3
6-94
2
6* 7"
1
+1 Beryllium
Be
4
9-1
1
9
^ Boron
B
5
10-9
2
10* 11"
.2 Carbon . . .
C
6
12-00
1
12
8 Nitrogen .
N
7
14-008
1
14
~ Oxygen . . .
8
16-80
1
16
S Fluorine .
F
9
19-00
1
19
<
" Neon. . . .
Ne
10
20-20
2
20 22*
j
jo Sodium .
Na
11
23-00
1
23
H Magnesium .
Mg
12
24-32*
3
24 a 25* 26 C
? Aluminium .
Al
13
26-96*
o Silicon
Si
14
28-3
2
28" 29* (30)
Phosphorus .
P
15
31-04
1
31
^ Sulphur .
S
16
32-06
1
32
"2 Chlorine .
Cl
17
35-46
2
35 37* (39)
Argon
A
18
39-9
2
36* 40"
Potassium
K
19
39-10
2
39 41*
Calcium .
Ca
20
40-07
(2)
40 (44)
Scandium
So
21
45-1*
Titanium .
Ti
22
48-1
Vanadium
V
23
51-0
co Chromium
Cr
24
52-0
^ Manganese .
Mn
25
54-93
Iron ....
Fe
26
55-84
'g Cobalt . . .
Co
27
58-97
rA *
3 Nickel . . .
Ni
28
58-68
2
58" 60*
& Copper .
A Zinc ....
Cu
Zn
29
30
63-57
65-37
(4)
(64 66* 68< 70")
*
t< Gallium .
Ga
31
70-10
y'"l
Germanium .
Ge
32
72-5
Arsenic .
As
33
74-96
1
75
Selenium .
Se
34
79-2
Bromine .
Br
35
79-92
2
79 81*
1 Krypton . . . ' Kr
36
82-92
6
78> 80 82 C 83 d 84 a 86*
APPENDIX I
143
Jg
|l
O 40
oj
si
Element
If
,0
BO
Masses of isotopes.
02
^fc
5
~
f Rubidium
Rb
37
85-45
2
85 a 87 6
Strontium
Sr
38
87-63
Yttrium .
Y
39
89-33
Zirconium
Zr
40
90-6
Niobium .
Nb
41
93-1
oo Molybdenum
Mo
42
96-0
*H
43
Ruthenium .
Ru
44
101-7
'o Rhodium.
Rh
45
102-9
*S Palladium .
Pd
46
106-7
PM Silver ....
Ag
47
107-88
A Cadmium
Cd
48
112-40
K3 Indium .
In
49
114-8
Tin ....
Sn
50
118-7
Antimony
Sb
51
120-2
Tellurium
Te
52
127-5
Iodine
I
53
126-92
1
127
Xenon
X
54
130-2
(7)5
(128) 129" (130) 131 C 132*
134* 136"
Caesium .
Cs
55
132-81
1
133
Barium .
Ba
56
137-37
Lanthanum .
La
57
139-0
Cerium .
Ce
58
140-25
Praseodymium .
Pr
59
140-6
Neodymium .
Nd
60
144-3
61
Samarium
Sm
62
150-4
Europium
Eu
63
152-0
Gadolinium .
Gd
64
157-3
Terbium .
Tb
65
159-2
_
Dysprosium .
Ds
66
162-5
Holmium
Ho
67
163-5
,_ Erbium .
Er
68
167-7
Thulium . . .
Tu
69
168-5
1 Ytterbium . .
Yb
70
173-5
'C Lutecuim
Lu
71
175
PM (Keltium) . .
(Kt)
72
rS Tantalum
Ta
73
181-5
<> Tungsten.
W
74
184-0
75
Osmium . .
Os
76
190-9
Iridium .
Ir
77
193-1
Platinum .
Pt
78
195-2
Gold ....
Au
79
197-2
Mercury .
Hg
80
200-6
(6)
(197-200) 202 204
Thallium . . .
Tl
81
204-0
IV
. Lead ....
Pb
82
207-2
XI
Bismuth .
Bi
83
209-0*
V
Polonium
Po
84
VII
85
1
Emanation
Em
86
222-0
III
-
87
.g Radium . . .
Ra
88
226-0
IV
Actinium.
Ac
89
II
~J Thorium .
Th
90
232-15
VI
Uranium X .
