HISTORICAL INTRODUCTION TO
CHEMISTRY

MACMILLAN AND CO., LIMITED

LONDON . BOMBAY . CALCUTTA
MELBOURNE

THE MACMILLAN COMPANY

NEW YORK . BOSTON . CHICAGO
DALLAS . SAN FRANCISCO

THE MACMILLAN CO. OF CANADA, LTD

TORONTO

•HISTORICAL

INTRODUCTION TO

CHEMISTRY

T. M. LOWRY, D.Sc., F.R.S.

MACMILLAN AND CO., LIMITED
ST. MARTIN'S STREET, LONDON

1915

\D

COPYRIGHT

PREFACE

IN the preparation of this volume the purpose has been
to present an historical account of the more important facts
and theories of chemistry, as these disclosed themselves to
the original workers in this branch of science. No attempt
has been made to write a formal History of Chemistry,
either as a survey of the various periods into which the
history of the science may be divided, or in the more usual
biographical form. The material has been classified by
subjects rather than by authors ; but it will be found that
under this system the work of individual experimenters is
described quite as fully as in a biographical survey, whilst
in the case of certain chemists, such as Priestley, Lavoisier
and Gay-Lussac, it has been possible to include detailed
descriptions of experimental work which could scarcely have
found a place in a brief biography.

The Biographical Index provides a key to the work of
each author as it is described in the text, and contains most
of the essential items for an account in narrative form of
the achievements of the great pioneers of chemistry.

In a few cases this index contains dates and titles of works
not included in the text, as, for example, Cavendish's work
on the density of the earth, and some of Faraday's physical
experiments ; but no attempt has been made either in the
biographical index or in the text to record the later and
more detailed developments of organic or of physical chem-
istry. Such a record would be out of place in a historic?^

35926

vi PREFACE

introduction dealing with fundamental facts and problems ;
only the most incidental references will be found therefore to
the work of Hofmann in organic chemistry, or of van t'Hoff
in the various branches of physical chemistry. This limita-
tion is of little importance, as the advanced student will
find ample descriptions of the achievements and discoveries
of these later workers in the two volumes of Memorial
Lectures issued by the Chemical Society.

In compiling the present volume, the standard histories
of Thomson and of Kopp have been invaluable as guides to
the literature, but the whole story has been written afresh
from the original sources. Almost without exception, every
reference and quotation has been checked directly in the
printed proofs against the original text ; if, however, any
errors should have survived, the author would be very grate-
ful to anyone who would direct his attention to them. The
few statements for which dates but no references are given
are made on the -authority of Kopp, but most of them refer
only to incidental points. Much of the narrative, even when
not enclosed between inverted commas, is in the actual
words of the original descriptions, a feature which the pre-
sent volume shares with the text-books of 100 years ago, but
which has gradually disappeared as the early history of
chemistry has become more and more a " twice-told tale."

In order to present the material in the most accessible form,
quotations have been taken so far as possible from modem
reprints, such as those of Scheele's Essays, or the Alembic
Club Reprints, or from collected Works such as those of
Lavoisier, Davy, and Stas. To guard against anachronism,
full use has been made of contemporary translations of such
works as Bergman's Essays and Berthollet's Chemical Statics ;
but as the object in view was to present a picture of the
development of chemistry rather than a formal history, it
was thought better not to introduce unnecessary confusion
by using the name " nitric acid " when nitrogen peroxide
was meant, nor to obtrude at every point the various names

PREFACE vii

by which oxygen and chlorine were described up to 1787
and 1 8 10 respectively; but whenever some more familiar
term has been introduced into the text, the alteration has
been indicated by square brackets.

The detailed descriptions of classical experiments, which
form a leading feature of the book, should be of value not
only to the student but also to the teacher of chemistry, since
they show not merely how the great facts of chemistry might
have been discovered, but the actual course of the discover-
ies themselves. Such information has proved of real value
in devising courses of instruction in elementary chemistry,
and forms a sure guarantee against an incorrect or illogical
sequence ; from this point of view the book may be com-
mended to those who are responsible for the training of
teachers, as well as to students who intend to become
teachers themselves. The historical method has also been
found to provide a complete solution of the difficult problem
of teaching mixed classes of students, some absolute begin-
ners and others with a considerable knowledge of elementary
text-books of chemistry. This problem is insistent in
training colleges and in medical schools, and is probably
but little less urgent in other departments of teaching. The
material included in the present volume has been proved,
by several years of actual practice, to provide a means of
interesting and instructing both types of students. Even
the laws of chemical combination acquire a new interest
when presented in the picturesque imagery of Proust, or
with some of the glow of Berzelius's early enthusiasm.

In contrast with most of the well-known histories of
chemistry, the volume is provided with many illustrations,
especially of the apparatus used by the earlier workers. It
is unfortunate that large-scale copper-plate illustrations can-
not be reduced in exact facsimile without destroying all
their technical merits ; but great care has been taken that
the wood-blocks shall reproduce as faithfully as possible all
the essential features of the original drawings. Thus,

viii PREFACE

Bunsen's burner (1866) has been removed, it is hoped finally,
from Dumas's apparatus for the composition of water (1841),
and the big spirit-lamp has been restored to its place ; and
Lavoisier's red-hot gun-barrel is again sealed with clay joints
instead of with rubber. The temptation to reconstruct early
apparatus (which has led to the association of the Bunsen
burner with Lavoisier's work) has been resisted even in the
pressing case of Cavendish's experiments on the composition
of water, where the gap has been filled by reproducing the
contemporary apparatus of Monge ; only when the text and
illustrations were complete was the discovery made that two
of the globes used in these experiments are still preserved in
the library of the Royal Institution. The single case in
which a figure has been consciously modified is a small
alteration in Priestley's blackboard and table (Fig. 19 b}
with the view of making the most of the limited space avail-
able for reproduction.

An exact historical narrative, such as is here presented,
could not have been written without free access to books and
journals, many of which are rare and some almost inacces-
sible. I wish to express my indebtedness to Mr. A. H. White,
of the Royal Society's Library, and to Mr. F. W. Clifford,
the Librarian of the Chemical Society, for their invariable
courtesy and helpfulness over a period of several years, as
well as to the Institution of Electrical Engineers for access to
their important collection of Volta's papers. I am indebted
to the late Miss Freund for a quotation from a copy of
Wenzel's Theory of Affinity in the library of the University
of Bonn, to Prof. Victor Henri of Paris for some information
in reference to early French publications, and to Prof. Ernst
Cohen of Utrecht for a number of historical details. The
early chapters of the book were written in collaboration
with the late Mr. G. C. Donington, whose double qualifica-
tion in History and Natural Science gave special value to
his opinions and criticisms. I have also derived great
benefit from the expert advice which Prof. R. A. Gregory

PREFACE ix

and Mr. A. T. Simmons have generously given to me
through the whole period occupied by the compilation of
the book. The proof-sheets have been read by Mr. W. A.
Davis, by Dr. Merriman, by Lieut. Victor Steele and by
Mr. H. S. Patterson, to whom I am grateful for much
valuable help and criticism.

T. M. LOWRY.
LONDON :
August, 1915.

CONTENTS

CHAPTER I

KAW MATERIALS AND PRIMITIVE MANUFACTURES . I

CHAPTER II

THE ACIDS 12

CHAPTER III

THE BURNING OF METALS AND THE DISCOVERY OF

OXYGEN .'.... . . . . • 26

CHAPTER IV

CHALK, LIME, AND THE ALKALIS 48

CHAPTER V

THE STUDY OF GASES . . . . , : . . . . 64
CHAPTER VI

THE COMPOSITION OF FIXED AIR. CARBON, CARBONIC

ACID, AND THE CARBONATES 94

CHAPTER VII

THE BURNING OF INFLAMMABLE AIR, AND THE COM-
POSITION OF WATER . • . . . . . . 112

CHAPTER VIII

THE BURNING OF INFLAMMABLE GASES, LIQUIDS, AND

SOLIDS 137

CHAPTER IX

SULPHUR AND PHOSPHORUS 162

xii CONTENTS

PAGE

CHAPTER X

NITRE, NITRIC ACID, AND NITROGEN . . . . 1 86

CHAPTER XI

MURIATIC ACID AND CHLORINE 2IO

CHAPTER XII

THE HALOGENS , . 233

CHAPTER XIII

THE DECOMPOSITION OF THE ALKALIS .... 257
CHAPTER XIV

THE ATOMIC THEORY 2QI

CHAPTER XV

THE MOLECULAR THEORY 320

CHAPTER XVI

ATOMIC WEIGHTS OF THE METALS 360

CHAPTER XVII

MOLECULAR ARCHITECTURE . . . . . .383

CHAPTER XVIII

THE CLASSIFICATION OF THE ELEMENTS .... 448

CHAPTER XIX

BALANCED ACTIONS 498

CHAPTER XX
DISSOCIATION 514

BIOGRAPHICAL INDEX . 539

SUBJECT INDEX 565

LIST OF ILLUSTRATIONS

t IG. PAGE

1. Cubic Crystal of Salt 3

2. Crystal of Saltpetre 4

3. Sal-ammoniac ...... . . . 4

4. The Colenso Diamond ....... 5

5. Crystals of Quartz '...". . 6

6. Arrowhead Crystals of Marcasite . . * ... 8

7. Nodule of Marcasite as described by Glauber . . 9

8. Striated Cubes of Pyrites 9

9. Crystal of Blue Vitriol or Sulphate of Copper . 1 7
io. Large 'Crystal of Gypsum or Selenite (Sulphate of

Lime) 19

11-14. Mayow's Apparatus . . . . . . . 33

15. Lavoisier's Apparatus for Calcining Lead and Tin in

Air over Water or Mercury ..... 34

16. Lavoisier's Apparatus for Heating Mercury in a

Limited Volume of Air 41

17. Cavendish's Apparatus for finding the Weight and

Density of Gases . . . . 68

1 8. Apparatus Used by Cavendish to Prepare a Gas

which " Lost its Elasticity by Contact with

Water" . . . 82

19 (a). Hales's Apparatus for Measuring the Volume of
Gas set free by Heating Animal, Vegetable, and

Mineral Substances . 85

19 (£). Hales's Apparatus (Improved Design) ... 85

20. Priestley's Apparatus (First Plate) .... 87

xiv LIST OF ILLUSTRATIONS

FIG. PAGB

20. Priestley's Apparatus (Second Plate) .... 89

21. Lavoisier's Apparatus for Collecting the Gas produced

by Heating Red-lead and Charcoal in an Iron
Retort 96

22. Lavoisier's large Burning Glass . . ... 100

23. Crystals of Calc-Spar . . . . . . .105

24. Stalactite, /.<?., Chalk deposited by the escape of

Fixed Air from Dripping Water .... 106

25. Lavoisier's Apparatus for decomposing Steam in a

red-hot gun-barrel 119

26. Monge's Apparatus for exploding Hydrogen and

Oxygen in an exhausted globe .* . . . 120

27. Volta's Eudiometer 123

28. Dumas's Apparatus for determining the Composition

of Water 126

29. Morley's Apparatus for measuring the Density of

Oxygen . • . 129

30. Morley's Palladium-tube for weighing Hydrogen . 130

31. Morley's Combustion-tube . . . .... 131

32. Morley's Apparatus for the Combustion of Hydrogen

and Oxygen . . 132

33. Lavoisier's Apparatus for the Combustion of Spirit

of Wine 147

34. Dumas and Stas's Apparatus for the Combustion of

Graphite and of Diamonds 149

35. Stas's Apparatus for the Combustion of Carbonic

Oxide 152

36. Rhombic Crystals of Sulphur 1.63

37. Stoppered flask used by Priestley in preparing

"Vitriolic Acid Air" . . . . . . . 165

38. Large Cube of Galena 173

39. Cavendish's Apparatus for Sparking Air over

Mercury 189

40. Gray's Apparatus for determining the Composition

of Nitric Oxide . . . - 207

LIST OF ILLUSTRATIONS xv

PAGE

Twinned Cubes of Fluor Spar 234

Moissan's Apparatus for Preparing Fluorine . . 252

The Voltaic Pile . . , 271

The Voltaic Battery or " Crown of Cups " . . . 272
Boyle's Pneumatic Engine or Air Pump . . .321
Tube used in Boyle's Experiments on the Condensa-
tion of the Air 325

Gay-Lussac's Apparatus for Measuring the Expan-
sion of Gases and Vapours . . . . . 328
Gay-Lussac's Apparatus for Comparing the Expan-
sion of Air with that of Soluble Gases and Vapours 329
Gay-Lussac and Thenard's Apparatus for the Com-
bustion of Organic Compounds . . . . 391
Berzelius's Apparatus for the Combustion of Organic

Compounds . 392

Liebig's Apparatus for the Combustion of Organic

Compounds .. . .••- ; . . . 395
Dewar's Atomic Heat Curve, with a Curve of Atomic

Volumes . 472-473

High-frequency Spectra 488

High-frequency Spectra of the Elements . . . 489

Deville's " Hot-cold " Tube 517

Pebal's Apparatus for Proving the Dissociation of

Sal-ammoniac 521

H. B. Baker's Apparatus for Heating Dried Mixtures

of Hydrogen and Oxygen . . . . 532

HISTORICAL INTRODUCTION
TO CHEMISTRY

PART I

ELEMENTS AND COMPOUNDS

CHAPTER I

RAW MATERIALS AND PRIMITIVE MANUFACTURES

THE study of Chemistry as a branch of natural knowledge
may be said to begin with the work of the Honourable
Robert Boyle (1627-1691). But, at the time when Boyle
commenced his work, the most important. chemical processes
and many well-defined substances were already familiar,
many of them having been used for practical purposes
from very early times. Early chemical discoveries group
themselves naturally into three periods :

1. The Prehistoric and Ancient Period culminated in
the civilisations of Egypt, Greece, and Rome. During this
period many of the raw materials, which it is the business
of Chemistry to study, were collected, purified, and brought
into common use in daily life ; but only in a few cases were
processes discovered for the preparation of new substances
by the action of these raw materials on one another.
Evidence as to the substances which were known during this
period is derived mainly from casual references in the writings
of ancient authors rather than from systematic works of a
scientific character.

2. The Earlier Alchemistic Period extended from the
early part of the Christian era to about 1500 A.D. During

2 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

this second period the search for the PHILOSOPHER'S STONE,
a substance by which baser metals could be converted
into gold, led to an exhaustive study of all available
materials. Their actions upon one another were also studied,
and the effect of heat upon them both separately and in
mixtures of varying complexity. In this way, many new pro-
cesses and compounds were discovered. The physician
Geber, who lived in Spain at the close of the eighth century,
was perhaps the greatest of the alchemists ; the writings
that have been attributed to him afford a clear picture
of the progress which the science had made in the hands of
the earlier alchemistic workers.

3. The Later Alchemistic Period extended roughly from
1500 to 1650 A.D. During this period the search for the
philosopher's stone (then regarded as a means of healing all
diseases) and for the equally imaginary ELIXIR OF LIFE
gradually gave place to deliberate investigations of the
action of drugs on the human body and to the preparation
of new substances for use in medicine. The writings of
Glauber (1603-1668) contain a description of many sub-
stances discovered during this period, and give a good idea
of the state of knowledge at the time when Boyle laid the
foundations of the modern science of Chemistry.

The following pages contain an account of some of the
most important materials which became known during the
earlier part of these three periods, but large groups of
substances (including, for instance, the acids and alkalis)
are reserved for separate treatment in subsequent chapters.
The various materials may be classified conveniently under
the following headings :

Soluble salts. — Among the substances in common use
from very early times was the salt obtained by the evapora-
tion by the sun's rays of sea-water collected in shallow
pools along the sea-shore. The product was a mixture of
several substances, remarkable for their sharp taste and

I RAW MATERIALS AND PRIMITIVE MANUFACTURES 3

their solubility in water, and contained a large proportion of
the substance now known as COMMON SALT (Fig. i). The
name SALT was applied subsequently to all similar solids.
Thus Boyle denned a " salt " as being characterised by two
qualities, that " it is easily dissoluble in water and that it
affects the palate with a savour, whether good or evil " (see
Experiments and Notes about the Produdbleness of Chymical
Principles, 1680, p. 3; Works, 1725, III. 365).

SODA (Latin natruni), known from the earliest times as a
natural deposit on the shores of the soda-lakes of Egypt,
was originally
called " nitre ";
it was employed
as a cleansing
agent and in the
manufacture of
glass, but was
almost u n -
known in West-
e r n Eu rope
until the eight-
eenth century,
when it was
prepared from
the ash of marine plants. POTASH, or " pearl ash," a
white solid closely resembling soda in many of its pro-
perties, probably received its name from the fact that it
was obtained by extracting the white ash of burnt wood
with water in earthenware pots. During the middle ages
the chief source of potash was the "lees" or sediment
of wine to which the name of TARTAR was given ; this
sediment was calcined, and the potash thus prepared, the
" burnt lees of wine," was known as CALCINED TARTAR, or
more simply as " tartar." These two substances, soda and
potash, were known as ALKALIS, and were remarkable for

B 2

FIG. T— CUBIC CRYSTAL OF SALT.
British Museum (Natural History).

4 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

their power of effervescing when mixed with acids. This
property served to distinguish the Egyptian nitre or soda
from the common nitre or saltpetre described in the next
paragraph. It is referred to in the proverb : " As he that
taketh away a garment in cold weather, and as vinegar upon
nitre, so is he that singeth songs to an heavy heart"
(Prov., xxv. 20). In reference to this passage Robert Boyle,

FIG. 2 — CRYSTAL OF

SALTPETRE.
British Museum
(Natural History).

FIG. 3— SAL-AMMONIAC.

who had received a specimen of Egyptian " nitre " from
Constantinople, wrote in 1680: —

" And here .... give me leave to take notice of a text
of the holy Scripture, that has sometimes puzzled not only
me, but far better Critics in the Hebrew tongue than I, ....
where to illustrate Things very incongruous to one another

I RAW MATERIALS AND PRIMITIVE MANUFACTURES 5

the disagreement of Vinegar and Nitre is mentioned, for
supposing the words to be rightly translated .... it seems
very hard to find what show of Antipathy there is between
Vinegar, and the Saltpetre that is commonly sold in our
shops for Nitre ; wherefore strongly presuming that Solomon
.... made use of Egyptian Nitre .... when once I
received the Nitre that I have mentioned, and saw it in signs
of an Alcalizate nature, I quickly poured upon it some strong
Vinegar, and found as I expected that there presently ensued
a manifest conflict, with noise, and store of bubbles, with
which Experiment I afterwards acquainted some Critics,
and other learned men who were not ill-pleased with it "
(Experiments and Notes about the Produdbleness of Chymical
Principles, 1680, p. 30; Works,
1725, III. 371).

SALTPETRE or NITRE (Fig. 2),
a salt-like substance formed
by the decay of animal matter
and found as an incrustation
in the neighbourhood of
stables, was introduced into
Europe from the East. SAL-
AMMONIAC (Fig. 3), a salt of
similar origin, was manufac- FIG. 4-THE COI.ENSO DIAMOND

ill,- i) j British Museum (Natural History)

tured by heating camels dung ;

it differed from other salt-like substances in that it could
be vaporised completely by gentle heat. The name was first
applied to a mixture of common salt and soda found near
the temple of Jupiter Ammon in Upper Egypt, but was
transferred by the early alchemists to the volatile salt just
referred to.

Earths and rocks. — A number of other substances, which
may be classed together as earths or rocks, were employed
for various purposes. FULLER'S EARTH, a white, friable clay,
was used as a cleansing agent prior to the manufacture
of soap. CHALK, LIMESTONE, and MARBLE were employed

6 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

largely in their natural condition as building materials,
but were also converted into lime by the action of heat and
used in the preparation of mortar. SAND was used with soda
in the manufacture of glass.