UX
91
II
I Uranium
Ur
92
238-2
II
APPENDIX II
The Periodic Table of the Elements. The atomic numbers are given h;
bold type, the atomic weights in italics and the isotopes, where known, iijj
ordinary numerals. The roman numerals indicate the chemical groups anc^
the most important associated valencies are given below them. Element!,
are placed to the left or to the right of the columns according to their chemicaj
properties, those in the same vertical line as each other have strong chemicaj
similarities. The Rare Earth group is surrounded by a thick line. Element ,
59-72 have no properties pronounced enough to give them definite place"
in the table. The properties of the missing elements can be predicted wit]
PERIODIC TABLE
1H
1-008
Valency
II
+ 2
III
+ 3
IV
+ 4
2 He
4-00
4
10 Ne
20-2
20, 22
3 Li
6-94
6. 7
11 Na
23-00
4 Be
9-1
12 Mg
24-32
24, 25, 26
5B
10-9
10, 11
13 Al
26-96
6C
12-00
12
14 Si
28-3
28,29
ISA
39-9
36, 40
19 K
39-1
39, 41
20 Ca
40-07
21 Sc
45-1
29 Cu
63-57
30 Zn
65-37
38 Kr
82-92
78, 80, 82, 83,
84, 86
37 Rb
85-45
85, 87
38 Sr
87-83
39 Y
89-33
47,
107-
48 Cd
112-40
22 Ti
48-1
31 G
70-1
32 Ge ,
72-5
40 Zr
90-6
49 In
114-8
50 Sn
118-7
54 Xe
130-2
129, 131, 132,
134, 136
55 Cs
132-81
133
56 Ba
137-37
57 La
139-0
58 Ce
140-25
59 Pr 60Nd 61 62 Sm 63 Eu 64 Gd 65 Tb
140-6 144-3 150-4 152-0 157-3 159-2
66 Ds 67 Ho 68 Ev 69 Tu 70 Yb 71 Lu 72 (Kt)
162-5 163'5 1677 t 168-5 173-5 175
79 Au
197-2
80 Hg
200-6
197-204
81 Tl
204-0
82 Pb
207-2
86 Em
222-0
87-
88 Ra
226-0
89 Ac
90 Th
232-15
144
APPENDIX II
nsiderable certainty from the positions of their atomic numbers. From
point of view of the construction of the atom the inert gases should mark
9 end of the periods as they are shown to do in the list of atomic weights
Appendix I, on the other hand it is more usual in chemistry to start with
lency 0. From principles of general convenience of arrangement the
ter plan is adopted in this table, which is intended to give the maximum
lount of chemical information. Hydrogen, which belongs equally well
group I or group VII, is best omitted from the table altogether.
HE ELEMENTS
V
3
VI
2
VII
1
VIII
7N
14-01
14
80
16-00
16
9F
19-00
19
15 P
31-04
31
163
32-06
32
17 Cl
35-46
35, 37
23V
51-0
33 As
74-96
75
24 Or
52-0
34 Se
79-2
25 Mn
54-93
35 Br
79-92
79, 81
26 Fe 27 Co 28 Ni
55-85 58-97 58-68
58, 60
41 Nb
93-5
51 Sb
120-2
42 Mo
96-0
52 Te
127-5
43
531
126-92
127
44 Ru 45 Rh 46 Pd
101-7 102-9 106-7
73 Ta
157-5
83 Bi
209-0
74 W
184-0
84 Po
75
85
76 Os 77 Ir 78 Pt
190-9 193-1 195-2
91 U X ii
92 U
238-2
145
APPENDIX III
Recent results obtained by Dempster. Thanks to a private
communication the writer is able to include some further results
obtained by Dempster and a diagram of his apparatus for obtaining
FIG. 19. Diagram of Anode in Dempster's latest apparatus.
positive rays from metals. A full account is to appear in the
Physical Review. Fig. 19 shows the new arrangement of
vaporising furnace A and ionising filament C. The analysing
apparatus has already been described on p. 31 and the results with
5-9
6-0
- 6-1
6-9 7-0
Atomic Weight.
FIG. 20. Curve for Lithium.
146
APPENDIX III
147
magnesium on p. 81. Fig. 20 shows one of the curves obtained
with lithium. It will be seen that the relative intensities of the
isotopes is entirely different from that found by the writer (p. 86)
and also disagrees very definitely with the chemical atomic weight.
Dempster describes these relative intensities as varying very
considerably. This is a most remarkable phenomenon and further
information upon it is very desirable. There seems just a possibility
that the 6 line is enhanced by doubly charged carbon but it is not
easy to see where such particles could be produced.
VoltS
928 913-5 899-5 886 873 860 847-5
62 63
64 65 66 67
Atomic Weight.
FIG. 21. Curve for Zinc.