Many crystals and precious stones were also known and
valued for personal adornment and for decorative purposes.
Amongst the first to attract attention were probably EMERALD,
TOPAZ, DIAMOND (Fig. 4), and QUARTZ (Fig. 5) ; the last

FIG. 5 — CRYSTALS OF QUARTZ.
British Museum (Natural History)

gave

substance, known also as ROCK-CRYSTAL, or CRYSTAL,
its name to the whole of this class of substances.

Substances of organic origin, — Many of the substances
known in ancient and mediaeval times were formed by the
agency of living organisms, either animals or plants. Among
these may be mentioned SUGAR (in the form of honey),
TURPENTINE, OILS, FATS, and WAXES (extracted from plants
and from the bodies of animals), AMBER (a fossil resin), and
PEARLS. WINE and VINEGAR were obtained by fermentation

I RAW MATERIALS AND PRIMITIVE MANUFACTURES 7

from the juice of the grape and other fruits, whilst saltpetre
and sal-ammoniac have been referred to already as products
of the decay of animal matter.

Substances prepared by the action of fire. — Many other
substances were prepared from natural materials by the
action of fire. LIME, obtained in this manner from lime-
stone or chalk, has been mentioned previously, but greater
importance attaches to the use of fire as an agent for the
preparation of metals. Two metals, GOLD and SILVER, are
distributed somewhat widely in a native state, and were
known from the earliest times. Native COPPER was also
found and used.1 MERCURY (Latin hydrargyrum, or liquid
silver), which occurs in minute droplets in certain rocks5
was known to the Greeks.

Other metals were obtained by smelting their ores, that
is, by heating them with charcoal. Amongst these was TIN,
obtained by smelting TINSTONE and valued highly as a
means of hardening copper. The hard alloy of copper and
tin is known as BRONZE. The stage of civilisation during
which this alloy came into common use has been called the
" Bronze Age," although the various European peoples
learnt to use it for the manufacture of weapons and imple-
ments at widely different times. The smelting of IRON was
a more difficult process, since a much higher temperature
was required than in the case of tin or copper. The use of
iron, therefore, follows that of bronze in the history of each
race. Thus, the Greeks, as described in the Homeric poems,
were accustomed to the use of bronze weapons and imple-
ments, but esteemed iron much more highly — a lump of
iron being described as a valuable prize in a contest. The
Romans had reached the " Iron Age " by early classical
times. Four hundred years later, at the opening of the
Christian era, the Germanic races still employed the earlier

1 "A land whose stones are iron, and out of whose hills thou
mayest dig brass " (Deut. viii. 9).

8 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

known metal, although the use of iron had superseded
that of bronze partially amongst the Gauls and other Celtic
races. LEAD was known to the Romans at the time of their
conquest of Britain ; the mining of lead ores was carried
on by them in Derbyshire and in other parts of the island.
Fire was also used as an agent in the purification of the
so-called " noble metals," gold and silver, from baser im-
purities, such as tin and lead, the latter being converted into
dross. The dross obtained by burning lead in order to
separate it from the silver which it contained, received

a special name,

LITHARGE, that

is, the stone
(Greek, Ai<9os)
obtained from
silver (Greek,
apyvpos) ; it was
valued because
it could be con-
verted by gentle
roasting into the
scarlet paint
known as

MINIUM Or RED

FIG. 6— ARROWHEAD CRYSTALS OF MARCASITE
British Museum (Natural History).

LEAD.

Substances produced by weathering or decay. — New

materials were also obtained by the natural processes of
weathering or decay. SALTPETRE, WINE, and VINEGAR have
been mentioned as examples of this kind, but special
reference may be made to the green, glassy substance which
is formed when the minerals MARCASITE and PYRITES (Figs.
6, 7, 8) are allowed to weather. The brilliant golden nodules
of marcasite, which in England are often found as " thunder-
bolts," embedded in the chalk, decay and become " rusty "
almost as easily as iron. When the rusty mass is extracted

I RAW MATERIALS AND PRIMITIVE MANUFACTURES 9

with water, a soluble substance is dissolved out and separates
in green crystals on allowing the washings to evaporate.
The glassy appearance of the crystals won for the substance the
name of viTRiOL(Latin vitrum, glass). It played a most impor-
tant part in the early development of chemistry on account of

FIG. 7 — NODULE OF MARCASITE AS DESCRIBED BY GLAUBER.

FIG. 8 — STRIATED CUBES OF PYRITES.

Marcasite and iron pyrites both consist of disulphide of iron (FeSa), but they

differ in crystalline form. The crystals of iron pyrites are cubic whilst those of

marcasite belong to the orthorhombic system.

the discovery of an oily product, OIL OF VITRIOL, to which it
gave rise when strongly heated (see Chapter II). The
following description of the way in which green vitriol was
prepared from nodules of marcasite was written by Glauber
about 1648 A.D., and translated into English in 1651 : —

io HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

" Commonly in all fat soils or clayey grounds, especially
in the white, there is found a kind of stones, round or oval

in form Which if it is cleansed from the earth,

and beaten to pieces, looks within of a fair yellow and in
streaks, like a gold Marcasite, or a rich gold Ore, but there
is no other taste to be perceived in it, than in another
ordinary stone ; . . . . Now this stone is nothing else, but
the best and purest Mineral (or Ore) of Vitriol, . . . out of
which there may be made an excellent medicine, as
followeth.

Take this Ore or Mineral beaten into pieces, and for some
space of time, lay or expose it to the cool air, and within
twenty or thirty days it will magnetically attract a certain
saltish moisture out of the air, and grow heavy by it, and at
last it falleth asunder to a black powder, which must remain
further lying there still, until it grow whitish, and that it do
taste sweet upon the tongue like vitriol. Afterward put it in
a glass vessel, and pour on so much fair rainwater, as that itv
cover it one or two inches ; stir it about several times in a
day, and after a few days the water will be coloured green,
which you must pour off, and pour on more fair water, and
proceed as before, stirring it often until that also come to be
green : this must be repeated so often, until no water more
will be coloured by standing upon it. Then let all the green
waters which you poured off run through filtering paper, for
to purify them ; and then in a glass-body [ i.e., a retort ] cut
off short let them evaporate till a skin appear at the top :
then set it in a cold place, and there will shoot little green
stones, which are nothing else but a pure vitriol : the re-
maining green water evaporate again, and let it shoot as
before : and this evaporating and crystallising must be
continued until no vitriol more will shoot " (Philosophical
Furnaces ; Part II. 1651, p. 71 ; Works, 1689, I. 21).

SUMMARY AND SUPPLEMENT

Scientific Chemistry begins with Robert Boyle (1627-1691),
but was preceded by three periods, namely : —

(i) An Ancient Period, extending up to the Christian era,
in which many primitive manufactures were developed ;

I RAW MATERIALS AND PRIMITIVE MANUFACTURES 11

(2) An Earlier Alchemistic Period, extending to about
1500 A.D., during which many substances were discovered in
the attempt to convert base metals into gold ;

(3) A Later Alchemistic Period, extending to about 1650 A.D.,
in which new substances were prepared and tested for
medicinal purposes.

Important substances known at a very early date include :

1. Soluble Salts.—

Common Salt (Sodium chloride, NaCl).
Washing Soda (Sodium carbonate, Na2CO3,ioH2O).
Potash (Potassium carbonate, K2CO3).
Saltpetre or Nitre (Potassium nitrate, KNO3).
Sal-ammoniac (Ammonium chloride, NH4C1).

2. Earths and Rocks.—

Fuller's Earth.

Chalk, Limestone, Marble (Calcium carbonate, CaCO3).

Sand, Quartz (Silicon dioxide, SiO3).

Emerald. Topaz. Diamond.

3. Organic Products.—

Sugar. Turpentine. Oils, Fats and Waxes. Amber
and Pearls. Wine and Vinegar. Saltpetre and
Sal-ammoniac.

4. Igneous Products. —

Lime from Limestone (CaCO3-> CaO + CO2).
Tin (stannum, Sn) from Tinstone

Iron (ferrum, Fe) from Ironstone

Litharge from Lead (2Pb + O2-> 2PbO).

Also Native Metals.—
Gold (aurum, Au).
Silver (argentum, Ag).
Copper (cuprum, Cu).
Mercury (hydrargyrum, Hg).

5. Products of Decay.— Especially, Iron Pyrites (FeS2)
-> Green Vitriol (Ferrous sulphate, FeSO4,7H2O)
-> Oil of Vitriol (Sulphuric acid, H,SO4). .

CHAPTER II

THE ACIDS

A. DISCOVERY OF THE COMMON ACIDS

Vegetable acids. — The earliest acids known were of
vegetable origin, but until the middle of the eighteenth
century scarcely any attempt was made to isolate them from
the "sour" liquids in which they occur, or even to distinguish
between various acids of similar origin. The most familiar
of the vegetable acids was sour wine or VINEGAR, which
was known to have a remarkable action upon soda
(Chapter I. p. 4). Its power of dissolving chalky materials is
illustrated by the story of Cleopatra and the pearls which
she dissolved and drank in a cup of vinegar, as well as by
Livy's fantastic story of the use of vinegar by Hannibal to
dissolve away the limestone rocks of the Alps. DISTILLED
VINEGAR was familiar to the alchemists from the time of
Geber, and was frequently used as a solvent, but it was not
until a much later period that the acid constituent, ACETIC
ACID, was isolated in a pure state.

A large number of crystalline acids of animal and vegetable
origin were, however, prepared at the close of the eighteenth
century by the Swedish chemist, Carl Wilhelm Scheele,
(1742-1786), whose work on these substances may be

CH. II THE ACIDS 13

regarded as the basis of the modern science of " Organic
Chemistry." Amongst the acids discovered by Scheele were
TARTARIC ACID prepared in 1770 from " tartar," in which it
is present in combination with potash ; BENZOIC ACID from
benzoin (1775); URIC ACID from bladder-stones (1776);
LACTIC ACID from sour milk (1780) ; OXALIC ACID by the
action of nitric acid on oils (1783); CITRIC ACID from
lemon-juice (1784); and MALIC ACID from apples (1786).

Oil of vitriol, or sulphuric acid. — A great advance was
made by the discovery in the early alchemistic period of
powerful acids of mineral origin. The first of these to be
prepared was undoubtedly oil of vitriol, which the writings
of Geber (800 A.D.) describe as obtained by the distillation
of alum. The acid can be prepared more easily by distil-
ling green vitriol, as described in the writings of Basil
Valentine and of Glauber, and it was from this method of
preparation that the acid obtained the name OIL OF VITRIOL.
The first stage of distillation results in the formation of clouds
of steam : subsequently, dense white fumes are produced,
which dissolve in the condensed steam to form oil of vitriol.
When the white fumes are condensed separately, a strongly
fuming liquid is produced which hisses violently when poured
into water : this fuming liquid is often called NORDHAUSEN
OIL OF VITRIOL from the name of the German town in which
it used to be manufactured. The red, rusty residue remaining
behind in the retort was known as COLCOTHAR, and is now
sold as ROUGE. This method of making oil of vitriol is
described by Basil Valentine as follows : —

" If you get such deep graduated and well prepared
Mineral, called Vitriol, then pray to God for understanding
and wisdom for your intention aqd after you have calcined
it, put it into a well coated Retort, drive it gently at first, then
increase the fire, there comes in the form of a white spirit of
vitriol in the manner of a horrid fume, or wind, and cometh
into the Receiver as long as it hath any such material in it."
(Last Will and Testament, p. 158.)

14 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

The acid is now prepared by more direct methods from
iron pyrites or from sulphur. The presence of sulphur in
pyrites l and in oil of vitriol was known in Boyle's time, but
the name SULPHURIC ACID by which the acid is now generally
known was not adopted until 1787.

Aqua fortis or nitric acid. — A second acid of mineral
origin was described by Geber as obtained by distilling a
mixture of saltpetre with green vitriol and alum. At a
later period Glauber showed that it could be prepared more
easily, and in a much purer condition, by distilling a mixture
of oil of vitriol and saltpetre from a glass retort heated gently
in a bath of hot sand over a furnace. From its remarkable
power of dissolving metals such as copper and silver, which
were not readily acted on by oil of vitriol, it came to be
known as AQUA FORTIS. In Boyle's time its acid properties,
its volatility, and its origin from nitre were indicated by the
name "acid spirit of nitre"; this was afterwards shortened
to nitrous or NITRIC ACID, the last name being introduced by
Lavoisier in 1787.

Aqua regia — By dissolving sal-ammoniac or salt in aqua
fortis, Geber prepared a still more powerful acid which was
capable of dissolving gold; it was therefore called AQUA

REGIS, Or AQUA REGIA.

Spirit of salt or muriatic acid.— The method of prepar-
ing spirit of salt by heating salt with oil of vitriol in a
glass retort is due to Glauber. He had previously made it
by throwing a mixture of salt, green vitriol, and alum upon
the hot fuel of a charcoal fire and passing the fumes into a
large glass globe. The action of oil of vitriol on salt produces
a pungent fume which escapes into the air and is lost ; but
Glauber found that this fume condensed readily in a receiver
half filled with water, giving an acid liquid which he described

1 "Vitriols are produced from the stone . . . called Marchasite, and
from it on the application of fire the flowers of common sulphur are
elicited in considerable abundance " (John Mayow, Medico-physical
Works, 1674 ; Alembic Club Reprints, XVII. 28).

n THE ACIDS 15

as SPIRIT OF SALT ; another common name for the acid is
MURIATIC ACID (Latin, murium, brine). The gas itself
was first isolated by Priestley, who showed that it could be
collected quite readily over mercury.

Glauber also showed that the mixture of spirit of salt and
aqua fortis, which is produced by distilling salt and saltpetre
with oil of vitriol, was capable of dissolving gold and had
the same properties as the aqua regia prepared by Geber's
method.

B. PROPERTIES OF THE ACIDS.

Taste. Action on vegetable dyes. — The most con-
Jspicuous property possessed by all the above acids was their
sour taste. To this may be added the fact that they were
capable of changing the colour of various vegetable dyes.
During the eighteenth century, SYRUP OF VIOLETS, which
changes from violet to red on the addition of an acid, was
largely used as a test or INDICATOR for these substances ; at
a later date " tournesol," or LITMUS, which changes from
blue to red more easily than syrup of violets, and various
artificial dyes, were used for this purpose.

When concentrated, the mineral acids were found to have
very corrosive properties. Glauber describes the charring of
a slip of wood and the ignition of turpentine and of spirit of
wine as properties of the strongest oil of vitriol. Aqua fortis
was equally corrosive, but both acids become harmless when
sufficiently diluted. The diluted mineral acids were at one
time employed in medicine and as substitutes for vinegar
and lemon-juice in the preparation of food.

Action on alkalis and chalk. — When added to soda or
potash all the acids were found to produce the violent effer-
vescence that was first noticed in the action of vinegar on
soda (p. 4). A similar " breaking out of air " took place
when chalky materials were acted on by acids : in this action

16 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

the chalk was usually dissolved, but oil of vitriol gave a
white solid which was proved by Margraaf, in 1750, to be
identical with the mineral GYPSUM or SELENITE (Fig. 10,
p. 19), from which "Plaster of Paris" is made by gentle
burning. The origin and the nature of the "air" which is
set free in these actions are discussed in Chapters IV.
and VI.

Action of acids on metals. — The acids also had the
property of corroding or dissolving metals. The poisonous
green powder known as VERDIGRIS was prepared at a very
early date by the action of vinegar on copper, and was used
as a paint during the classical period. The mineral acids
were found -to act on metals in a much more powerful
way ; by the time of Geber (circ. 800 A.D.) methods had"
been discovered for dissolving all the metals that were
known. The action of acids on metals was often accompanied
(as in the case of soda and chalk) by the breaking out of a
gas, but for many centuries no method was known by which
these fugitive products could be collected. The important
discoveries which were made when at last it was found
possible to isolate and examine them are described in
Chapter V.

C. PREPARATION OF NEW SALTS AND NOMENCLATURE
OF SALTS

New salts. — One result of the discovery of the acids
was to add very greatly to the number of "salts" which
were known. When an acid acts on a metal, on chalk, or
on an alkali, a solution is produced which no longer has the
sour taste of the acid. These solutions contain a variety
of salt-like substances, which can be isolated by evaporating
either to dryness or until crystals begin to separate. In this
way many beautiful and useful salts were obtained. At
first a special name was given to each salt ; but later a

THE ACIDS

1 7

system was devised in which each salt was named after the
acid and the BASE (metal, alkali, or earth) from which it was
derived, e.g., nitrate of silver, sulphate of potash, muriate of
lime. This system was initiated about the middle of the
eighteenth century and completed by a group of French
chemists in 1787*; it led inevitably to the inclusion in the
category of " salts " of many insoluble and tasteless sub-
stances. Selenite, for instance, when prepared by the action
of oil of vitriol on lime or chalk, could scarcely be excluded
from the category of salts merely because it was only slightly
soluble in water ; for the same reason it was necessary to
regard as a salt the
insoluble muriate of
silver, which could
scarcely be separated
in an arbitrary way
from the soluble muri-
ates of copper and gold.

Some of the more
important salts pre-
pared with the help of
the acids are described
below.

Vitriols or sulphates.— The salts prepared by the action
of oil of vitriol on metals were of a glassy crystalline
character, which won for them the name of VITRIOLS.

"Out of all Metals there can be made a Vitriol or
Chrystal (Chrystal and Vitriol is taken for one)." (Basil
Valentine, Last' Will and Testament, p. 157.)

This name was afterwards limited to crystals prepared
from, or related to, oil of vitriol. The most important of
these were GREEN VITRIOL, and BLUE VITRIOL (Fig. 9),

1 The Methode de Nomenclature Chimique, by MM. de Morveau,
Lavoisier, Berthollet, and Fourcroy, was published in Paris in 1787,
and translated into English in 1788 and 1796.

FIG. 9— CRYSTAL OF BLUE VITRIOL
OR SULPHATE OF COPPER.

18 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

obtained by the weathering of different varieties of pyrites.
The blue vitriol was of special interest because iron dipped
into it became coated with copper, and seemed to have
been transmuted into that metal. These two vitriols were
prepared artificially by the action of oil of vitriol on plates
of iron or copper, by Glauber, who writes as follows : —

" Take of your heavy oil .... as much as you please,
put it into a glass body together with plates of copper or
iron, set it in a warm sand, and let it boil until that the oil
will dissolve no more of the metal, then pour off the liquor,
filter it through brown paper, and put it into a low gourd
glass and set it in sand, and let the phlegm evaporate until
there appear a skin at the top, then let the fire go out, and
the glass grow cold, then set it in a cold place, and within
some days there will shoot fair Crystals ; if of Iron, greenish ;
if of Copper, something bluish ; take them out and dry them
upon filtering paper, the remaining liquor, which will not
shoot into Vitriol, evaporate again in sand, and then let it
shoot as before ; continue this proceeding, until all the
solution (or filtered liquor) be turned to Vitriol" (Philoso-
phical Furnaces. Part IV; Works, I. 18).

These two vitriols were described by Lavoisier and his
colleagues as SULPHATE OF IRON and SULPHATE OF COPPER,
after the acid and metals from which they were derived.

GLAUBER'S SALT was prepared by the action of oil of
vitriol on common salt. When this action was carried out
in a glass retort, the salt was separated easily in a pure con-
dition by crystallisation. Its discoverer attributed to it
almost miraculous properties and called it " sal mirabile."
It was afterwards prepared from oil of vitriol and soda, and
was therefore called SULPHATE OF SODA. The corresponding
salt prepared by the action of oil of vitriol on potash was
known in Boyle's time as VITRIOLATED TARTAR, but in
Lavoisier's system became SULPHATE OF POTASH.

GYPSUM or SELENITE (Fig. 10), which could be prepared

THE ACIDS

2,

artificially from oil of vitriol and chalk or lime (p. n),
was called SULPHATE OF LIME. EPSOM SALTS, a purgative
salt derived from mineral springs, was shown by Black in
J755 to contain a base MAGNESIA in combination with oil
of vitriol ; it was therefore called SULPHATE OF MAGNESIA.

Nitres or nitrates. — NITRE or SALTPETRE (Fig.
p. 4), for many centuries
the only source from which
nitric acid and the nitrates
could be derived, was pre-
pared artificially in 1674
by John Mayow, a friend
and contemporary of
Boyle, by recombining the
nitric acid with potash ;
its systematic name was
therefore NITRATE OF

POTASH. A NITRATE OF

SODA was prepared by
Geber ; enormous deposits
of this salt have been
discovered in the desert
regions of Chile, and mil-
lions of tons are now ex-
ported every year for use
in agriculture under the
name of CHILE SALTPETRE.
The nitrates of the metals
were well-known to the
Alchemists. Geber describes LUNAR CAUSTIC, the NITRATE
OF SILVER, as prepared in the form of " small fusible stones
like crystal," by dissolving silver in aqua fortis and boiling
away two-thirds of the water in a long-necked flask. The
nitrates of lead, copper, iron, and mercury, were also prepared
at an early date.

c 2

FIG. 10 — LARGE CRYSTAL OK GYPSUM OR
SELENITE (Sulphate of Lime).

.

.
20 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

Muriates. — Glauber, who discovered the first efficient
method of making spirit of salt or muriatic acid, prepared
the MURIATES OF IRON, COPPER, GOLD, and other metals, by the
action of spirit of salt, or of aqua regia upon the metals ; he
obtained them in the form of strong solutions which he
described as "oil of Mars," "oil of Venus," etc., in accordance
with the alchemistic system (which survives in the case of
Mercury) of calling each of the common metals after a
planet.

Common salt, which can be prepared artificially by recom-
bining muriatic acid with soda, is a MURIATE OF SODA. The
corresponding MURIATE OF POTASH, known as SAL SYLVII,
or SYLVINE, was prepared by the action of* muriatic acid
on wood-ashes or potash. Extensive deposits of the salt
have been found at Stassfurt in Germany ; the mineral is
used on a large scale as a fertiliser in agriculture, and is
one of the chief sources from which potash is derived.

Black in 1755 prepared the MURIATE OF MAGNESIA, and
compared its properties with those of MURIATE OF LIME.

Acetates. — Of the salts derived from organic acids the most
important were the ACETATES prepared from vinegar, or acetic
.acid. The ACETATE OF SODA and ACETATE OF LIME prepared
by the action of vinegar on soda and on chalk (as described
on p. 15), were amongst the first salts to be prepared arti-
ficially. Mention may also be made of the ACETATE OF LEAD
which Basil Valentine prepared from vinegar and litharge,
and which acquired the name SUGAR OF LEAD on account of
its sweet taste ; also of the ACETATE OF COPPER, which he
prepared by the action of vinegar on verdigris.

" There is extracted from calcined Saturn [*>., burnt lead
or litharge] with distilled Vinegar a Crystalline Salt."

"Take some pounds of Verdigris, extract its Tincture with
distilled Vinegar, let it shoot, then you have a glorious
Vitriol." (Basil Valentine, Last Will and Testament, pp.
349 and 351.)

ii THE ACIDS 21

Black in 1755 prepared the ACETATE OF MAGNESIA, in
addition to the muriate, nitrate, and sulphate, for comparison
with the corresponding salts of lime.

Classification of salts. — The salts described above were
prepared by the action of acids (i) on metals such as iron
and copper ; (2) on the alkalis, soda and potash ; (3) on
earths such as lime and magnesia. The distinction, which
was at first made between these three classes of salts, was
rendered of little value by the observation that the salts of
metals could be prepared much more easily from the earthy
calces which are formed when the metals are burnt : Glauber,
for instance, showed that the muriate of copper was prepared
easily by the action of muriatic acid on the calcined metal,
although the metal itself was attacked but slowly. The
distinction was broken down finally at the commencement
of the nineteenth century by the discovery that the alkalis
and earths were themselves calces of easily-burnt metals.
When this had been proved the custom arose of describing
all salts as derivatives of metals. Thus gypsum, which was
called by Lavoisier " sulphate of lime," is now described as
CALCIUM SULPHATE, and Glauber's salt is called SODIUM
SULPHATE instead of " sulphate of soda. " The older
names are, however, still used in commerce and in
pharmacy.

Strength of acids. — The fact that oil of vitriol could
displace muriatic acid from common salt, and nitric acid from
nitre, was recorded by Glauber. Attention was directed at
first mainly to the liberated acid, but it was recognised soon
that a soluble substance was left behind which contained the
" fixed salt " (i.e., the base or alkali) of the original substance
in combination with the stronger vitriolic acid. Thus,
when oil of vitriol acted upon common salt, the residue of
Glauber's salt was found to be identical with sulphate of
soda prepared by the action of oil of vitriol on soda ; when
nitre was used, the residue was a "vitriolated tartar" or

22 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

sulphate of potash identical with that prepared from oil of
vitriol and " tartar" or potash.

These actions were studied carefully by John Mayow, a
London Physician (1645-1697), who pointed out in 1674
that " although [acids] and [alkalis] pass into a neutral sub-
stance when they meet, yet they do not, as is generally
supposed, entirely destroy each other," since when the
conditions are suitable both the acid and the alkali may
be recovered from the salt. Thus: —

" If oil of vitriol is poured upon nitre, which consists of
an [alkali] and of a volatile [acid] (as was shown above),
the fixed salt [i.e., the base] of the nitre will soon leave its
own acid and will enter into union with the acid of the
vitriol, which is more concordant with it .... That the
case is so, is clear, for if nitre mixed with oil of vitriol be
distilled, the spirit, or [acid] of the nitre will pass under
a mild heat into the receiving vessel, while yet in other
circumstances that spirit will not be carried up except by a
very vehement fire .... It is a corroboration of this view
that the mass left in the retort after a distillation of this
kind, closely resembles vitriolated tartar, and can be properly
substituted for it." (" Of the combination of contrary salts,"
Medico-physical Works, A.C.R. XVII. 161—162.)

As oil of vitriol was found to liberate nitric acid from any
nitrate and muriatic acid from any muriate, it was regarded
as stronger than either of these acids, whilst vinegar was
found by similar tests to be weaker than the three mineral
acids. It is recognised now that this rough and ready
method is not the best test of the strength of an acid, as it
depends too largely on the readiness with which the various
acids can be driven off in the form of vapour. But
observations such as these were of great importance
because they showed clearly that all salts possessed a
dual structure ; they thus prepared the way for the system
of classifying salts which is described in the preceding
paragraphs.

n THE ACIDS 23

SUMMARY AND SUPPLEMENT.
A and B.— DISCOVERY AND PROPERTIES OF THE ACIDS.

Vegetable Acids include :—

Acetic acid, C2H4O2, from Vinegar (known to Geber) ;
Tartaric acid, C4H6O6, from Tartar (Scheele, 1770) ;
Benzoic acid, CrH6O2, from Benzoin (Scheele, 1775) ;
Uric acid, C5H4N4O3, from Bladder-stones (Scheele, 1776) ;
Lactic acid, C3H6O3, from Sour Milk (Scheele, 1780) ;
Oxalic acid, C2H2O4, from Oils (Scheele, 1783) ;
Citric acid, C6H8O7, from Lemons (Scheele, 1784) ;
Malic acid, C4H6O6, from Apples (Scheele, 1786).

Mineral Acids include :—

Oil of Vitriol, Vitriolic acid, or Sulphuric acid, H2SO4, pre-
pared by distilling Alum (Geber) or Green Vitriol (Glauber).

Aqua Fortis, Acid Spirit of Nitre, or Nitric Acid, HNO3, pre-
pared by distilling Nitre and Alum (Geber) or Nitre and Oil
of Vitriol (Glauber), 2KNO3 + H,,SO4->K2SO4 + 2HNO3.

Spirit of Salt, Muriatic acid, or Hydrochloric acid, HC1, pre-
pared by heating Salt and Oil of Vitriol, and collected in
water (Glauber), 2NaCl + H2SO4-> Na2SO4 + 2HC1.

Aqua Regia, a mixture of Aqua Fortis and Spirit of Salt
prepared by adding Sal-ammoniac to Aqua Fortis (Geber),
NH4Cl + HNO3-> NH4N03+HC1, or by distilling a mixture
of Nitre and Salt with Oil of Vitriol (Glauber).

The acids as a class possess the following properties, though
some of them may be lacking in the case of individual acids : —

(a) General Properties : Sour taste ; acids turn syrup of violets
red and redden blue litmus ; the mineral acids are corro-
sive when concentrated but harmless when dilute ;

(b] Acids dissolve Alkalis and Chalk, liberating gas, e.g.,
"Vinegar upon Nitre (i.e., soda)" and upon Chalk,

Na2CO3 + 2C2H4O2->2NaC2H3O2 + H2O + CO2
CaCO3 + 2C2H4O2->Ca(C2H302)2 + H20 + CO2 ;
Oil of Vitriol with Soda gives " Glauber's salt," with Chalk gives
" Gypsum" or "Selenite" (almost insoluble in water)
Na2CO3 + H2SO4-> Na2SO4 + H2O + CO2
CaCO3 + H2SO4->CaSO4 + H

24 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

(c] Adds dissolve metals, e.g., Oil of Vitriol and Muriatic acid
dissolve Iron ; Oil of Vitriol (concentrated) and Aqua Fortis
dissolve also Copper, Mercury and Silver ; Aqua Regia
dissolves Gold.

The action of acids and alkalis on vegetable tinctures such
as the blue-fluorescent extract of " Lignum nephriticum," syrup
of violets and cornflowers, and the purple juice of ripe privet-
berries was described by Boyle (Experiments on Colours, 1663).

D. SALTS.

Many salts can be prepared artificially by the action of Acids
on Metals, Alkalis (Soda and Potash) and earths. Amongst
the salts which were prepared and examined at an early date
the following may be noted : —

(a] Vitriols or Sulphates.— From Metals : Green Vitriol,
Sulphate of Iron, or Ferrous Sulphate, FeSO4, 7H2O ; Blue
Vitriol, Sulphate of Copper, or Copper Sulphate, CuSO4,5H2O.

From Alkalis : Glauber's Salt, Sulphate of Soda, or Sodium
Sulphate, Na2SO4, ioH2O ; Vitriolated Tartar, Sulphate of
Potash, or Potassium Sulphate, K2SO4.

From Earths : Gypsum or Selenite, Sulphate of Lime, or
Calcium Sulphate, CaSO4, 2H2O ; Epsom Salts, Sulphate of
Magnesia, or Magnesium Sulphate, MgSO4, 7H2O ; Alum,
Sulphate of Potash and Alumina, or Potassium Aluminium
Sulphate, KA1(SO4)2,I2H2O.

(b] Nitres or Nitrates.— Nitre or Saltpetre, Nitrate of Potash,
or Potassium Nitrate KNO3 ;

Chile Saltpetre, Nitrate of Soda, or Sodium Nitrate, NaNO3 ;
Lunar Caustic, Nitrate of Silver, or Silver Nitrate, AgNO3.
Also Nitrates of Lead, Copper, Iron and Mercury.

(c] Muriates or Chlorides. — Muriates or Chlorides of Iron
(FeCl3), Copper (CuCl2), and Gold (AuCl3). Common Salt
Muriate of Soda, or Sodium Chloride, NaCl ;

Sylvine, Muriate of Potash, or Potassium Chloride, KC1 ;
Muriate of Magnesia or Magnesium Chloride, MgCl2 ;
Muriate of Lime or Calcium Chloride, CaCl2.

(a] Acetates. — Sugar of Lead, Acetate of Lead, or Lead Ace-
tate, Pb(C2H302)2 ;
Acetate of Copper or Copper Acetate, Cu(C2H3O2)2, pre-

ii THE ACIDS 25

pared from Verdigris or Basic Acetate of Copper,

Cu(C2H30.2)2 + Cu(OH)2;

Acetate of Soda or Sodium Acetate, NaC2H3O2 ;
Acetate of Lime or Calcium Acetate, Ca(C2H3O2)2 ;
Acetate of Magnesia, or Magnesium Acetate, Mg(C2H3O2)2.

Three systems of nomenclature are seen in the names given
to the salts set out above: — (i) At first each salt received a
special name, in many cases recalling the origin or properties of
the salt. (2) The system of naming salts after the acid and
the metal, alkali or earth from which they are derived was
elaborated by de Morveau, Lavoisier, Bertholiet, and Fourcroy
in their Methode de Nomenclature Chimique, published in
1787. (3) This system became obsolete when Davy, twenty
years later, showed that the alkalis and earths contained
metals ; it was then possible to name every salt after the acid
and the metal from which it was derived and to abandon the
use of the alkalis and earths in naming salts.

The idea that salts still contained the acid and alkali from which
they were derived was put forward by John Mayow in his Medico-
Physical Works published in 1674 ; he showed that Ammonia
could be displaced from its Salts by Potash, 2NH4C1 + K2CO3
->2KC1 + (NH4)2CO3 and Aqua Fortis by Oil of Vitriol,
2KNO3 + H2SO4-> K2SO4 + 2HNO3, and was impressed with
the idea of the unequal " concordance " of the two acids with
the alkali, an idea that is essentially the same as that of the
unequal strengths of different acids. Rouelle, to whom we owe
the terms base (Mem. Acad., 1754, 572) and water of crystalli-
sation (ibid., 1744, 356) described how neutral salts had been
restricted at first to " salts formed by the union of acids with
alkalis, which are soluble in water, and imprint on the tongue
a saline taste. . . . The number of neutral salts was at first
very small, scarcely any were known but sea salt and nitre ; but
the number was soon increased, above all by the work of
Glauber. Others have since been added of which the bases are
the volatile alkali and an absorbent earth. Finally there have
been added salts formed by the union of acids with metallic
substances" (Mem. Acad., 1754, 572). He himself, in 1744, had
defined as a neutral salt, " a salt formed by the union of an acid
with any substance whatever, which serves as a base for it and
imparts to it a concrete or solid form " (ibid., 573).

CHAPTER III

THE BURNING OF METALS AND THE DISCOVERY OF OXYGEN

A. THE BURNING OF METALS.

Jean Rey (1630) shows that lead and tin increase in
weight when burnt — It has been known from very early
times that the metals, except gold and silver, are, by heating,
gradually changed to powders of various colours. These
powders were called CALCES from the resemblance which
they showed to lime (Latin, calx), and the process of
burning was called CALCINATION. A casual examination
showed that the calx was a lighter material than the metal
from which it was formed; it was, therefore, natural to suppose
that the burning of the metal had resulted in a loss of
weight, just as is obviously the case when wood or coal is
burnt. The fact "that tin and lead increase in weight
when they are calcined," was therefore "observed with
astonishment " by those who first put the matter to the test
of experiment.

The Essays of Jean Rey (1630), a French Doctor of
Medicine, " On an Enquiry into the cause wherefore Tin
and Lead increase in weight on Calcination" (Alembic Club
Reprints, No. XI), contain an account of one of the earliest
chemical researches of which a clear record has been pre-
served. He records that Brun, an apothecary of Bergerac,

" having placed two pounds six ounces of fine English tin

in an iron vessel and heated it strongly on an open furnace

26

CH. in THE BURNING OF METALS 27

for the space of six hours with continual agitation and
without adding anything to it, he recovered two pounds
thirteen ounces of a white calx ; which filled him with
amazement, and with a desire to know whence the seven
ounces of surplus had come " (A. C. R. XL 36).

It had been suggested that the gain in weight was due
to soot or vapour from the furnace condensing on the tin,
or to the disintegration of the iron vessel ; Rey showed
that these explanations were unlikely, and concluded (on
the basis of argument rather than of experiment)

" That this increase in weight comes from the air, which
in the vessel has been rendered denser, heavier, and in some
measure adhesive, by the vehement and long-continued heat
of the furnace : which air mixes with the calx (frequent
agitation aiding) and becomes attached to its most minute
particles : not otherwise than water makes heavier sand
which you throw into it and agitate, by moistening it and
adhering to the smallest of its grains " (A. C. R. XL 36).

Rey supported his argument by quoting an experiment
of Hamerus Poppius, who had calcined a cone of anti-
mony on a marble slab by means of a burning mirror,
and had found the weight to be augmented instead of
diminished, in spite of the copious exhalation of vapours and
fumes (A. C. R. XL 49) ; in this case at least, the gain in
weight could not be attributed to contamination of the metal
by the furnace or vessel, and must surely be due to the
condensation of air. A similar explanation was given in
the case of lead, of which it had been recorded that " it is a
remarkable thing that lead on calcination increases in weight
by eight or ten pounds per cent " (A. C. R. XL 41).

Robert Boyle (1673) confirms Key's statement that
metals gain in weight on calcination. —Jean Rey himself
does not appear to have made any experiments on the gain
in weight of tin and lead. But the fact that metals gain in
weight when calcined was confirmed forty-three years later

28 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

by the careful experiments of Robert Boyle on the calcina-
tion of copper and iron, as well as of tin and lead. Boyle's
observations showed a gain of : —

30 to 49 grains on 480 grains of copper x
60 „ „ 480 „ „ tin

66 „ „ 240 „ „ iron

7 „ „ 480 „ ,, lead (in spite of loss)

whilst silver showed a gain of only 2 grains on 212.
Boyle's experiments were carried out as follows : —

" Into a very shallow crucible, we put an ounce of copper-
plates, and set it in a cupelling furnace, where it was kept
for two hours • and then being taken out, we weighed the
copper, which had not been melted (having first blown off
all the ashes), and found it had gained thirty grains."

A similar experiment with an ounce of copper filings
gave an increase of 49 grains.

" Upon a good cupel, we put an ounce of English tin, of
the better sort ; and having placed it in the furnace, under
a muffler, though it presently melted, yet it did not forsake its
place, but remain upon the concave surface of the cupel,
till, at the end of about two hours, it appeared to have been
well-calcined ; and then being taken out, and weighed by
itself, the ounce of metal was found to have gained no less
than one dram."

" An ounce of lead was put upon a cupel, made of calcined
hartshorn, and placed under a muffler, after the cupel was
first made hot, and then weighed. This lead did not enter
the cupel, but was turned into a kind of litharge on the top of
it, and broke the cupel, whereby some part of the latter was
lost in the furnace ; yet the rest, together with the litharge,
weighed seven grains more than the lead and heated cupel,
when they were put in."

" Four drams of the filings of steel, being kept two hours

1 20 grains = I scruple
3 scruples = I dram
8 drams = I ounce.

in THE BURNING OF METALS 29

on a cupel, under a muffler, acquired one dram, six grains
and a quarter, increase of weight."

" A piece of refined silver, being put upon a cupel under
a muffler, and kept there for an hour and a half, was
taken out, and weighed again ; and as before it weighed
three drams, thirty-two grains and a quarter, it now weighed
in the same scales, three drams, thirty-four grains and a
quarter" (Works, 1725, II. 389-390).

Boyle calcines tin in sealed flasks. — Although Boyle was
convinced that metals really gained in weight when burnt,
he regarded these experiments as unsatisfactory because the
metals were exposed to the smoke and dust of the furnace.
To get over this difficulty he heated the metals between
two crucibles cemented together with clay, and finally (in
the case of the more fusible metals) made use of glass
flasks, the necks of which were loosely stoppered, drawn
out to a fine point, or sealed up altogether. Boyle's experi-
ments on the calcination of tin in sealed flasks are of special
interest as having provided the basis for Lavoisier's experi-
ments on the same subject.

"To prevent all suspicion of any increase of weight, in
the metals, arising from smoke, or saline particles, getting
in at the mouth of the vessel, I made the experiment in
glasses, hermetically sealed, as follows. Eight ounces of
good tin, carefully weighed, we hermetically sealed up in a
new, small retort, with a long neck, by which it was held in
the hand near a charcoal fire, that kept the metal in fusion ;
being now and then shaken for almost half an hour ; in
which time, it seemed to have acquired, on the surface, such
a dark colour, as argued a beginning calcination ; and it
both emitted fumes that played up and down, and also,
afforded two or three drops of liquor, in the neck of the
retort. The glass was, at length, laid on quick coals,
where the metal continued above a quarter of an hour longer
in fusion ; but, before the time was come, that I intended,
to suffer it to cool, in order to its removal, it suddenly broke
into a great multitude of pieces, and with a noise, like the
report of a gun."