68 69
Fig. 21 gives a remarkable curve obtained from zinc. This
indicates three strong isotopes and a faint fourth. The absolute
scale of atomic weight is not known with certainty, and the values
63, 65, 67, 69 are given by Dempster as those in best agreement
with the atomic weight 65-37. Considering that the error hi the
148 APPENDIX III
mean atomic weight of lithium, when calculated on these lines ,
is about 5 per cent, it would appear possible that these might be a
unit too high or too low. The probability of this is strengthened
very much by the rule given on p. 110 connecting even atomic
number with even atomic weight.
Results with calcium show only one line. This makes it extremely
probable that this is a simple element of atomic weight 40 and
therefore an isobare of argon. 1
Note. In a still later communication Dempster states that he
has been successful in using an anode of calcium to which a small
quantity of zinc had been added. By this means he is able to
compare the masses of the zinc isotopes with the strong calcium
maximum, assumed as 40. This gives the atomic weights as 64,
66, 68 and 70. The intensities are quite different to those in the
curve given above for zinc. 64 is now the strongest, 66 and 68
fainter, while 70 is very faint indeed. No explanation is yet
advanced for these remarkable irregularities in relative intensity.
He has also observed a small maximum at 44 invariably accom-
panying the strong calcium maximum 40. This he considers to be
probably due to an isotope of that element present in small quantity
as suggested by the atomic weight 40 - 07.
The above values are included provisionally in the tables on
pages 89 and 142.
1 F. p. 88.
INDEX
Abnormal hydrides, 98
Abundance of the elements, 111
Accuracy of mass-spectrograph, 60
Actinium chain, 14, 15
Additive law of mass, 99
Alkali metals, mass-spectra of, 83
Alpha ray changes, 13
Analysis of the elements, 63
ANDRADE and RUTHERFORD, 11
Anode, composite, 80, 86
hot, 80, 83, 84
Anticathode, silica, 48
Antimony, 78
Argon, 66
ARONBERG, 123
and HARKINS, 124
Atmolysis, separation by, 127
Atomic number, 13, 93
theory, 2
volume of isotopes, 18
weights, tables of, 89, 141
weights of radio -elements, 13,
141 .
Atoms, structure of, 90
BALKE, OWENS and KREMERS, 142
BARKLA, 93
BATTTECAS and MOLES, 141
BAXTER and HODGES, 142
and PARSONS, 113
and STARKWEATHER, 141
and WILSON, 142
TANI and CHAPIN, 142
WEATHERELL and HOLMES,
73, 142
Beryllium, 88
Beta ray change, 13
BOHR, 94, 95, 121, 122, 123
,, atom, 95
BOLTWOOD, 1, 7
Boron, 72
anomalous atomic weight ot,
114
trifluoride, 73
Bracketing, method of, 59, 69
BRAUN and KREPELKA, 141
BROEK, VAN DEN, 93, 94, 116
Bromine, 76
BRONSTED and HEVESY, 135, 136, 139
BROSSLERA, 102, 104
BRUYLANTS and MICHIELSON, 142
Caesium, 87
anomalous atomic weight
of, 114
Calcium, 88, 148
Calibration curve, 55
Camera of mass-spectrograph, 51
positive ray, 26
Canalstrahlen, 22
Carbon, 63
Carnotite, lead from, 124
Cathode rays, 22, 24
CHADWICK, 94
and RUTHERFORD, 103
CHAPIN, BAXTER and TANI, 142
CHAPMAN, 130
and DOOTSON, 130
Chemical action, separation by, 133
law of radioactive change,
11
Chlorine, 65, 113
separation of the isotopes
of, 136
CLASSEN, 31
and WEY, 142
CLAUDE, 35
Cleveite, lead from, 17
Coincidence, method of, 57
Composite anode, 80, 86
Constancy of chemical atomic weights,
22
Cosmical effect of change of mass, 103
CROOKES, 3, 4, 24, 115, 117