30 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

In order to reduce the risk of explosion the flask was next
heated before sealing.

" Two ounces of filings of tin, were carefully weighed, and
put into a little retort, whose neck was afterwards drawn
slender to a very small apex ; then the glass was placed on
kindled coals, which drove out fumes at a small orifice of
the neck, for a pretty while. Afterwards, the glass, being
sealed at the apex, was kept in the fire for above two hours ;
and then being taken off, was broken at the same apex :
whereupon I heard the external air rush in, because, when
the retort was sealed, the air, within it, was highly rarefied.
Then the body of the glass being broken, the tin was taken
out, consisting of a lump, about which there appeared some
grey calx, and . some very small globules, which seemed to
have been filings melted into that form. The whole weighed
two ounces, and twelve grains " ( Works, II. 393-394).

"Fire and flame weighed in a balance" by Robert
Boyle. — Although the observations on calcination which
are described under this title l ( Works, 1725, II. 388) agreed
closely with those quoted by Jean Rey, Boyle gave a different
explanation of the gain in weight, which he attributed to the
absorption of heat instead of to the condensation of air.
Boyle's failure to recognise the essential part played by air
in combustion may be attributed to his observations " Of
the strangely difficult Propagation of Actual Flame in Vacuo
Boyliano " (New Experiments, touching the Relation betwixt
Flame and Air, London, 1672 ; compare Works, 1725, II. 517)
in which he found that various substances including sulphur,
gunpowder, and fulminating gold, could be fired, although
with difficulty, by contact with hot iron in a vessel from which
much of the air had been removed by means of an air-pump.

Boyle's opinion forms the basis of the "Phlogiston"
theory of Becher and Stahl. — Boyle's view that fire and
flame were material things which could be " weighed in a

1 The title of the original tract is ' ' New Experiments to make the
Parts of Fire and Flame stable and ponderable." London, 1673.

Ill THE BURNING OF METALS 31

balance " appeared again in a modified form in the THEORY
OF PHLOGISTON which dominated the science of chemistry
during the next hundred years, but was finally overthrown
by the work of Lavoisier from 1770 to 1787. This theory
was elaborated by two German philosophers, John Joachim
Becher (1635-1682), and George Ernest Stahl (1660-
1734), who sought to explain the phenomena of combustion
as due to a fire-element, or principle of inflammability, to
which Stahl gave the name PHLOGISTON. It was supposed
that phlogiston was present in all combustible substances ;
the largest proportion was contained in soot, which was
thought to be almost pure phlogiston, since it left only the
smallest residue of ash when burnt. In the smelting of
metals the phlogiston of the fuel combined with the ore to
produce the metal ; when the metal was burnt it parted with
the phlogiston it had taken from the fuel and was converted
into an incombustible calx, or ash. It will be noticed that
whilst Boyle regarded the calx as a compound of the metal
with igneous particles from the fire, thus

Metal '+ Fire = Calx,

Stahl regarded the calx as a simple substance, and the
metal as a compound of the calx with phlogiston, thus,

Calx + Phlogiston = Metal.

Metal- Phlogiston = Calx.

As the calx is heavier than the metal from which it is
derived it was clear that the phlogiston which escaped
during calcination must have a negative weight ; this curious
conclusion, although not much considered at first, led
ultimately to the destruction of the theory.

In Stahl's opinion, air was required in combustion merely
to absorb the phlogiston set free by the burning substance ;
air which had become exhausted by combustion was
thought to be saturated with phlogiston, and was called

PHLOGISTICATED AIR.

32 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

Hooke (1665) & Mayow (1674) suggest that air contains
an active principle similar to nitre. — The air, to which
Jean Rey had attributed the gain in weight of lead and tin
during calcination, was regarded by Boyle as having little
or nothing to do with this change. It was, therefore, left
to his contemporaries, Robert Hooke and John Mayow, to
follow up the clue which Rey had provided and to demon-
strate the part which air plays, not only in the burning of
metals, but in other cases of combustion.

Robert Hooke, in his " Micrographia " (1665), called
attention to the close resemblance between the action of
nitre, or saltpetre, and of air in various cases of burning.
He regarded air as a solvent for the burning substance, and
attributed its activity to the presence in it of a substance
similar to (or even identical with) melted saltpetre, but in a
very attenuated condition.

The similarity between air and nitre also formed the basis
of the theory of combustion put forward by John Mayow
in his " Medico-Physical Works " (Alembic Club Reprints,
No. XVII.), published in 1674. He showed, as Boyle had
done, that air was not needed for the burning of gun-
powder, since damp gunpowder would continue to burn
when the tube in which it was contained was plunged under
water (A. C. R. XVII. 9). There was, therefore, something
in the nitre which would take the place of air in enabling
charcoal and sulphur to burn. To this common principle,
present in air and in nitre, he gave the name SPIRITUS

NITRCWEREUS Or NITRO-^ERIAL SPIRIT (A. C. R. XVII. l).

Mayow (1674) proves that air is absorbed during
combustion, — Hitherto no attempt had been made to
collect and examine gases. When substances were dis-
tilled the volatile products were condensed in a cold
receiver, in which water was sometimes placed, but gases
and vapours which were not condensed in this way had
always been lost. To Mayow belongs the credit of intro-

Ill

THE BURNING OF METALS

33

ducing into chemistry the method of collecting gases in
flasks or bottles inverted over water, and of studying

FIG. 12.

FIG. ii.

FIG. 13.

FIG. 14.

MAYOW'S APPARATUS.

Fig. it shows a candle burning in a glass vessel inverted over water, and

also a piece of camphor being fired by a burning-glass.
Fig. 12 shows the diminution of the volume of air by a mouse breathing in

it ; a stretched bladder is sucked inwards as the air diminishes.
Fig. 13 shows the same experiment carried out with air contained in a glass

vessel inverted over water.
Fig. 14 shows the apparatus used to collect gases prepared artificially by

the action of acids on iron.

changes of volume in the gas by noticing the position of
the water in the glass vessel. Mayow's apparatus is shown
in Figs, u, 12, 13 and 14 (A.C.R. XVII. Plate 5). With

D

34 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

this apparatus he showed that a decrease in the volume of
air occurred when a candle was burned in an inverted flask,
or when camphor was fired in it by means of a burning
glass (Fig. n): this decrease he attributed to the disap-
pearance of the nitre-air during burning. He also showed
that a decrease in the volume of air was caused by a mouse
breathing in it (Figs. 12 and 13), and that the mouse died
almost immediately if placed in a jar of air in which a
flame was burning : the nitre-air used in burning was
therefore also necessary for breathing, and was used up in
just the same way as by
a burning candle.

Lavoisier (1774)
proves that the gain in
weight when tin is
calcined is due to ab-
sorption of air. — The
overthrow of the phlog-
iston theory and the
establishment of the
theory of combustion,
were the work of a
French nobleman, An-
toine Laurent Lavoisier
(1743-1794). Lavoisier met with an untimely death dur-
ing the French Revolution, but himself brought about a
revolution in the science of chemistry, so complete as
almost to justify the dictum of a fellow countryman,
" Chemistry is a French science. Its founder was Lavoisier
of immortal memory." l His experiments on combustion

1 This is the opening passage of Wurtz's Atomic Theory. A
contrary opinion was expressed by Brande, Professor of Chemistry in
the Royal Institution, who wrote in 1819 : "It is, I think, among our
own countrymen that we discover the fathers of chemical philosophy :
for Bacon, Boyle, Hooke, and Newton, present unequivocal claims to
that distinctive title" (Manual of Chemistry, p. xviii.).

FIG. 15 — LAVOISIER'S APPARATUS FOR CAL-
CINING LEAD AND TIN IN AIR OVER
WATER OR MERCURY.

in THE BURNING OF METALS 35

were carried out with a full knowledge of the work of his
predecessors, and were modelled largely upon the work of
Mayow and Boyle.

By means of a burning-glass 33 inches in diameter,
Lavoisier calcined lead and tin in air enclosed in a glass
vessel inverted over water or mercury1 (Fig. 15), and found
that " the volume of air diminished by about a twentieth
as a result of the calcination, and that the weight of the
metal is increased by an amount almost equal to that of
the air destroyed or absorbed." He concluded " that a
portion of the air .... combined with the metals during
their calcination, and that the increase in weight of the
metallic calces was due to this cause " (" Memoir on the
Calcination of Tin in Closed Vessels," 1774, Works, II. 105.)
Thus in contrast to the current view that
Metal = Calx + Phlogiston
Lavoisier reverted to the suggestion of Jean Rey that

Metal + Air = Calx,

the metal being regarded as a simple substance and the
calx as a compound.

Lavoisier calcines tin in a sealed flask. — In order to test
the view that the gain in weight of metals during calcination
was due to the absorption of air, Lavoisier, in 1774, repeated
Boyle's experiment of heating lead and tin in sealed glass
vessels. But, unlike Boyle, he carried out the critical experi-
ment of weighing the sealed retort both before and after
heating. In making these experiments Lavoisier had the
advantage of using a balance, more sensitive than any that
had been constructed previously, with which he was able to
detect changes in weight of yj^ of a grain.

The experiment was carried out as follows : Eight ounces
of tin were weighed into a retort of 43 cubic inches capacity,
heated to drive out part of the air, sealed up and again
weighed. The tin was then heated in the sealed retort for

1 Compare Priestley, Experiments on Air, I. 136.

D 2

36 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

an hour and a quarter, until the surface of the molten metal
ceased to tarnish and a considerable quantity of a black 1
calx had collected. In spite of the calcination that had
taken place the sealed retort had only changed in weight
by the loss of a quarter of a grain. If, therefore, the
metal had gained in weight "it was necessary to look
for the cause in the interior of the retort" (Works,

II. 112).

The retort was then broken into halves by cracking it
with a hot coal, and the whole was weighed again. A
comparison with the first weighing of the tin and" retort
showed a gain of 3-13 grains; further weighings showed
that the tin in the retort had gained 3-12 grains, whilst the
retort itself had not changed in weight. Lavoisier was
able to calculate that the weight of air which he had
sealed up in the retort was 15 \ grains, and therefore
concluded that one-fifth of this air had been absorbed by
the tin.

In a second experiment 20 ounces of tin were heated
during two and a-half hours in a large retort of 250 cubic
inches capacity. After allowing air to enter through a small
crack, the retort and its contents were found to be io'o6
grains heavier than when they were first weighed : the tin
had gained in weight by 10*00 grains or \ to \ of the
weight of air in the retort. The agreement between these
two figures could not be predicted, as the cracked retort
did not contain ordinary air ; but it is referred to below
as evidence that the ordinary air entering the retort had
almost the same density as the portion which had been
absorbed by the tin.

Lavoisier's figures are given on p. 37.

Lavoisier concluded :

" i. That only a limited quantity of tin can be calcined
in a given quantity of air.

1 When tin is heated in an excess of air, a white calx is produced.

THE BURNING OF METALS

37

voOO t^ 00 Q -1

r^oo oo oo p oo

,

8

CO

COOO Tj- t^ 1-1 CO

b

0

b

§ ? 1 ? §

„!

^ II II II II II II

II

II

II

§|

.c 1C

vp

If

O & : '• : : :

1

g

w

£

•*
1

1

8

1C

|: = 1 : :

oo

VO

tooO vO 1 vO OJ vO

CO

IN

-II

^•00 ^^ I OO O ****

CO

**O

'^ 1

vO vO II vO vo^

II

fg

u

II II II II II II

gS>

O

11

II
8;

^ Qu
X

o co :

Tf

co
VO

1

I

W

+

1

I

i

vO5

N

S^

'o : : : : :

CO

^

^

^ . - - -

vO

V3-

*

"rt "-S

•

:

:

: c : fed : :

:

:

:

rt C

c

*"O — *

•a

: ts • .S £

• o • *•*-• *+ CU

^ « § o

ID

1

"53

*

'o

o

o •""" -2 c •—

c3

t3

"5 " o 2^ *G ^

c

Jj

**

"T C ^3 "rt ^

c

0

-

.s | | 1* t

'5

"E

u< S "^ "^ *~ ^L

15

"53

^

0 " .J3 "o § ««

•g

^

5 ^ 5 ^ ^

^

c

fcjO tx fcjO

s

Cfl

'53 ^ '5 '53

*rt

y?

^ £ £

O

1—1

38 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

" 2. That the quantity of tin calcined is greater in a large
retort than in a small one. . . .

" 3. That the sealed retorts, weighed before and after the
calcination of a portion of the tin which they contain, show
no difference in weight, proving that the gain in weight of
the metal does not come from the fire nor from anything
outside the flask.

"4. . . . That in the calcination of tin, the gain in weight
of the metal is almost exactly equal to the weight of the
quantity of air admitted, proving that the part of the air
which combines with the metal during calcination has almost
the same density as that of atmospheric air " ( Works, II.
118-119).

Having thus shown that air was made up of two parts,
one of which was absorbed by tin, whilst the other was not
acted on, Lavoisier set himself to obtain separately the
part of the air which combines with metals and enables
substances to burn. He was able to imagine what
the properties of this gas must be : it must support com-
bustion, and probably substances would burn in it with great
ease. He tried to obtain it from metallic calces in which
he knew it was present, but was unable to set it free by
any of the methods at his disposal. The recovery of this
lost air, which Lavoisier required to complete the proof of
his theory of combustion, was, however, accomplished within
a few months by the independent work of Priestley in
England and of Scheele in Sweden.

B. THE DISCOVERY OF OxvGEN.1

Priestley (1774) discovers a gas richer than common
air. — The facts which Lavoisier required in order to com-
plete the proof of his theory of combustion were supplied

1 The discovery of oxygen followed that of most of the gases
described in Chapter V. It is described here in order to complete
the story of the calcination of metals.

in THE DISCOVERY OF OXYGEN 39

almost immediately by Joseph Priestley (1733— 1804), a
nonconformist pastor of Leeds, who was at this time making
a large number of experiments with the object of finding out
what "airs," or gases, were formed when various substances
were heated. Having procured a new burning glass twelve
inches in diameter, Priestley " proceeded with great alacrity
to examine, by the help of it, what kind of air a great
variety of substances, natural and factitious, would yield,
putting them into vessels .... filled with quicksilver, and
kept inverted in a bason of the same." (Experiments and
Observations on Different Kinds of Air, 1774, II. 28;
A.C.R. VII. 8.)

With this apparatus, on the ist of August, 1774,
Priestley endeavoured to extract air from " mercurius
calcinatus per se,," the RED CALX OF MERCURY prepared by
heating the metal gently in air. Having " found that by
means of this lens, air was expelled from it very readily,"
Priestley proceeded to examine the product, and discovered
to his great astonishment that " a candle burned in this air
with a remarkably vigorous flame .... and a piece of
red-hot wood sparkled in it, exactly like paper dipped in a
solution of nitre, and it consumed very fast " (A.C.R.
VII. 10). He placed a mouse in a vessel filled with the gas
where " it remained perfectly at its ease another full half
hour," twice as long as it would have lived in ordinary air
(A.C.R. VII. 17). Later he had the curiosity to breathe
the gas himself, and fancied that his " breast felt peculiarly
light and easy for some time afterwards " (A.C.R.
VII. 54). Priestley also obtained the gas by heating other
substances, including red-lead, a red powder formed by
gently roasting white lead, or litharge, and capable, like the
mercury calx, of being decomposed when heated more
strongly.

The only explanation which Priestley could give of his
discovery of a gas better and richer than air was that

40 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

ordinary air must contain some phlogiston, and that in this
new gas he had, for the first time, obtained air free from
phlogiston. He therefore described it as DEPHLOGISTICATED
AIR, and attributed its superior power of supporting com-
bustion, to the fact that it was capable of receiving more
phlogiston from burning substances, and could therefore
maintain a flame for a longer period than ordinary air.

Lavoisier makes quantitative experiments on the
calcination of mercury. — In October of the same year
Priestley, when on a visit to Paris, informed Lavoisier of
his experiments. Lavoisier concluded that the new gas
was the active part of the air which he had tried without
success to separate. In November, 1774, he repeated
Priestley's experiments, and in the spring of the following
year read before the Academy of Sciences at Paris a
memoir " On the Nature of the Principle which Combines
with the Metals during their Calcination and Increases
their Weight" (Works, II. 122). In this memoir he
described as VITAL AIR or EMINENTLY RESPIRABLE AIR the
gas obtained by heating the red calx of mercury, and
confirmed Priestley's observations as to its properties, but
without making any reference to the source from which
he had obtained his first information as to the behaviour of
the calx.

In his " Elementary Treatise on Chemistry," published
in 1789, Lavoisier described a series of experiments on the
calcination of mercury (Works, I. 35-38) which display to
the full his genius for exact and careful measurements. In
order to study the part that air played in the formation of
the red calx, he placed four ounces of pure mercury in a
retort, the neck of which was bent so that it passed
up into a bell-jar of air inverted over mercury (Fig. 16).
The total volume of air thus enclosed in the retort and the
bell-jar amounted to 50 cubic inches. He then heated the
retort, and observed the formation of a red scale on the

THE DISCOVERY OF OXYGEN

surface of the mercury contained in it. At the end of
twelve days, when further heating did not cause the forma-
tion of any more of the red scale, the retort was allowed to
cool. The mercury rose in the bell-jar and the total volume
of air was found to have been reduced to 42 or 43 cubic
inches, a loss of 7 or 8 cubic inches. The calx on the
mercury was collected and found to weigh 45 grains.
" The air which remained after this operation, and which had
been reduced to five-sixths of its volume by the calcination
of the mercury, was no longer fit for respiration nor
combustion ; since animals introduced into it perished in a
few moments, and lights were extinguished in it at once,
as if they had been plunged into water " (loc. at. p. 37).

As the mercury calx (unlike the black calx from the tin)
could be decom-
posed by heating
it, Lavoisier trans-
ferre d the 4 5
grains to a small
retort and ob-
tained from its de-
composition 41^
grains of mercury,
and 7 to 8 cubic
inches of gas,
which was " much
more fit than
atmospheric air to support respiration and combustion,"
since " a candle plunged into it, gave out a dazzling
light; charcoal, instead of burning quietly as in ordinary
air, burnt with a flame .... and a brightness of light
which the eye could scarcely bear" (loc. cit. pp. 37-38).

From these experiments it was clear that the whole of
the 7 or 8 cubic inches of air absorbed in the first experi-
ment had been liberated in an intensely active form in

FIG. 16 — LAVOISIER'S APPARATUS FOR HEATING
MERCURY IN A LIMITED VOLUME OF AIR.
No illustration is given of the retort used
afterwards to decompose the calx.

42 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

the second. A further experiment showed that a mixture
of 8 cubic inches of the respirable air with 42 cubic inches
of non-respirable air behaved in all respects like ordinary
air. There was, therefore, no doubt that the air which
disappeared during calcination had actually combined with
the mercury, and had been released from it by heating it
more strongly.

In order to complete the proof it was only necessary to
show that the loss in weight of the calx (45 grains), when
reconverted into mercury (41^ grains), was equal to the
weight of the gas which had been liberated : a knowledge
of the density of the active gas showed that the quantity
collected would weigh 3^ to 4 grains, agreeing closely with
the loss of 3 J grains already recorded.

Lavoisier completes his " oxygen " theory of com-
bustion.—Lavoisier's proof was now complete : he had
shown that the calcination of a metal meant the combina-
tion of the metal with an active constituent of the air,
which in the case of mercury could be recovered from the
calx by heating it : the calx was, therefore, a compound, and
the combustible metal a simpler substance. The same
constituent of the air was required for the burning of fuel
and for respiration, two processes which differed from the
burning of metals mainly in giving rise to gaseous instead of
solid products (see Chapter VI).

Lavoisier showed that the same constituent of air was
also concerned in the burning of sulphur and of phosphorus;
but the products of combustion differed completely from
the metallic calces, dissolving readily in water and producing
acid solutions. Lavoisier regarded his active gas as an
essential constituent of these and of all other acids. He
therefore selected for it the name OXYGEN or "acid-pro-
ducer" (compare German, sauerstof), deriving it "from
the Greek words o£us, acid, and yeivo/xat, I beget, on account
of the property of this principle, the basis of vital air, to

in THE DISCOVERY OF OXYGEN 43

change a great many of the substances with which it unites
into the state of acid, or rather because it appears to be a
principle necessary to acidity" (Chemical Nomenclature,
p. 24 ; compare Works, II. 249, where the name " oxygen "
is first used ) This name has been retained to the present day
in spite of the discovery (see Chapter XII) of acids in which
no oxygen is present. The compounds in which oxygen
was present were described as OXIDES ; under this title were
included, not only the acid oxides of sulphur and phos-
phorus, but also the basic calces derived from the metals.

By making use of the discovery of oxygen, Lavoisier
had been able to explain all the main facts on which the
phlogiston theory was based ; the phlogiston theory was
therefore no longer necessary, and gradually ceased to be
used. Lavoisier's theory was completed in the year 1777.
It won its first converts nearly ten years later, when it was
accepted by de Morveau, Fourcroy, and Berthollet. The
" System of Chemical Nomenclature " which they issued in
1787 was one of the first manifestoes of the new faith.

Scheele's discovery of oxygen. — The separation of the
gases of the atmosphere was accomplished independently
by the Swedish chemist Scheele in a research, "On Air
and Fire " (A. C. R. VIII), completed about the same time
as Priestley's work, but not published until 1777, and not
known either to Priestley or to Lavoisier at the time when
they were carrying out their experiments.

Scheele's experiments led him to the belief that common
air consisted of two gases, one of which could be removed
by various substances. Thus when damp iron filings rusted
in air, in a tightly closed bottle which was afterwards opened
under water, about a quarter of the air was found to be
absorbed. Similar results were obtained when phosphorus
was burnt in a thin flask tightly corked, or when it was
simply allowed to stand for six weeks in a closed flask until
it ceased to glow. The part of the air remaining at the

44 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

close of these experiments was slightly less dense than
ordinary air, and did not allow a candle to burn in it, or
even the smallest spark to continue glowing. This inactive
portion of the air was separated by no fewer than sixteen
different methods, but the recovery of the "lost air," as
Scheele called it, was a problem of much greater difficulty.
Scheele actually obtained it from a variety of substances,
amongst which was nitric acid, thus confirming the view of
Hooke and Mayow as to the close relationship existing
between nitre and air. He found that it could be prepared
most readily either from nitre (saltpetre) or from common
RED PRECIPITATE (obtained by dissolving mercury in nitric
acid and heating the residue until the product became red)
which he proved to be identical with calcined mercury (red
oxide of mercury).

Having separated this PURE FIRE-AIR Scheele repeated
the experiments which he had carried out with ordinary air
and proved that the substances which diminished the
volume of ordinary air when left or burnt in it, were
capable of absorbing " fire-air " entirely. For example,
phosphorus, heated in a closed flask filled with " fire-air,"
burnt with remarkable brilliancy ; on allowing the flask
to cool and opening it under water, it was found that
water entered the bottle and filled it almost completely.
Scheele showed that his " fire-air" (Lavoisier's "oxygen ")
was slightly denser than air, and soluble in water. He also
proved it to be essential to the breathing of animals, and
the growth of plants. By mixing it with 3 J times its volume
of " foul air," left after burning phosphorus in a closed
vessel, there was obtained a gas which in every respect
resembled common air. Scheele rightly concluded that
common air is a mixture of " fire-air " with about four times
its volume of " foul air."

Azote. — The residue left after removing the oxygen from
air was described by Priestley as " phlogisticated air," by

in THE DISCOVERY OF OXYGEN 45

Scheele as " foul air " (Air and Fire, p. 54), whilst Lavoisier
usually called it the "atmospheric mofette"; the French
chemists in 1787 (Chemical Nomenclature, p. 26) described
it as AZOTE, " from the Greek privative a and £o»j, life" in
order to indicate its inability to support life. It was shown
by Cavendish to consist mainly of a gas, present in
nitre, to which Chaptal gave the name NITROGEN. (See
Chapter X.)

SUMMARY AND SUPPLEMENT.

Jean Rey, in 1630, " On an enquiry into the cause wherefore
tin and lead increase in weight on calcination," concluded that
the gain in weight on calcining tin, lead, and antimony must be
due to condensation of air.

Robert Boyle, in 1673, in an essay entitled " New Experi-
ments to make Fire and Flame stable and ponderable," proved
that copper, iron, tin and lead gained in weight when burned,
but attributed this to the absorption of igneous particles. This
materialistic view of the nature of fire was elaborated in the
" phlogiston theory " of Becher and Stahl ; they assumed that
combustion and calcination involved an escape of phlogiston,
whilst in smelting an ore, or calx, the addition of phlogiston
from the fuel revivified the metal.

Robert Hooke, in 1665, in his " Micrographia," and John
Mayow, in 1674, in his "Medico-physical Essays," suggested
that air must contain an active principle similar to nitre ;
Mayow described this as " spiritus nitro-aereus," or " nitro-aerial
spirit." Mayow showed that damp gunpowder would continue
to burn under water. He made experiments with air trapped
over water, and showed that the volume of air was diminished
by the burning of a candle or of camphor, as well as by the
breathing of a mouse.

Antoine Laurent Lavoisier, in 1774, "Memoir on the
Calcination of Tin in Closed Vessels," showed that a similar
decrease of volume took place when lead and tin were heated
by a burning-glass in air confined over water or mercury. He
repeated the experiments in which Boyle had calcined tin in
sealed glass flasks, but proved that no change in weight took
place until the flask was opened ; as the flask did not change in

46 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

weight, the gain in weight of the tin must have been accom-
panied by the disappearance of an equal weight of air. When
the flask was opened a quantity of air rushed in, the weight of
which was almost equal to that gained by the tin ; this showed
that the part of the air absorbed by the tin (about £) did not
differ greatly in density from the air outside.

Joseph Priestley, in 1774, isolated the active part of the air
by heating the red calx of mercury, and showed that it was a
brilliant supporter of combustion and maintained respiration
much longer than ordinary air. He regarded it as air minus
phlogiston, and called it " dephlogisticated air." Lavoisier
then carried out quantitative experiments on the calcination of
mercury, in which he showed that the volume of gas set free
when the calx was heated was identical with the volume of air
absorbed in its preparation, whilst the weight of the gas set
free was equal to the loss in weight of the calx when reconverted
into metal.

Since sulphur and phosphorus were converted into acids by
combination with the active part of the air, Lavoisier gave to it
the name oxygen, i.e., acid-producer, the same idea being-
implied in the expressive German name, sauerstoff.

Oxygen was also prepared independently about 1774 by
Carl Wilhelm Scheele, who obtained it first by heating nitric
acid and nitre. He proved in many ways that air was a mixture
of an active and an inactive constituent, and showed that
ordinary air could be reproduced by adding " fire-air," or
oxygen, to the " foul air " remaining after rusting or burning
had taken place in it.

The chemical changes described in this chapter may be
represented by the following equations : —

Calcination of tin . . Sn + O2 = SnO2 (white calx).

(Stannic
oxide.)

2Sn + O2 = 2SnO (black calx).

(Stannous
oxide.)

„ „ lead . 2Pb + O2=2PbO (yellow calx).

(Litharge.)

-. 2Cu + O2 = 2CuO (black calx).

(Cupric
oxide.)

in THE DISCOVERY OF OXYGEN 47

Calcination of iron . 3Fe + 2O2 = Fe3O4 (blue-black scale or

(Magnetic calx.)
oxide of
iron.)

Preparation of red oxide of mercury —
(a) " Mercurius calcinatus per se "

(b) " Red precipitate "

(Mercuric (Nitric

nitrate.) oxide.)

2Hg(N03)2 = 2HgO + 4N02 + 02.

(Mercuric (Nitrogen
oxide.) peroxide.)

Decomposition of red oxide of mercury —

2HgO =
„ „ nitric acid —

4
„ „ nitre—

(Potassium (Potassium .
nitrate.) nitrite.)

Red lead contains more oxygen than litharge, but has not a
definite composition, and cannot be represented by any exact
formula. Its preparation and decomposition may be shown
thus—

where x is a fraction which is always less than 0-5.

CHAPTER IV

CHALK, LIME, AND THE ALKALIS

A. CHALK AND LIME

The burning of chalk to lime. — The changes wrought
by the action of fire have been described (Chapter I, p. 7)
as leading at an early period to the preparation of tin from
tinstone, iron from ironstone, and so forth. In addition to
these metallic ores, a number of rocks were known which
were converted by burning into a caustic substance, known
as LIME, which was used in the preparation of mortar.
Joseph Black (1728-1799), an Edinburgh • physican who
afterwards occupied the chairs of chemistry at Glasgow and
at Edinburgh, was one of the first to make a careful study of
these substances. He regarded chalk as the typical source
of lime, and included in the calcareous class of substances
" all those that are converted into a perfect quick-lime in
a strong fire, such as limestone, marble, chalk " and " those
spars and marls which effervesce with aqua fortis n
(A.C.R. I. 10).

The substance formed by heating chalk has remarkable
properties. When brought into contact with the skin, it
produces blisters and wounds resembling those caused by
fire. Hence it was known as a caustic. This property has
long been used for removing the hairs from hides in the
manufacture of leather. The lime drawn from the kilns is

CH. iv CHALK, LIME, AND THE ALKALIS 49

described as QUICKLIME because it becomes hot and steamy
when water is poured upon it ; the soft powder formed when
the hard lumps of quicklime are wetted with water, or are
left exposed to the air, is known as SLAKED LIME. Slaked
lime dissolves easily in acids and to a slight extent in water ;
the solution in water is known as LIME-WATER.

At the time when Black began his investigations, it was
generally believed that the caustic properties of lime were
due to the absorption of igneous particles from the fire in
which it had been burnt ; this view, like the phlogiston
theory of calcination, was found to be untenable as soon as
exact quantitative measurements were made. Black's experi-
ments were described in a paper " Experiments upon
Magnesia Alba, Quick-lime and some other Alkaline
Substances" (A. C. R. No. I.) published in 1755. They
were of importance on account of his success in solving the
problem of the relationship between chalk and lime, and also
because of the stimulus that they gave to exact quantitative
work. The principal points of this investigation are set out
in the following paragraphs.

Chalk loses in weight when burnt to lime owing to
the escape of fixed air. — Black found that "a piece of
perfect quick-lime made from two drams of chalk . . .
weighed one dram and eight grains " (A. C. R. I. 28), i.e.,
120 grains of chalk gave 68 grains of lime, a loss in weight
of 52 grains, or nearly 44%. This large loss in weight
could only be due to the escape of a gas, since nothing but
a little water could be condensed, when chalk was burnt to
lime in a retort.

" That the calcareous earths really lose a large quantity of
air when they are burnt to quick-lime, seems sufficiently
proved by an experiment of Mr. Margraaf. . . He subjected
eight ounces, of [chalk] to distillation in an earthen retort,
finishing his processs with the most violent fire of a
reverberatory, and caught in the receiver only two drams of

50 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

water, which by its smell and properties shewed itself to be
slightly alkaline " (A. C. R. I. 23).

Black had himself carried out the experiment of heating
magnesia (a compound resembling chalk, but more easily
decomposed by heat), and had " found only five drams of a
whitish water in the receiver," although the three ounces
(24 drams) of magnesia in the retort " had lost more than
the half of its weight" (A. C. R. I. 15).

From these experiments it was clear that the burning of
chalk to lime resulted in the escape of an invisible gas ; to
this gas Black applied the name FIXED AIR. He regarded
chalk as " a peculiar acrid earth rendered mild by its union
with fixed air." According to his view : —

" When the calcareous earths are exposed to the action of
a violent fire, and are thereby converted into quick-lime,
they suffer no other change in their composition than the
loss of a small quantity of water and of their fixed air. The
remarkable acrimony which we perceive in them after this
process, was not supposed to proceed from any additional
matter received in the fire, but seemed to be an essential
property of the pure earth " (A. C. R. I. 22).

The contrast between Black's theory and that which it
replaced may be expressed by the equations :

Old theory : Limestone + phlogiston = lime.
New theory : Limestone -fixed air = lime.

Lime combines with water and with fixed air. — In

the process of slaking, lime combines vigorously with water
to form slaked lime ; but in presence of fixed air it releases
the water, recombines with the fixed air and is reconverted
into chalk. Black writes : — -

" A calcareous earth deprived of its air, or in the state of
quick-lime, greedily absorbs a considerable quantity of water,
becomes soluble in that fluid, and is then said to be slaked :
but as soon as it meets with fixed air it is supposed to quit

iv CHALK, LIME, AND THE ALKALIS 51

the water and join itself to the air, for which it has a
superior attraction, and is therefore restored to its first state
of mildness and insolubility in water " (A. C. R. I. 24).

The conversion of soluble lime into insoluble chalk was
thus a test for the presence of fixed air. The test was most
sensitive when lime-water was used, because the chalk could
then be seen immediately. But in all important cases
Black preferred to rely on quantitative experiments in which
the chalk, after burning to lime, was recovered in such a
way as to show that the original weight of chalk had been
reproduced.

Fixed air is present in common air and in air
dissolved by water. — By noticing its action upon lime,
Black was able to show that a small quantity of fixed air
was dissolved in ordinary water, since : —

'' When slaked lime is mixed with water, the fixed air in
the water is attracted by the lime, and saturates a small
portion of it, which then becomes again incapable of dis-
solution, but part of the remaining slaked lime is dissolved
and composes lime-water " (A. C. R. I. 24).

The presence of fixed air in common air was also shown
by its action on lime-water, for : —

" If this fluid be exposed to the open air, the particles of
quick-lime which are nearest the surface gradually attract
the particles of fixed air which float in the atmosphere.
But at the same time that a particle of lime is thus saturated
with air, it is also restored to its native state of mildness
and insolubility ; and as the whole of this change must
happen at the surface, the whole of the lime is successively
collected there under^ its original form of an insipid
calcareous earth, called* the cream or crusts of lime-water"
(A. C. R. I. 24).

The fixed air forms, however, only a small proportion
both of ordinary air and of the air which is dissolved in
water, for(i) "lime-water, which soon attracts air, and forms

E 2

52 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

a crust when exposed in open and shallow vessels, may be
preserved, for any time, in bottles which are but slightly
corked, or closed in such a manner as would allow free
access to elastic air, were a vacuum formed in the bottle,"
and (2) under an exhausted receiver the same quantity of
air escapes from common water, and from water from
which the fixed air has been removed by the addition of
lime (A. C. R. I. 30).

Fixed air is liberated from chalk by the action of
acids, — Although Black was able to show that the loss of
weight in burning chalk to lime was due to the escape of
an invisible gas, it did not occur to him to attempt to
collect the gas, or even to render it visible by causing it to
bubble through water. But no special methods were
needed to render obvious the " violent breaking out of air "
which takes place when chalk is acted on by acids, nor the
absence of this effervescence when well-burnt quick-lime is
substituted for chalk. Remembering that
Chalk = lime + fixed air,

the contrast between the action of acids in the two cases
was sufficient to suggest that the gas so easily set free by
acids must be the same as the fixed air which escapes
when chalk is burnt to lime.

It is remarkable that Black did not attempt to test the
gas by passing it into lime-water,1 but apparatus such as is
now used for this purpose had not yet been devised, and
the systematic study of gases did not begin until nearly
twenty years later.

A very satisfactory proof of the identity of the two gases
was, however, obtained by showing .(i) that the weight of
gas which escapes is the same, whether the chalk is decom-
posed by acids or by heating it in a furnace ; and (2) that
the same weight of acid is required to dissolve the chalk

1 See footnote, p. 385, for Black's use of lime-water -as a test for
fixed air.

iv CHALK, LIME, AND THE ALKALIS 53

before and after burning it to lime. In the first experi-
ment Black saturated 120 grains of chalk with diluted spirit
of salt in a long-necked Florentine flask, and found a loss
in weight of 48 grains, as compared with 52 grains when
an equal weight of chalk was burnt to lime. In the second
experiment he found that 120 grains of unburnt chalk were
dissolved by 421 grains of diluted spirit of salt, whilst the
same quantity of chalk, burnt to quick-lime and slaked with
an ounce of water, required 414 grains of the acid, but
dissolved "without any sensible effervescence or loss of
weight" (A. C. R. I. 28).

Fixed air is present in mild alkalis such as soda and
potash. — It had been known from early times that the mild
alkalis resembled chalk in that they effervesced when acted
on by acids. This property was regarded as a chief
characteristic of the alkalis, — an idea that is preserved in
the use of the word " kali " to describe a mixture of sugar
with a mild alkali and a solid acid, which gives an
effervescent drink when added to water. As in the case of
the action of acids on chalk, Black did not test the gas by
passing it into lime-water, but he obtained an even more
satisfactory proof of the presence of fixed air in the alkalis
by showing that they could reconvert lime into chalk
quantitatively. This process had the advantage that it
rendered the use of an acid quite unnecessary in proving
the composition of the alkalis.

" A piece of perfect quick-lime made from two drams of
chalk, and which weighed one dram and eight grains, was
reduced to a very fine powder, and thrown into a filtrated
mixture of an ounce of a fixed alkaline salt, and two ounces
of water. After a slight digestion, the powder being well
washed and dried, weighed one dram fifty-eight grains.1
It was similar in every trial to a fine powder of ordinary

1 i.e. One hundred and twenty grains of chalk gave 68 grains of lime
from which 118 grains of chalk were recovered by the action of an
alkali.

54 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

chalk, and was therefore saturated with air which must
have been furnished by the alkali " (A. C. R. I. 28).

B. THE ALKALIS.

Mild alkalis are made caustic by the removal of fixed
air. — The preparation of CAUSTIC ALKALIS by the action of
caustic lime on the MILD ALKALIS, soda and potash, was
described by Geber. The caustic liquid l thus prepared
was used from an early period in the manufacture of soap by
the action of the hot alkali on fat ; it was, therefore, generally
called a " soap lye." Its caustic properties had been
attributed to lime dissolved in it ; but Black was not able
to prepare from it any trace either of chalk or of gypsum ;
he therefore concluded that " the acrimony of the caustic
alkali does not depend on any part of the lime adhering to
it" (A. C. R. I. 26).

On the other hand, Black found that a mild alkali made
caustic by lime was acted on by acids " without the least
effervescence or diminution of weight" (A. C. R. I. 32) nor
did it produce more than a slight cloudiness with lime-water.
The caustic alkali, when properly prepared, was thus free
both from lime and from fixed air.

The action of lime in rendering the alkali caustic was
therefore due, not to its ability to dissolve in the liquid or
to impart to it some fiery principle which it had acquired in
the lime-kiln, but on the contrary to its property of remov-
ing from the alkali the fixed air which had rendered it mild.

Caustic alkalis are rendered mild by exposure to air.—
Black found that after a fortnight's exposure to the air in an
open shallow vessel his caustic alkali " became entirely mild,
effervesced as violently with acids, and had the same effect
upon lime water as a solution of an ordinary alkali "

1 "A very hellish spirit, in which great mysteries lie hid" (Basil
Valentine, Last Will and Testament, p. 302).

iv CHALK, LIME, AND THE ALKALIS 55

(A. C. R. I. 32). The caustic alkali therefore resembled
slaked lime in its power of absorbing fixed air from the
atmosphere, but differed from it in its much greater
solubility in water.

On account of their great solubility the caustic alkalis
had not been isolated previously. Black attempted to
separate them by evaporating the caustic lye in an earthen-
ware bowl, but found that the inside of the bowl became
corroded and pitted with holes. By using a silver dish,
however, he succeeded in evaporating all the water, and
obtained the caustic potash as a fused mass which solidified
on cooling, but dissolved again on adding a small quantity
of water (A. C. R. I. 33).