dark space, 24, 35
theory of the evolution of
elements, 117
CURIE, MLLE. I., 113
,, M., 18
Dalton's hypothesis, 2
DARWIN, 15
DAVIES and HORTON, 68
Deflection of positive rays, 27
DEMPSTER, 31, 80, 81, 86, 114, 146
149
150
INDEX
Dempster's method of analysis, 31,146
Density balance, 35
of isotopic leads, 17, 18
Diffusion of neon, 39
separation by, 127
velocity, determination of,
20
Disintegration theory of the evolu-
tion of elements, 116
Distillation of neon, 37
Distribution of lines on mass-
spectrum, 54
DOOTSON and CHAPMAN, 130
Du Bois magnet, 51
EDDINGTON, 104
Einstein's theory of relativity, 103
Electrical theory of matter, 90
Electric discharge in gases, 23
,, field of mass-spectrograph,
50
Electricity as an element, 115
Electrochemical properties of isotopes,
10
Electron, the, 91
Element, meaning of the word, 115
ENSKOG, 130
EPSTEIN, 95
EXNEB and HASCHEK, 121
FA JANS, 11
First order lines, 61
FLECK, 12
Fluorine, 72, 97
Focussing positive rays, 44
FOWLER, 123
and ASTON, 45
Fractional distillation, separation by,
133
FRANCE: and KNIPPING, 68
GEHRCKE, 102
and REICHENHEIM, 80, 83,
88
GEIGER and NUTTALL, 10, 13
GOLDSTEIN, 22
Gravitation effect on spectra, 121
separation by, 131
GROH and HEVESY, 20, 135
HAHN, 8
and MEITNER, 8
Halation effect, 60
Half-tone plates, 25
HALL and HARKINS, 116
HARKINS, 102, 111, 116, 129
and ARONBERG, 124
and HALL, 116
and WILSON, 116
HASCHEK and EXNER, 121
Helium, 67, 69, 106
HEVESY, 10, 12, 19
,, and BRONSTED, 135, 136,
139
and GROH, 20, 135
and PANETH, 11
and ZECHMEISTER, 20
HODGES and BAXTER, 142
HOLMES, BAXTER and WEATHERELL,
73, 141
HONIGSCHMID, 17, 18, 141, 142
and HOROVITZ, 18,
121
HOROVITZ and HONIGSCHMID, 18, 121
HORTON and DAVIES, 68
Hot anode, 80, 83, 84
Hydrochloric acid, diffusion of, 129
Hydrogen, 67, 69, 106
HYMAN and SODDY, 17, 121
IBBS, 130
IMES, 125, 126
Indicators, radioactive, 19
Infra-red spectrum of isotopes, 125
Intensity of positive rays, 44
Iodine, 78
Ionic dissociation theory, proof of, 20
lonisation in discharge tube, 24
Ionium, 1, 7, 9, 18
atomic weight of, 18
Isobares, 12, 13, 97, 110
Isotopes, definition of, 12
,, diagrams of, 97
,, discovery of, 5
,, melting point- of, 18
,, refractive index of, 18
,, separation of, 127
,, solubility of, 18
table of, 89, 141
JAMES and STEWART, 142
JOLY and POOLE, 133
KEETMAN, 7
Kernel of atom, 98
KIRCHOFF, 116
KNIPPING and FRANCK, 68
KOHLWEILER, 116
KRATZER, 126
KREMERS, OWENS and BALKE, 142
KREPELKA and BRAUN, 141
and RICHARDS, 141
Krypton, 70
anomalous atomic weight
of, 114
LANDAUER and WENDT, 70
LANGMTJIR, 95, 96, 99
Lead, atomic weight of, 16
from carnotite, 124
,, from thorite, 17
,, isotopes of, 14, 15
INDEX
151
LEMBEBT and RICHARDS, 17, 121
Lewis -Langmuir atom, 95
LINDEMANN, 102, 124, 134, 139
and ASTON, 131
Lines of first and second order, 61, 75
of reference, 55, 64
Lithium, 86, 97, 146
LOOMIS, 125, 126
LUDLAM, 129
Me ALPINE and WILLARD, 142
Magnesium, 80
Magnetic field of mass-spectrograph,
51
MARCKWALD, 7, 8
Mass, change of, 100
deduced from parabolas, 28
deduced from mass -spectrum,
55
Mass-spectrograph, 43
Mass -spectrum, 47, 54
Measurement of lines on mass-
spectrum, 59
MEITNER, 21
and HAHN, 8
Melting point of isotopes, 18
Mercury, 72, 80
,, parabolas of, 30
separation of the isotopes
of, 134
MERTON, 121, 123, 124, 125
Mesothorium, 8, 10
Meta-elements, 4
Metallic elements, mass -spectra of, 80
Meteoric nickel, 113
MICHIELSON and BRUYLANTS, 142