At a later date, strong solutions of the caustic alkalis
were used in order to absorb (and so to collect and weigh)
fixed air from various sources, the greater concentration of
the solutions enabling them to take up much more fixed air
than in the case of lime-water; but Black, at that date, did
not understand the manipulation of gases, and had no
opportunity of making use of this valuable quality.

The alkalies are not easily decomposed by heat—
From the above observations it was clear that the relation-
ship between the mild and caustic alkalis was much the
same as that between chalk and slaked lime. Soda and
potash were closely analogous to chalk, from which they
differed chiefly in being soluble in water; whilst caustic
soda and caustic potash were similar to slaked lime, but
differed from it in their extraordinary solubility in water.

As potash parts with its fixed air so readily when mixed
with lime, it seemed probable "that alkalis might be
entirely deprived of their air, or rendered perfectly caustic,
by a fire somewhat weaker than that which is sufficient to
produce the same change " in chalk (A. C. R. I. 37). This
was found not to be the case : potash was " exposed, for
several hours, in a covered crucible," to the action of a

56 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

strong fire, black lead being added to soak up the potash
and so prevent it from corroding the vessel ; the potash
was found to " lose a part of its air, and acquire a degree of
causticity," but this was only slight, and it was not found
possible to convert the alkalis into a perfect caustic by the
action of heat alone. Black points out, however, that " the
alkali newly obtained from the ashes of vegetables is
generally of the more acrid kind " (A. C. R. I. 39) owing
to its partial conversion into a caustic alkali during
calcination.

The caustic alkalis, prepared in the form of solutions,
evidently correspond with slaked lime rather than with
quick-lime. But they are even less easily decomposed by
heat than the mild alkalis ; the form of the alkali corre-
sponding with quick-lime is therefore very difficult to prepare,
and was not discovered until Davy (Chapter XII) had
succeeded in isolating from the alkalis the very refractory
metals which they contain.

Sal volatile is a mild alkali : spirit of hartshorn is a
caustic alkali. — It had long been known that sal ammoniac,
the volatile salt described in Chapter I, could be converted
into a volatile alkali, SAL VOLATILE, by heating it with soda
or potash. This volatile alkali resembled the fixed alkalis
in that it effervesced when acted upon by acids. In this
action the volatile alkali was reconverted into a salt ; for
instance, sal ammoniac could be formed from sal volatile
and spirit of salt. These facts were used by Mayow in
1674 to prove that the alkali as well as the acid still exists
in the salt prepared by mixing them.

"Although [acids] and alkalis pass into a neutral
substance when they meet, yet they do not, as is generally
supposed, entirely destroy each other. For example, when
the acid spirit of salt is coagulated with a volatile
[alkali] .... although the mixed salts seem to be des-
troyed, yet they may be separated from each other with

iv CHALK; LIME, AND THE ALKALIS 57

their forces unimpaired, as takes place when sal ammoniac
.... is distilled with salt of tartar [i.e., potash] (A. C. R.
XVII. 160).

When sal ammoniac is acted on by lime a pungent gas
is liberated, to which, at a comparatively late date, the
name of AMMONIA was given. During the alchemistic period
this pungent vapour was (like spirit of salt) known only in
the form of a solution which was called the VOLATILE SPIRIT
OF SAL AMMONIAC. It was also prepared by distilling horn,
and was therefore known as SPIRIT OF HARTSHORN.1
Priestley, who was the first to collect spirit of salt as a gas
over mercury, was also the first to collect ammonia in the
gaseous state (see Chapter V).

Black showed that the "mild spirit of sal ammoniac"
(described above as sal volatile) could be rendered caustic
by removing its fixed air with the help of magnesia; it
then "emitted a most intolerably pungent smell," effervesced
only slightly with acids, and produced only a turbidity
when mixed with lime-water. It was therefore evident that
the pungent gas was a caustic alkali which could be rendered
mild by union with fixed air. Final proof of these facts
was supplied when Priestley showed that gaseous ammonia
could be combined with fixed air to produce solid sal
volatile, whilst with gaseous spirit of salt it united to form
sal ammoniac.

C. PREPARATION OF NEW EARTHS.

Earths can be prepared from salts by the action of
alkalis. — At the close of Chapter II it was shown how the

1 "Of the spirit and oil of Harts-horn. — Take Harts-horn, cut it
with a saw into pieces, of the bigness of a ringer, and cast in one at
a time into the aforesaid distilling vessel, and when the spirits are
settled, then another, and continue this until you have spirits enough :
and the vessel being filled with the pieces that were cast in, take them
out with the tongs, and cast in others, and do this as often as is
needful."— (Glauber, Philosophical Furnaces, Part II. ; Works, I. 51)

58 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

discovery of the acids led to the preparation of a number of
new salts. It may now be pointed out how the alkalis were
used from a very early period to prepare earths, or bases, from
the salts in which they were combined. Thus it was found
that gold, which could not be converted into a calx by
heating, was transformed into an earth by dissolving it in
aqua regia and precipitating again with an alkali. This
earth was decomposed when heated, leaving behind a
residue of " revivified " metal. By precisely similar methods,
silver (which likewise resisted calcination) was converted
into an earth from which the metal could be recovered by
the action of heat. At a much later date these unstable
earths were used by Scheele for the preparation of oxygen.
The earth prepared by precipitating gold from its solutions
by means of the volatile alkali, exploded violently when
heated, and was called AURUM FULMINANS or FULMINATING
GOLD ; its composition is discussed in Chapter XII.

These actions were studied by Mayow, who suggested
that the metallic earth was precipitated because the acid
had left it in order to combine with the stronger alkali, just
as in the action of oil of vitriol on nitre the alkali of the
nitre left it in order to combine with the stronger acid of
the vitriol.

Preparation of magnesia from Epsom salts. — The paper
"On Magnesia Alba," in which Black's experiments on
chalk and lime are described, derived its name from a white
earth, prepared by the action of alkalis on a variety of saline
liquids, and used as a mild aperient. It was first made from
MOTHER OF NITRE, the mother liquor left after crystallising
out saltpetre ; it was here present in combination with nitric
acid, which could be removed either " by the addition of an
alkali which attracted the acid to itself" or " by exposing
the compound to a strong fire in which the acid was
dissipated" (A.C.R. I. 6). Black prepared it from "the
bitter saline liquor called BITTERN, which remains in the

iv CHALK, LIME, AND THE ALKALIS 59

pans after the evaporation of sea- water," but he " afterwards
made use of a salt called EPSOM SALT, which is separated
from bittern by crystallisation, and is evidently composed of
magnesia and the vitriolic acid " (A. C. R. I. 7).

The magnesia prepared with the help of an alkali
effervesced with acids in just the same way as chalk, but
gave rise to a totally different series of salts ; in particular it
was dissolved by oil of vitriol, reproducing the Epsom salt
from which it had been prepared, instead of giving a
sparingly soluble residue of gypsum.

Like chalk, it was decomposed by heat, its weight
decreasing to a remarkable extent. The product resembled
lime in dissolving without effervescence in acids, but differed
from it in that it could not be slaked, and was not soluble
in water.

" An ounce [480 grains] of magnesia was exposed in a
crucible for about an hour to such a heat as is sufficient to
melt copper. When taken out, it weighed three drams and
one scruple [200 grains], or had lost 7/12 of its former
weight"

" I repeated, with the magnesia prepared in this manner,
most of those experiments I had already made upon it
before calcination, and the result was as follows :

" It dissolves in all acids, and with these composes salts
exactly similar to those described in the first set of
experiments : but what is particularly to be remarked, it is
dissolved without any the least degree of effervescence "
(A. C. R. I. 14).

This decomposition could be effected in a glass retort,
but although the magnesia " lost more than the half of its
weight," only 5 drams of water could be collected from
24 drams of the earth.

" We may, therefore, safely conclude, that the volatile
matter lost in the calcination of magnesia, is mostly air;
and hence the calcined magnesia does not emit air, or make
an effervescence, when mixed with acids."

60 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

The work upon " magnesia " was of special importance
because it provided Black with the knowledge which
enabled him to solve the more difficult problem of the
relationship between chalk and lime.

Preparation of an earth from alum. — Black also made
use of the alkalis in order to separate from alum an
earth to which the name of ALUMINA was afterwards given.
This earth is present in the alum as a vitriol, or sulphate. It
does not combine with fixed air, and this gas is therefore set
free when the alum is precipitated by means of a mild alkali.

Preparation of earths from salts by heat — A second
method of preparing earths from salts, which was practised
from the earliest period of alchemy, depended on dissipating
the acid of the salt by heat. When salts were thrown into a
charcoal fire, or were distilled from a retort, it was often
found that an acid vapour escaped, leaving behind an earthy
residue to which the name of CAPUT MORTUUM was given.
Amongst the salts distilled in this way were green vitriol and
alum, both used in the preparation of oil of vitriol ; the
residue of " colcothar " from the green vitriol was a red-brown
earth somewhat resembling rust; the alum left behind a
whitish residue of alumina. The nitrates were also
frequently decomposed, as for instance, in the preparation of
" red precipitate " by heating nitrate of mercury. Black
prepared magnesia in this way from mother of nitre ; and
Boyle in 1680 (Works, 1725, III, 372) prepared an alkali
from sea-salt by converting it into a nitrate and igniting this
with charcoal. As a further illustration, it will be shown in a
later chapter that fixed air may be regarded as an acid, and
the decomposition of chalk by heat as that of a salt into
acid and base.

The method of decomposing salts by heat, although very
simple, was not always available. Some salts were not
changed in the fire ; others (such as sal-ammoniac and salt)
were dissipated without leaving any residue : in the case of

iv CHALK, LIME, AND THE ALKALIS 61

silver and gold the residue left on heating a salt often con-
sisted of the revivified metal. In cases such as these the
base of the salt could only be separated with the help of an
alkali.

SUMMARY AND SUPPLEMENT
A. CHALK AND LIME

Joseph Black (1728-1799), in his "Experiments upon Mag-
nesia Alba, Quick-lime, and some other Alcaline Substances,"
read in June 1755, showed

(r) That chalk loses in weight by about 44 per cent, when
burnt to lime.

(2) That, as nothing but a trace of water could be condensed
by cooling the vapour (Margraaf), this loss in weight must be
due to the escape of gas, to which he gave the name fixed air.

(3) That lime combines with water to form slaked lime, but
releases it when it re-combrnes with fixed air to form chalk.

(4) That fixed air is present in small quantities in common
air and in air dissolved in water.

(5) That fixed air is liberated from chalk by the action of
acids ; apart from this liberation of gas, the action of acids upon
chalk and lime is identical both qualitatively and quantitatively.

These changes may be represented by the following equa-
tions :

Burning of chalk, CaCO3 -> CaO + CO2

chalk -> lime + fixed air

(Calcium (Calcium (Carbon
carbonate.) oxide.) dioxide.)

Slaking of lime, CaO-f OH2 -> Ca(OH)2

lime + water -> slaked lime

(Calcium hydroxide.)

Action of fixed air on slaked lime —

Ca(OH)2 + CO2 -> CaCO3 + OH2
slaked lime + fixed air -> chalk 4- water

Action of acids on chalk, slaked lime, and quicklime —

(CaCO3 + H9SO4 -» CaSO4 +
i. Oil of mtriol Ca(OH)2 + H2SO4 -> CaSO4 + 2H2
+H2SO4 -> CaSO4 + H2O

62 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

In each case, if dilute acid is used, solid gypsum or selenite,
CaSO4,2H2O, is produced.

(CaCO3 + 2HC1 -> CaCl2 + H2O-f CO2
2. Muriatic aa'd\Ca(OH)z + 2HC\ -> CaCl2 + 2H2O
ICaO +2HC1 -> CaCl2+H20

In each case a solution of muriate of lime (calcium chloride,
CaCl2) is produced.

B. THE ALKALIS

Boyle recognised an alkali by the facts that " it had a fiery
taste upon the tongue," that it would " make an ebullition with
acid spirits and precipitate diverse spirits made with them " and
would " turn syrup of violets green " (Producibleness of Chymical
Principles, 1680, pp. 35 and 37 ; Works, 1725, III. 372-373).

Black's observations on the alkalis are set out with modern
equations in the following paragraphs :

1. Mild alkalis contain fixed air since they are able to
re-convert lime into chalk —

K2CO3 + Ca(OH)2 -» 2KOH + CaCO3
potash + slaked lime -> caustic potash + chalk

(Potassium (Calcium (Potassium (Calcium

carbonate.) hydroxide.) hydroxide.) carbonate.)

2. The caustic alkali which is produced in this action con-
tains neither lime nor fixed air. It is very corrosive, but can be
separated by evaporating in a silver dish.

3. The caustic alkali absorbs fixed air from the atmosphere,
and is rendered mild thereby —

2KOH + CO2 -> K2CO2-r-H2O.

4. The caustic alkalis, unlike the mild alkalis, do not effer-
vesce with acids, e.g. —

KOH + HC1 -> KC1 +H2O

caustic potash + muriatic acid -> sal sylvii + water
K2CO3+ 2HC1 -> CO2 + 2KC1 + H2O
potash + muriatic acid ->fixed air + sal sylvii-h water

5. Sal volatile is a mild alkali, which can be rendered caustic
by means of lime—

(NH3)2CO2 + CaO-> 2NH3 +CaCO3
sal volatile + lime -> ammonia + chalk

(Ammonium carbamate.)

IV

CHALK, LIME, AND THE ALKALIS 63

C. PREPARATION OF EARTHS.

Black prepared magnesia alba (magnesium carbonate,
MgCO3) by the action of alkalis on —

1. Mother of nitre (containing magnesium nitrate),

Mg(NO3)2+K2GO3 -» MgCO3 + 2KNO3

2. Bittern (containing magnesium chloride),

MgCl2+K2CO3 -> MgCO3 + 2KCl.

3. Rpsom salts (magnesium sulphate),

MgSO4+K2CO3 -> MgCO3+K2SO4.

This magnesia is easily decomposed by heat, liberating fixed
air and losing more than half its weight,

MgC03-> MgO + C02.

It effervesces with acids, but after it has been burnt it dis-
solves without effervescence,

MgCO3+H2SO4 -> MgSO4 + H2O + CO2
MgO + H2SO4 -> MgSO4 + H2O

Magnesia, free from fixed air, was also prepared by igniting
the nitrate,

2Mg(N03)2 ->

CHAPTER V

THE STUDY OF GASES

A. FIXED AIR AND INFLAMMABLE AIR

Van Helmont recognises the existence of gases diffe-
ring from ordinary air.— The fact that metals, insoluble
in most liquids, are dissolved by acids, was known and
used from very early times. The liberation of gases during
this process must have been noticed from the first, but it
was not until the latter part of the seventeenth century that
any attempt was made to collect and examine these volatile
products. Van Helmont (1577-1644), to whom we owe
the name of " gas," recognised the existence of a poisonous
GAS SYLVESTRE, i.e. " wood-gas," which possessed the power
of extinguishing a lighted candle ; he detected it in the air
of a cavern, in the fumes from a charcoal fire, and as a
product of the fermentation of wine and beer; he also
recognised what he thought to be the same gas as a product
of the action of nitric acid on silver, and of distilled vinegar
on chalk. In contrast to this, he found in the large intes-
tine, and as a product of the fermentation of dung, an
inflammable gas to which he applied the name GAS PINGUE.
The names used by van Helmont were applied broadly to
two types of gas, analogous with the " choke-damp " and
" fire-damp " of miners ; no attempt was made to distinguish

64

v THE STUDY OF GASES 65

between one inflammable gas and another, or between one
poisonous gas and another.

Gases collected over water by Mayow (1674),— A few
years later, about 1674, the familiar method of collecting
gases over water was described by John Mayow, who used
it both for investigating the reduction in volume of air
during burning and breathing, and for examining gases
prepared artificially.1 For the latter purpose Mayow used a
flask inverted in a trough filled with diluted oil of vitriol
(Fig. 14, p. 33). The gas was produced by the action of
the acid on two or three iron spheres in the neck of the
flask. When nitric acid was used, much of the gas dissolved
in the liquid, and repeated action of the acid on the metal
was needed to fill the flask with gas. With oil of vitriol no
such contraction occurred. Mayow proved that these gases
possessed the same elastic properties as air,2 and that, when
added to a limited volume of ordinary air, they did not in
any degree prolong the life of a mouse confined in it ; but
no other tests were made, and Mayow could not be certain
"whether air of this kind is really common air or not"
(A. C. R. XVII. 113).

Cavendish prepares "Factitious Air" (1766),— Nearly
one hundred years intervened between the experiments of
Mayow and the appearance in the Philosophical Transac-
tions of the Royal Society for 1766 of a remarkable series
of three papers by the Honourable Henry Cavendish
(1731-1810) entitled "Experiments on Factitious Air."
Black had already investigated the part played by " fixed
air " in the burning of chalk to lime, and in other related
processes. But although he showed that fixed air was
present in the atmosphere, he did not attempt to collect or

1 As described by Boyle in 1660 (Works^ 1725, II. 432).

- Boyle's tract " Touching the Spring of Air and its Effects" (1660)
had recently appeared, so that this point was of special interest at
the time (see Chapter XV. p. 320).

F

66 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

handle it. It was therefore left to Cavendish to make the
first careful examination of those gases whose existence had
been vaguely recognised during the preceding one hundred
and fifty years. Cavendish writes :

" By factitious air, I mean in general any kind of air
which is contained in other bodies in an unelastic state,
and is produced from thence by art."

" By fixed air, I mean that particular species of factitious
air, which is separated from alkaline substances by solution
in acids or by calcination ; and to which Dr. Black has
given that name in his treatise on quick-lime (Phil. Trans.,
1766, 56, MI).

Describing next the action of acids on metals he writes :

" I know of only three metallic substances, namely, zinc,
iron and tin, that generate inflammable air x by solution in
acids ; and those only by solution in the diluted vitriolic
acid, or spirit of salt."

" Zinc dissolves with great rapidity in both these acids ;
and, unless they are very much diluted, generates consider-
able heat. One ounce of zinc produces about 356 ounce
measures of air 2 : the quantity seems just the same whichever
of these acids it is dissolved in. Iron dissolves readily in
the diluted vitriolic acid, but not near so readily as
zinc. One ounce of iron wire produces about 412
ounce measures of air: the quantity was just the
same, whether the oil of vitriol was diluted with ij, or
7 times its weight of water : so that the quantity of air
seems not at all to depend on the strength of the acid "
(ibid., p. 144).

Inflammable air was also obtained by the action of spirit
of salt on iron, but the quantity was not measured. One
ounce of tinfoil dissolved in strong spirit of salt yielded
202 ounce measures of inflammable air ; the same gas
was also produced slowly by the action of vitriol on tin.

1 The name " inflammable air " is here used for the first time.

2 The "ounce-measure" as used by Cavendish was the volume
occupied by an ounce of water.

v THE STUDY OF GASES 67

The identification of inflammable air and fixed air
by Cavendish. — To Cavendish belongs the credit of intro-
ducing the method of identifying gases by careful measure-
ments of their physical properties. For this purpose he
measured the density of inflammable air and of fixed air
from various sources ; in the case of fixed air the solubility
in water was also found. The densities were measured by
filling a bladder with the gas, and finding the change in
weight when the bladder was emptied. If the gas was
heavier than ordinary air the bladder became lighter when
emptied, but if the gas was lighter than air a gain in weight
was observed. In either case the density of the gas was
calculated from the known weight of the bladder-full of
air, and the change of weight on emptying the bladder.

In the case of inflammable air a gain of weight was
observed amounting in four experiments to 4of, 40 J, 41 J
and 41 grains, when using 80 ounce-measures of gas pre-
pared by the action of zinc on vitriolic acid and on spirit
of salt, of iron on oil of vitriol, and of tin on spirit of salt
respectively. The gas produced by the four methods was
therefore the same ; as air was 800 times lighter than water,
80 ounce-measures of common air weighed 80 x 480 l -=- 800 =
48 grains ; the same volume of inflammable air weighed
48 - 41 = 7 grains ; this gas was therefore 48 + 7 = 6-9 times
lighter than common air or 80 x 480 -r 7 = 5490 times lighter
than water.2

In the case of fixed air there was a loss in weight of
34 grains on a volume of 100 ounces ; the air weighed
i oo x 480 -v- 800 = 60 grains ; the fixed air weighed 60 4- 34 =
94 grains ; it was therefore iT5(/o times heavier than air or
511 times lighter than water.2 The values for fixed air
were trustworthy; those for inflammable air were un-

1 480 grains = I ounce.

2 The correct values are : for inflammable air, 14*4 and 11600 ; for
fixed air, I '53 and 530.

F 2

68 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

trustworthy, because they depended so largely on an exact
knowledge of the density of the common air displaced by
the bladder, e.g., taking the density of common air as being
850 (instead of 800) times less than that of water, Cavendish
found inflammable air to be "9,200 times lighter than
water, or ioT8o lighter than common air."

Cavendish devises a new method of
weighing gases. — In order to overcome
this difficulty Cavendish devised a method
of weighing directly the gas liberated by
the action of acids on a known weight of
metal or chalk. If the volume of the gas
had been determined in a separate ex-
periment the density could easily be calcu-
lated. Cavendish writes :

" I endeavoured to find the weight of the
air discharged from a given quantity of
zinc by solution in the vitriolic acid in the
manner represented in Fig. 17. A is a
bottle filled near full with oil of vitriol
diluted with about six times its weight of
water : B is a glass-tube fitted into its
mouth and secured with lute i1 C is a
glass cylinder fastened on the end of the
tube, and secured also with lute. The
cylinder has a small tube at its upper end
to let the inflammable air escape, and is
filled with dry pearl-ashes2 in coarse powder.
The whole apparatus together with the zinc, which was.
intended to be put in, and the lute which was to be used
in securing the tube to the neck of the bottle, were first
weighed carefully; its weight was 11930 grains. The zinc
was then put in, and the tube put in its place. By this

1 "The lute used for this purpose .... is composed of almond
powder, made into a paste with glue, and beat a good deal with a
heavy hammer. This is the strongest and most convenient lute I know
of" (loc. «?., p. 143, footnote).

2 /.<?., potash.

FIG. 17 — CAVEN-
DISH'S APPAR-
ATUS FOR FIND-
ING THE WEIGHT
AND DENSITY OF
GASES.

v THE STUDY OF GASES 69

means the inflammable air was made to pass through the
dry pearl-ashes ; whereby it must have been pretty effec-
tively deprived of any acid or watery vapours that could
have ascended along with it (Phil. Trans. 1766,^66, 153).

The loss in weight was nf grains, but one grain of this
was due to the displacement of the common air in the
bottle by inflammable air ; the true weight of the inflam-
mable air was therefore lof grains.

The weight of zinc used in this experiment was 254 grains.
A previous experiment (p. 66) had shown that one grain
of zinc gave 356 grain-measures of inflammable air.
The volume from 254 grains would therefore be
254x356 = 90424 grain-measures of gas.

By combining the two experiments it was seen that
i of grains of the gas occupied the same volume as
90424 grains of water. The gas was therefore 8410 times
lighter than water, or roj times lighter than common air.

Cavendish determines the weight of fixed air set
free by the action of acids on chalk and the alkalis,—
The same method was applied to find the weight of fixed
air liberated by acids from various substances, blotting-
paper being substituted for the pearl-ashes because the
greater weight of gas demanded less careful drying.
Cavendish found that :

1000 parts of marble lost 407 parts of fixed air.

1000 parts of sal volatile lost 528 to 538 parts of fixed air.

1000 parts of pearl-ashes lost 284 to 287 parts of fixed air.

1000 parts of a crystalline salt obtained by saturating
pearl-ashes with fixed air lost 423 parts of fixed air.

The quantity of acid required to liberate the fixed air
from each substance was also measured, and compared with
the weight of fixed air set free. It was found that the acid
which liberated 100 parts of fixed air from marble set free
109 parts from pearl-ashes, 217 from sal volatile, and
2 1 1 parts from the crystals prepared by the action of fixed

70 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

air on potash ; the two latter compounds thus contained
twice as much fixed air, relatively to their neutralising
power, as the two former.

Cavendish measures the solubility of fixed air. — In
the case of fixed air, its solubility in water was measured
and used to identify the gas. Cavendish states that :

" Water when the thermometer is about 55° will absorb
rather more than an equal bulk." " Water heated to the
boiling point is so far from absorbing air, that it parts with
what it has already absorbed " (Phil. Trans., 1766, 56, 163).

The gas was also lost by exposure to air.

By measuring the solubility, as .well as the density,
Cavendish was able to prove that the gas set free during
fermentation was identical with fixed air prepared from
chalk and the alkalis.

In his investigation of fixed air Cavendish introduced the
method of storing over mercury the gas which he had col-
lected by the displacement of water from an inverted bottle.
This use of mercury in place of water for manipulating
soluble gases proved to be of very great value in Priestley's
experiments a few years later.

B. GASES DERIVED FROM NITRIC ACID.

Priestley's work on gases. — Whilst Cavendish was the
first to make accurate measurements of the physical pro-
perties of gases, the discovery of their chemical properties
was mainly the work of Priestley. The state of knowledge
when Priestley began his work upon gases is described in
the following paragraphs :

" Van Helmont, and other chymists who succeeded him,
were acquainted with the property of some vapours to suffo-
cate, and extinguish flame, and of others to be ignited . • . .
But they had no idea that the substances (if, indeed, they
knew that they were substances, and not merely properties,
and affections of bodies which produced those effects) were

v THE STUDY OF GASES 71

capable of being separately exhibited in the form of a
permanently elastic vapour, not condensible by cold, to
which I give the name of air, any more than the thing that
constitutes smell. In fact they knew nothing at all of any
air besides common air, and therefore they applied the term
to no other substances whatever.

" Mr. Boyle was, I believe, the first who discovered that
what we now call fixed air, and also inflammable air, are
really elastic fluids, capable of being exhibited in a state
unmixed with common air ....

" Besides these two kinds of factitious air, that which I
call nitrous air obtruded itself upon Dr. Hales ; but even
he had no idea of there being more than one kind of air,
loaded with different vapours ; and was far from imagining
that they differed from one another so very essentially as
they are now known to do. And though Mr. Boyle, Dr.
Hales, and others, could not but be acquainted with the
effluvium of spirit of salt, and also of volatile alkali, they
could have no idea that the substance which had those
powers was capable of being separated from common air,
and of being exhibited free from moisture, in the form
of a permanently elastic vapour, to appearance exactly
like that which constitutes the common atmosphere ....

" Even Mr. Cavendish, whose experiments relating to air
immediately preceded my own, appears not to have had so
much as a suspicion of this kind. For he relates an
experiment of his, on the solution of copper in the marine
acid, as inexplicable, except on the hypothesis of there being
a kind of air that lost its elasticity by the contact of water,
which admits of the easiest solution imaginable, on the
supposition of the spirit of salt emitting a vapour, which
though capable of being confined by quicksilver, and of
being by that means exhibited in the form of air, was
instantly absorbed by water, which would thereupon become
possessed of all the properties of common spirit of salt.

" In fact, none of the chy mists appear to have had the
least idea of its being even possible to separate the acid or
alkaline principles from the water with which they are
always found combined ; and therefore, though they did
suppose them capable of further concentration, they still

72 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

considered a certain portion of water as absolutely essential
to them; and consequently all the experiments that have
hitherto been made on the affinities of the acids, and alkalis
are, in fact, nothing more than the affinities of compound
substances, consisting of adds or alkali, and water. I have
been so particular in stating these historical facts, for the
sake of those chymists who can see nothing new in my
experiments on the several acids and alkali, divested of
water, and exhibited in the form of air " (Experiments on
Air, 1777, Vol. III. pp. 325 et seq.).

Priestley (1772) prepares and examines nitrous air.—

The gas which Mayow prepared by the action of iron on nitric
acid was dismissed by Cavendish with the remark that iron,
zinc and tin, " dissolve readily in the acid and generate air ;
but the air is not at all inflammable." Priestley, however,
after examining fixed air and inflammable air, proceeded to
make a detailed study of this action. By dissolving various
metals (amongst them brass, iron, copper, tin, silver, mercury)
in nitric acid, Priestley collected over water a colourless gas,
which he named NITROUS AIR. (Experiments on Air, 1774,
I. 109). The gas was noxious to animals, and extinguished a
lighted taper. But it differed from fixed air in that it did
not precipitate lime-water and was only slightly soluble in
common water; water, he says, "absorbs one-tenth of its
bulk of nitrous air." The gas is, however, freely soluble in
a solution of green vitriol in water, which can be made to
"absorb more than ten times its bulk of nitrous air, without
any sensible approach to saturation." (Experiments and
Observations, 1779, IV. 48) ; this fact was afterwards used
by Humboldt and by Davy to test the purity of different
samples. The solution in which the gas has been absorbed
" becomes of a very dark colour ; but becomes green again
on being exposed to the open air " (Experiments on Air,
1777,111. Preface p. xxxiii). Davy showed that the green
colour could be restored and pure nitrous air recovered
by gently heating the solution ( Works, III. 99).

v THE STUDY OF GASES 73

Priestley studies the combination of nitrouj air with
common air and with oxygen. — The most remarkable
property of nitrous air was that of combining with common
air to form brown NITROUS FUMES 1 which dissolved at once
in water. Priestley made a careful study of this action and
showed that the diminution of volume occasioned by the
addition of an equal volume of nitrous air could be used
to measure the goodness of common air. Speaking of
nitrous air he says :

" One of the most conspicuous properties of this kind of
air is the great diminution of any quantity of common air
with which it is mixed, attended with a turbid red, or deep
orange colour, and a considerable heat "

" The diminution of a mixture of this and common air is
not an equal diminution of both the kinds . . . but of about
one-fifth of the common air, and as much of the nitrous air
as is necessary to produce that effect ; which, as I have
found by many trials, is about one-half as much as the
original quantity of common air. . . ."

" If, after this full saturation of common air with nitrous
air, more nitrous air be put to it, it makes an addition equal
to its own bulk, without producing the least redness, or any
other visible effect. . . ."

" It is exceedingly remarkable that this effervescence and
diminution, occasioned by the mixture of nitrous air, is
peculiar to common air, or air fit for respiration • and, as
far as I can judge, from a great number of observations, is
at least very nearly, if not exactly, in proportion to its fitness
for this purpose ; so that by this means the goodness of air
may be distinguished much more accurately than it can be
done by putting mice, or any other animals, to breathe in it"
(Experiments on Air, I774,2 I. 110-115).

1 These fumes were generally regarded as the vapour of nitric acid ;
Priestley described them as "nitrous acid vapour," Davy (1800) as
" nitrous acid gas," Gay-Lussac in 1809 as " nitric acid " and in 1816
as "nitrous acid"; in order to avoid confusion, the term "nitrous
fumes" is used in the succeeding pages. The name " nitrogen per-
oxide" was not introduced until about 1850.

2 The first part of Vol. I., published in 1774, describes experiments
carried out in 1772.

74 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

Priestley tests the goodness of common air by mixing
it with nitrous air. — Priestley's method of testing air
was described in detail a few years later as follows : —

" I first provide a phial, containing about half an ounce
of water, which I call the air-measure. This I fill with air
by having first filled it with water, and placed it over the
opening of the funnel in my shelf, and when it is filled I
slide it along the shelf, always observing that there be a
little more air than I want. The phial being thus exactly
filled with the air which I am about to examine, and care
being taken that it be not warmed by holding in the
hand, &c. I empty it into a jar about an inch and a half
in diameter, and then introduce to .it the same measure of
nitrous air, and let them continue together about two
minutes. I choose to have an overplus of nitrous air, that
I may be sure to have phlogiston enough to saturate all the
common air. If I find the diminution with these measures
to be very considerable, I introduce another measure of
nitrous air ; but the purest dephlogisticated air will not, I
believe, require more than two equal measures of nitrous
air."

"Sometimes I leave the common and nitrous air in the
jar all night, or a whole day ; but always take care that,
whatever kinds of air I be comparing together, they remain
the same space of time before I proceed to note the degree
of diminution."

" When the preceding part of the process is over, I
transfer the air into a glass tube, about three feet long, and
one-third cf an inch wide, carefully graduated according to
the air measure, and divided into tenths and hundredth
parts ; so that one of the latter will be about a sixth or an
eighth of an inch. Then immersing the tube in a trough of
water, so that the water in the inside of the tube shall be on
a level with the water on the outside, I observe the space
occupied by them both, and express the result in measures
and decimal parts of a measure, according to the graduation
of the tube " (Experiments and Observations, 1779, IV.
Introduction, pp. xxx. — xxxii.).

Using this method, Priestley in 1772 thought he perceived

v THE STUDY OF GASES 75

a slight inferiority in the air of his study as compared
with that outside, and in a sample of air from York as com-
pared with that of Leeds.

Two years later he discovered the gas, richer than com-
mon air, to which Lavoisier gave the name oxygen. On
adding nitrous air to the gas prepared from red precipitate
or from mercury calx, he found that it was about five times
" as good as the best common air " he had ever tested.

It was soon recognised that the diminution of volume of
common air was due to the absorption of oxygen, but the
variations which Priestley thought he had found in the
behaviour of different samples were proved to be due to
experimental errors. The careful experiments made by
Cavendish in 1783 showed that the composition of air
is remarkably constant, the value found for 100 volumes
of air being :

Diminution by nitrous air (oxygen) 20*84 volumes.
Residue (azote) 79' 1 6 volumes.

The graduated glass tube which Priestley used for
measuring the goodness of air is still known as a
EUDIOMETER (Greek cv, good, /uerpov, a measure: see
Priestley, Experiments on Air, 1777, III. 379 and 380).

Priestley (1777) describes the properties of nitrous
fumes. — On account of their solubility and corrosive pro-
perties, the brown nitrous fumes produced by combination
of nitrous air with oxygen, or by the action of strong nitric
acid on metals, could not be collected by the methods which
Priestley employed for other gases. He was, therefore,
obliged to collect the gas by displacing the air from narrow-
necked flasks provided with glass stoppers. This gas, as
Priestley prepared it by the action of nitric acid upon
bismuth, was obviously contaminated both with nitrous
air and with common air. He writes :

" Being disappointed, as has been seen, in my expecta-

76 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

tions of confining the nitrous acid vapour by animal oils,
it occurred to me, that, in lieu of this, it might not be wholly
without its use, if I could shut up this vapour in dry glass
phials, with ground stopples. And though, in this method
of procuring it, by the solution of bismuth, or other things
with which it unites most rapidly, there is necessarily a
mixture of nitrous air, it is inconsiderable in proportion to
the quantity of pure nitrous vapour itself. And although a
mixture of common air also would necessarily remain in the
phial, it could only serve to dilute the acid vapour, and
could not materially alter the properties of it. Also, if the
mouths of the phials were small, they might be opened, and
various substances admitted to the vapour, without much
loss of the acid ; especially as all acid vapours, I had reason
to think, were heavier than common air" (Experiments on
Air, 1777, III. 184).

The most conspicuous quality of the nitrous fumes is
their brown colour, by which their presence is disclosed in
many operations in which nitric acid takes part. The
colour is permanent, and has the remarkable property of
becoming much intensified by heating. This observation
interested Priestley so much that he was in the habit of
carrying a bottle of the gas in his pocket in order to show
its curious behaviour to his friends. Priestley writes :

" The change of colour given to this vapour by heat is
not a little remarkable, for it is altogether independent of
gravity or condensation. In order to make some experi-
ments of this kind to proper advantage, I procured a glass
tube, three feet long, and about an inch wide, closed at one
end, and fitted with a ground stopple at the other. This
tube I easily filled with red vapour, in consequence of its
being much heavier than common air ; and closing the
open end with the stopple, observed, that that part of the
tube which I held in my hand was manifestly of a deeper
colour than any other part of the tube. On this I held
one end of it to the fire, and found that that end grew
most intensely red, three or four times more so than the
rest of the tube. The direction in which the tube was

v THE STUDY OF GASES 77

held made no difference with respect to the red part of it ;
the part that was hottest being always of the deepest colour,
whether it was held upwards or downwards ; so that whether
the heated vapour ascended or descended, it did not retain
its colour in the smallest degree, after it had been opposite
to the heated part of the glass.

" That this extraordinary redness was not occasioned by
the vapour being more rarefied in that particular place,
appeared by the whole tube assuming the same deep red
colour when the whole length of it was made equally hot :
for the vapour being closely confined, the density of it
within the tube must necessarily have continued the same
in all the variations of heat or cold. This redness, there-
fore, must be the proper effect of heat, on the phlogiston,
as I should imagine, of the vapour. Repeating this experi-
ment very often, with the same tube, and the same vapour,
it became alternately of a deeper or lighter colour, according
as it was kept hot or cold, without any sensible change,
except that which depended upon this single circumstance.
This is really a striking experiment, and especially when
the tube contains first so much vapour as to be nearly
transparent when it is cold ; so that the heat alone gives it
all the colour that it acquires.

" In order to observe the utmost effect of heat on this
vapour, I placed the closed end of the tube near the fire,
and bringing it gradually nearer and nearer, observed that
the colour deepened uniformly with the increase of heat,
till, the glass actually melting, the compressed vapour burst
its way out" (Experiments on Air, 1777, III. 186-188).

Priestley (1772) discovers " diminished nitrous air " or
" laughing gas." — One of the agents used by Priestley
for " phlogisticating " or deoxidising common air was a
mixture of iron filings and brimstone, or sometimes iron
filings alone. The action of this agent on nitrous air gave
rise to a new gas possessing remarkable properties. Priestley
describes this discovery as follows :

" The diminution of common air by a mixture of nitrous
air, is not so extraordinary as the diminution which nitrous

78 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

air is subject to from a mixture of iron filings and brim-
stone, made into a paste with water. This mixture, as I
have already observed, diminishes common air between
one-fifth and one-fourth, but has no such effect upon any
kind of air that has been diminished, and rendered noxious
by any other process ; but when it is put to a quantity of
nitrous air, it diminishes it so much, that no more than
one-fourth of the original quantity will be left."

" Nitrous air thus diminished has not so strong a smell
as nitrous air itself, but smells like common air in which
the same mixture has stood" {Experiments on Air, 1774,
I. 118-119).

Priestley was astonished to find that an agent which he
had employed previously to remove all the goodness from
common air, had actually converted the inert nitrous air
into a DIMINISHED NITROUS AIR " with properties which,
at the time of my first publication on this subject, I
should not have hesitated to pronounce impossible, viz.
air in which a candle burns quite naturally and freely, and
which is yet in the highest degree noxious to animals,
insomuch that they die the moment they are put into it "
(ibid., p. 215).

As " diminished nitrous air " is soluble in cold water, it
was best prepared over mercury ; the gas was then almost
like oxygen in its power of supporting combustion. On
the other hand, long contact with iron over water destroyed
this property, the diminished nitrous air being either
absorbed or wholly deoxidised. Priestley found that :

" A candle burned with an enlarged flame ... in nitrous
air, which had been in contact with iron in quicksilver,
about six months.

" Water being admitted to the remainder of this air, it
began to be absorbed as usual.

" Nitrous air, which had been confined above a year in
contact with iron, standing in water, was, in all respects like
phlogisticated common air : it neither diminished common

v THE STUDY OF GASES 79

air, nor was diminished by nitrous air, and extinguished a
candle" (Experiments on Air, 1775, EL I77)-

The action of nitric acid on metals gives rise to
nitrous fumes, nitrous air, and diminished nitrous air,
— In the action of nitric acid on metals, diminished nitrous
air is often formed, especially if weak acid be used ; being
soluble in cold water, but expelled by heat, the gas may be
set free by warming cold nitric acid in which iron, zinc, or
tin has been dissolved.