Microbalance for density, 35
MILLIKAN, 22, 91
Molecular lines of second order, 75
MOLES and BATUECAS, 141
MOSELEY, 11, 93, 115
MULLER, 142
Multiply charged rays, 30
Natural numbers and atomic weights,
111
Negatively charged rays, 29, 62
Negative mass-spectra, 62, 66
Neon, 1, 33, 64, 97
NEUBERGER, 21
Nickel, 79
,, meteoric, 113
Nitrogen, 67, 110
Nomenclature of isotopes, 61
Nucleus atom, 10, 92, 97, 125
,, structure of, 101
NUTTALL and GEIGER, 10, 13
Order, lines of first and second, 61
OWENS, BALKE and KREMERS, 142
Oxygen, 63
Packing effect, 100
PANETH and HEVESY, 11
Parabola method of analysis, 25
PARSONS and BAXTER, 113
Perforated electrodes, 22, 24
Periodic law, 11, 12, 34
table of the elements, 144,
145
Period of radio -elements, 13
PERRIN, 104
Phosphorus, 77
Photochemical separation, 137
Photographic plates for positive rays,
25
Planck's quantum, 95
Planetary electrons, 92
POOLE, 133
,, and JOLY, 133
Positive ray parabolas, 28
rays, 22
separation by, 136
Potassium, 87
Pressure diffusion, 131
Proton, the, 92
Protyle, 90, 118
Prout's hypothesis, 2, 90, 100
Radioactive isotopes, 7, 14
classification of,
21
transformations, 13, 14,
15
Radium B and lead, 1 1
D and lead, 11
RAMSAY, 115
and COLLIE, 39
and TRAVERS, 33
RATNER, 24
RAYLEIGH, 127
Reference lines, 55, 64
Refractive index of isotopes, 18
REICHENHEIM and GEHRCKE, 80, 83,
88
RENZ, 139
Resolving power of mass-spectro-
graph, 60
RICHARDS 17
and KREPELKA, 141
and LEMBERT, 17, 121
and WADSWORTH, 17
RICHARDSON, 85
Rossi and RUSSELL, 9, 120
Rubidium, 87
RUSSELL, 11
and Rossi, 9, 120
RUTHERFORD, SIR E., 7, 9, 13, 92, 93,
102
and CHADWICK, 103
and ANDRADE, 11
RYDBERG, 141
152
INDEX
SCHUTZENBERGER, 3
Screens, willemite, 25
Secondary rays, 29
Second order, lines of the, 61
Selenium, 77
Separation of isotopes, 127
Silicon, 72
fluoride, 74
SKAUPY, 139
Slit system of mass-spectrograph, 49
SMITH and VAN HAAGEN, 72
SODDY, 6, 8, 10, 11, 12, 13, 14, 16, 17,
35
and HYMAN, 17, 121
Sodium, 86
Solubility of isotopes, 18
SOMMERFELD, 95
Spectra of isotopes, 9, 121,
Spectrum lines, form of, 53
SPENCER, 91
STARKWEATHER and BAXTER, 141
STAS, 91
Statistical relation of isotopes, 109
STEWART, 11, 12
and JAMES, 142
Sulphur, 76
TANI, BAXTER and CHAPIN, 142
Tellurium, 77
Thermal diffusion, 129
Third order line of argon, 67
lines of, 61
THOMSON, G. P., 86, 88
SIR J. J., 1, 22, 29, 33, 62,
70, 72, 75, 84, 91, 129
Thorite, 17, 18
Thorium, 7, 9, 14, 15, 18, 120
Thorium, chain, 17, 18, 116
,, atomic weight of, 18
Tin, 78
TRAVERS, 39
and RAMSAY, 33
Triatomic hydrogen, 70
Unitary theory of matter, 90
Uranium, 10, 120
chain, 15
Valency electrons, 98
VAN HAAGEN and SMITH, 72
WADSWORTH and RICHARDS, 17
WATSON, 33
and ASTON, 24, 35
WEATHERELL, BAXTER and HOLMES,
73, 141
WELSBACH, 8
WENDT and LANDATJER, 70
WEY and CLASSEN, 142
Whole number rule, 90
WIEN, 22
WILLARD and MCALPINE, 142
Willemite screens, 25
WILSON and BAXTER, 142
and HARKINS, 116
Xenon, 70
anomalous atomic weight of,
114
X-ray spectra of isotopes, 1 1
ZECHMEISTER and HEVESY, 20
Zinc, 147
Printed in Great Britain by Butler & Tanner, Frome and London.
UNIVERSITY OF CALIFORNIA LIBRARY
BERKELEY
Return to desk from which borrowed.
This book is DUE on the last date stamped below.
27Mar'50CK
FEB 7 195fc
JAN 2 2 1955 Uf
I3Way'57LS
Rt LD
r
,337
00 j- 2 5 1957
UBRARY USE
AU6 3 1960
Lrr21-100m-ll,'49 (B
LD
MARi3'64. 10
SEP 2 3 1976 3
PC.C1R. W d 76
2003
49(B7146sl6)476
LJ I