"The solution of iron in spirit of nitre is known to
produce nitrous air ; but when all the nitrous air is pro-
duced in this manner, without foreign heat, if a candle be
applied to the solution, more air will be procured ; and
this will be possessed of the peculiar kind of inflammability
above mentioned.

" I have also procured this kind of air in a direct process
by the solution of zinc and tin."

" When ... I used only a very weak spirit of nitre
in the solution of zinc ... I got no other than this
kind of air in which a candle burned with an enlarged
flame ; and the air was of the very same kind from the
beginning to the end of the process " (" Of Nitrous
Air in which a Candle Burns," Experiments on Air,
1777, III., pp. 133, 134, 139).

Berthollet (1785) prepares " diminished nitrous air "
or " laughing gas " by heating nitrate of ammonia. —

Berthollet discovered in 1785 that when nitrate of ammonia
is heated to about i5o°C., it is decomposed into water and
" diminished nitrous air." This method of preparation was
used in the year 1799 by Sir Humphry Davy (1778-1829),
who found that the pure gas made by this method had the
following properties :

a. A candle burnt in it with a brilliant flame, and
crackling noise.

b. Phosphorus introduced into it in a state of inflamma-
tion, burnt with infinitely greater vividness than before.

8o HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

c. Sulphur introduced into it when burning with a
feeble blue flame, was instantly extinguished ; but when in
a state of active inflammation ... it burnt with a beautiful
and vivid rose-coloured flame.

//. Inflamed charcoal . , . burnt with much greater
vividness than in the atmosphere.

e. To some fine twisted iron wire a small piece of cork was
affixed : this was inflamed, and the whole introduced into
a jar of the air. The iron burned with great vividness, and
threw out bright sparks as in oxygen.

f. 30 measures of it exposed to water previously boiled,
was rapidly absorbed ; when the diminution was complete,
rather more than a measure remained.

g. Pure water saturated with it, gave it out again on
ebullition, and the gas thus produced retained all its former
properties..

h. It was absorbed by red-cabbage juice ; but no alteration
of colour took place.

/. Its taste was distinctly sweet, and its odour slight, but
agreeable.

/. It underwent no diminution when mingled with oxygen
or nitrous gas (Davy's Works, III. 54).

Davy (1799) uses " laughing gas " as an anaesthetic.—

Although the gas had been credited with the most deadly
properties, and with the power of producing plague and
other contagious diseases, Davy found from personal
experiments in the spring of 1799 that it could be breathed
without harm, and had indeed remarkable stimulating and
exhilarating qualities. Robert Southey, one of many friends
who submitted themselves to the action of the new.
intoxicant, describes his feelings as follows :

" My first definite sensation was a dizziness, a fullness in
the head, such as to induce a fear of falling. This was
momentary. When I took the bag from my mouth, I
immediately laughed. The laugh was involuntary, but
highly pleasurable, accompanied by a thrill all through me ;
and a tingling in my toes and fingers, a sensation perfectly
new and delightful " ( Works, III. 301).

v THE STUDY OF GASES 81

Effects such as these gave to the gas the popular name of
" laughing gas." Its permanent utility as an anaesthetic in
dentistry is forshadowed in Davy's own experience :

"The power of the immediate operation of the gas in
removing intense physical pain, I had a very good
opportunity of asertaining.

"In cutting one of the unlucky teeth called dentes
sapientiae, I experienced an extensive inflammation of the
gum, accompanied with great pain, which equally destroyed
the power of repose, and of consistent action.

" On the day when inflammation was most trouble-
some, I breathed three large doses of nitrous oxide. The
pain always diminished after the first four or five inspirations ;
the thrilling came on as usual, and uneasiness was for a few
minutes swallowed up in pleasure. As the former state of
mind however returned, the state of organ returned with it ;
and I once imagined that the pain was more severe after
the experiment than before" (Works, III. 276).

C ACID AIR AND ALKALINE AIR.

Cavendish (1766) collects "marine acid air," — From
the time of Glauber it had been customary to prepare
muriatic acid by distilling into water the pungent gas
produced by the action of strong oil of vitriol on common
salt (Chap. II, p. 14). The gas was first collected by
Cavendish in " an experiment with design to see, whether
copper produced any inflammable air by solution in spirit of
salt." He found that he "could not procure any inflammable
air thereby," but, on the application of heat, he was able to
drive off a considerable quantity of gas which had the
peculiar property of " losing its elasticity when in contact with
water." The gas set free in the bottle containing the copper
and acid (Fig. 18) was at first separated from the water
in the receiver by a "barrier of common air," but as soon
as this was removed the dissolution of the gas proceeded

G

82 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

rapidly, so that the water in the receiver " rushed violently
into the bottle and filled it almost entirely full " (Phil. Trans.
1766, 56, 157-158).

Priestley (1772) isolates "marine acid air. "—By making
use of mercury instead of water, Priestley had no difficulty
in obtaining this ACID AIR in a permanent form. He
showed that it could be driven off by heating spirit of salt
alone, the presence of copper being unnecessary. Its
peculiar behaviour depended entirely on its extreme solubility

in water, which he
found to absorb 5 76
times its volume of
the gas.1 Spirit of
salt was, indeed,
merely a solution of
acid air in perhaps
twice its weight of
water. Priestley also
prepared the gas by
the action of oil of
vitriol on salt, a more
efficient method

FIG. 18.— APPARATUS USED BY CAVENDISH TO PREPARE which had the ad-

A GAS WHICH " LOST ITS ELASTICITY BY CONTACT r i

WITH WATER." vantage of produc-

ing a dry gas. He

found that "acid air, extinguishes flame, and is much
heavier than common air" (Experiments on Air, 1774, I.

I43-I47)-

Priestley isolates "alkaline air." — Encouraged by his
success in separating an elastic air from spirit of salt,
Priestley next endeavoured to collect an ALKALINE AIR from
the "volatile spirit of sal-ammoniac." In this he was
successful : the gas which was set free by heating the

1 Two and a-half grains of water dissolved three ounce-measures
of the gas.

v THE STUDY OF GASES 83

volatile spirit remained permanently elastic when collected
over mercury, but collapsed immediately when brought into
contact with water. Priestley found that alkaline air was
only a little less soluble than acid air, one volume of water
dissolving 336 volumes of the gas.1 It also differed from
acid air in being lighter instead of heavier than common air.
He found that the gas (to which the name of ammonia was
given by Bergman in 1782), could be prepared directly by
heating crystals of sal-ammoniac with slaked lime in a gun
barrel : by passing the product into water he obtained a
very strong " spirit " from which he could expel the alkaline
;iir as required {Experiments on Air, 1774, I. 163-169).

Priestley prepares sal-ammoniac by mixing alkaline
air with acid air. — Thinking that gases possessed of such
opposite properties were likely to combine together to form
.1 neutral substance, possibly identical with common air,
Priestley brought together vessels containing alkaline air and
icid air. To his great astonishment he obtained a solid
oroduct which he identified as sal-ammoniac ; this volatile
salt was therefore a compound formed by the union of
alkaline air with acid air.

" Having satisfied myself with respect to the relation that
alkaline air bears to water, I was impatient to find what
would be the consequence of mixing this new air with the
other kinds with which I was acquainted before, and
especially with acid air ; having a notion that these two airs,
being of opposite natures, might compose a neutral air,
and perhaps the very same thing with common air. But
the moment that these two kinds of air came into contact,
a beautiful white cloud was formed, and presently filled the
whole vessel in which they were contained. At the same
time the quantity of air began to diminish, and, at
length, when the cloud was subsided, there appeared to be
formed a solid white salt, which was found to be the common

1 One and a-quarter grains of water dissolved seven-eighths ounce-
measures of gas.

G 2

84 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

sal-ammoniac, or the marine acid united to the volatile
alkali" (Experiments on Air, 1774, I. 169-170).

Sal- volatile formed by the union of alkaline air with
fixed air. — On mixing alkaline air with fixed air Priestley
again obtained a solid product which he identified as the
volatile alkali " sal-volatile."

" Fixed air admitted to alkaline air formed long and
slender crystals, which crossed one another and covered the
sides of the vessel in the form of net-work. These crystals
must be the same thing with the volatile alkalis which
chemists get in a solid form, by the distillation of sal-
ammoniac with fixed alkaline salts " (Experiments on Air,
1774, I. 171).

Apparatus for experiments on gases. — The first
apparatus used for experiments on gases, as distinguished
from condensible vapours, was that of Mayow, Figs, n, 12,
13, 1 6. It should be noticed that, in preparing gases
artificially by the action of acids on iron, Mayow was not
able to collect the gas over water, but was obliged to fill
the whole apparatus (Fig. 14, p. 33) with acid.

This disadvantage was removed by Stephen Hales
(1677 — 1761), botanist, chemist, and Vicar of Tedding-
ton, who was one of the first to separate the GENERATOR
from the RECEIVER of the gas. His Vegetable Statics (1727)
is chiefly concerned with hydrostatic experiments on the
pressure of the sap in plants. But the sixth chapter, occupy-
ing nearly one-third of the book, deals with "an attempt to
analyse the Air by means of a great variety of chymio-
statical experiments, which shew in how great a proportion
Air is wrought into the composition of animal, vegetable,
and mineral Substances, and withal how readily it resumes
its former elastick state, when in the dissolution of those
Substances it is disengaged from them."

In order to measure the volume of gas set free by heating
animal, vegetable, and mineral substances, Hales used a

THE STUDY OF GASES

tflass retort r (Fig. iqa) luted securely to a gauge ab,
standing in a trough of water xx. The gauge was

made from a long-
necked flask,
pierced at the
bottom to admit a
syphon-tube y, by
means of which air
could be drawn
out and water
sucked up as far

"IG. 19— (a) HALES'S APPARATUS FOR MEASURING THE
VOLUME OF GAS SET FREE BY HEATING ANIMAL,
VEGETABLE, AND MINERAL SUBSTANCES.

as z. The fall or
rise of the level
of the water in the
j;auge, after heating the contents of the retort, showed how
much air had been liberated or absorbed.

This apparatus had the disadvantage that the joint at a
Tvas liable to leak,
especially when an
iron retort had to
1) e used for
stronger heating,
or when the gas
had to be kept
for several days
before its volume
became constant.
These diffi-
culties were over-
come in a second
form of the ap-
paratus (Fig. 19^).
This consisted of
u retort rr made

an iron gun- FIG. i9-(J) HALES'S APPARATUS (Improved design).

86 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

barrel, with a lead syphon attached to the end. The gas
was carried by the syphon through a vessel of water xx
into an inverted flask ab, where it could be stored for any
length of time without risk of leakage.

The great importance of this new design will be realised
on comparing it with the apparatus used by Cavendish
(Fig. 1 8), by Lavoisier (Figs. 16, 21, 25), and by Priestley
(Fig. 20).

Priestley's apparatus. —The apparatus which Priestley
used for manipulating gases is described at the beginning
of the first volume of his Experiments on Air and illustrated
by two plates (numbered and lettered consecutively) which
are reproduced in Fig. 20.1

The most important feature was the trough, often called
a PNEUMATIC TROUGH, shown with all its accessories at i in
the first plate. The original trough a was of earthenware,
about eight inches deep, with thin flat stones, bb, just below
the surface ; afterwards a large wooden trough was used
with a shelf fixed an inch from the top. The various gases
were stored in cylindrical jars, cc, which could be immersed
in the trough, or stood on the shelf, or lifted out in dishes
as at 2 2 2. Mice were stored as at 3 in a receiver standing
on a perforated tin plate and provided with a perforated
cover held down by weights ; in order to test the goodness
of a sample of air, a mouse was held by the back of the
neck and passed through the water into a tall beer-glass, d, of
two or three ounces capacity, in which a mouse could
usually live from twenty to thirty minutes. The tapering
cork, 4, was used to close or open a small bottle inside a jar
of air, whilst the wire stand, 5, served to support small
dishes as at /. The funnel, 6, was used to pour gases from
a vessel with a wide neck into a vessel with a narrow neck,

1 In the second plate, the blackboard has been moved and the table
altered a little in order to reproduce on as large a scale as possible.

THE STUDY OF GASES

88 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

filled with water and supported near the surface of the
trough.

The gun-barrel, 7 (second plate : compare Hales, Fig.
19^), with a tobacco-pipe stem, or glass tube luted to the
open end, was used to expel gases from solid substances.
To reduce the volume of air, the gun-barrel was filled up
with dry sand. In the figure, the gas is being collected
over mercury ; Cavendish had transferred to a basin of
mercury, a bottle full of fixed air which was too soluble to
be stored long over water, but this is the first use of
mercury for collecting gases ; a carefully-shaped MERCURY
TROUGH was introduced by Lavoisier (Fig. 33). An
apparatus for . collecting over mercury gases expelled from
liquids by heat, or set free by the action of acids, is shown
at 8, where a is a basin of mercury, b a tube filled with it,
c the tube from which the gas is expelled and d a glass trap
to condense out moisture ; the apparatus used for generating
and collecting gases not freely soluble in water is shown at e
in the first plate.

The bladder 9, provided with funnel and delivery tube,
was used to receive gas from a jar standing in water and
then to transfer it free from water to a vessel standing in
mercury. A bladder is also shown in the apparatus, 10,
used for impregnating liquid with gas, e.g., water with fixed
air ; the gas generated in c was collected in the bladder and
then squeezed out through the flexible leathern tube, d, into
the flask a.

The candle, i2<z (first plate), mounted on the end of a
wire b, was used to test the air in the narrow tube 1 1 ; the
candle, c, was used in the wider cylinders, from which it could
be withdrawn through the water as soon as it was extin-
guished.

The syphon, 13 (second plate), was used to suck air out
from a cylinder as at /(compare Hales, Fig. iqa).

The receiver, 14, exhausted by an air-pump, was used

THE STUDY OF GASES

89

90 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

when a dry powder had to be surrounded by gas drawn from
a vessel standing in water.

The eudiometer? 15, was used for mixing minute quantities
of nitrous air and common air ; two bubbles were introduced
into the narrow tube a with the help of the iron-wire
plunger b, then measured, mixed, and measured again.

Apparatus for sparking small quantities of gas is shown in
the second plate at 16, 17, 18, 19, anticipating in part the
apparatus used by Volta (Fig. 27) and by Cavendish
(Fig. 39).

SUMMARY AND SUPPLEMENT
A. FIXED AIR AND INFLAMMABLE AIR

Van Helmont (b. 1577 ; d. 1644) recognised the existence of
two "gases," a poisonous "gas sylvestre" and an inflammable
" gas pingue," analogous with the " choke-damp " and " fire
damp " of miners.

John Mayow, Medico-physical Works, 1674, collected gases
prepared artificially by the action of iron on aqua fortis and on
dilute oil of vitriol contained in an inverted flask ; he showed
that they possessed the same elastic properties as common air,
but was not sure that they were essentially different from it.

Stephen Hales, Vegetable Statics, 1727, by heating different
substances, prepared many gases which he collected over water
in a separate "receiver"; he also mixed different gases and
studied the changes of volume produced in them by the action
of different agents ; but he did not recognise these gases
as distinct substances, and supposed that all had the same
density.

Joseph Black, Experiments upon Magnesia Alba, 1755,
proved that chalk and the alkalis contain a gas which he called
fixed air (carbonic anhydride or carbon dioxide, CO2), but he
did not at that time collect or handle it.

Henry Cavendish, On Factitious Air, 1766, measured the
density of fixed air from a variety of sources, and of in-
flammable air (hydrogen) prepared by the action of acids on
different metals, thus : —

1 This name was first used in 1776 (Experiments on Air], 1777, III.
379 and 380).

v THE STUDY OF GASES 91

Zinc on Vitriolic Acid, Zn + H 2S O4 ->Zn S O4 + H „ ;
Zinc on Spirit of Salt, Zn+ 2HCl->ZnCl2 + H2 ;
Iron on Vitriolic Acid, Fe + H2SO4->FeSO4+H2 ;
Tin on Spirit of Salt, Sn + 2HCl->SnCl2 + H2.

By these measurements he proved the identity of the
different samples, and established the method of investigating
gases by the exact measurement of their physical properties.
He found that water dissolved an equal volume of fixed air,
which escaped when the water was boiled or exposed to the
air ; fixed air could, however, be stored permanently in a bottle
inverted over mercury.

Joseph- Priestley, Experiments on Air, Vol. I. 1774 ; Vol.
II. 1775; Vol. III. 1777, prepared a large number of gases
and studied their chemical properties very thoroughly. He
examined fixed air and inflammable air, prepared three gases
by the action of nitric acid on metals, discovered oxygen, and
collected over mercury a series of gases which were too soluble
to be collected over water.

B. GASES DERIVED FROM NITRIC ACID

Nitrous air (nitric oxide, NO), was prepared by van Hel-
mont, by Mayow, and by Hales, by the action of nitric acid
on metals, but was first recognised as a distinct substance by
Priestley (1772), who named it nitrous air.

It dissolves in ten volumes of water, but is freely soluble in
solutions of green vitriol (ferrous sulphate), forming a dark
brown liquid from which the gas can.be recovered by gentle
heating. The preparation of the gas from copper and nitric
acid is often represented by the equation —

3Cu + 8HN03 -> 3Cu(NO3)2 + 2NO + 4H2O,

but many other products are formed at the same time. • .

Diminished nitrous air, or " laughing gas " (nitrous oxide,
N2O), was prepared by Priestley (1772) by the action of iron
filings on nitrous air.

Berthollet (1785) prepared it by heating ammonium nitrate.
(NH4)NO3 -> N2O + 2H2O.

92 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

It is readily soluble in water, but is expelled by boiling. A
candle burns in it with an enlarged flame, as in oxygen. The
pure gas can be breathed with impunity (Davy), and has been
used as an anaesthetic in dentistry since 1800.

The formation of soluble brown nitrous fumes (nitrogen
dioxide, NO2)on mixing nitrous air with common air had been
noticed by Hales. In presence of water the action is more
complex, both nitric and nitrous acid being produced (see
Chapter X), thus :

Without water 2NO + O2 = 2NO2

(Nitrogen dioxide.)

With water 4N O + 3O2 + 2 H2O = 4 H N O3

(Nitric acid.)

(Nitrous acid.)

When common air is mixed over water with an excess of
nitrous air (nitric oxide) all the oxygen is absorbed, together
with as much of the nitrous air as has combined with it ;
Priestley used the contraction produced by mixing equal volumes
of the two gases in a eudiometer as a measure of the "good-
ness" of air.

The brown nitrous fumes were prepared in an impure state
by Priestley by the action of bismuth on nitric acid —

Bi + 6HNO3 -

(Nitric (Bismuth (Nitrogen

acid.) nitrate.) dioxide.)

But this gas is better prepared by heating lead nitrate —
2Pb(NO§)2 -> 2PbO + 4NO2 + O2.

(Lead nitrate.) (Litharge.) (Nitrogen
dioxide.)

It can be purified by condensing it to a liquid. Priestley
noticed that the gas darkened in colour when heated, but
became light when cooled again. The loss of colour is due
to the formation of a more complex compound (dinitrogen
tetroxide) —

2NO2 =± N2O4,

(Brown.) (Colourless.)

which can be frozen out as a colourless ice (Chapter XX,
p. 525), but begins to decompose as soon as it is melted. The
mixture of these two oxides is called "nitrogen peroxide."

v THE STUDY OF GASES 93

C. ACID AIR AND ALKALINE AIR

Acid air (hydrogen chloride, HC1). Glauber prepared
" spirit of salt" by the action of oil of vitriol on salt,

2NaCl + H2SO4 -> Na2SO4 + 2HCl,

but he was only able to collect the gas by absorbing it in water.
Priestley (1774) collected it over mercury as a "permanently
elastic fluid." He found it to be extremely soluble in water,
which dissolved 576 times its volume of the gas.

Alkaline air (ammonia, NH3). The pungent gas present
in " spirit of hartshorn " was expelled by warming, and collected
over mercury by Priestley (1774), who called it "alkaline air" ;
the name "ammonia" was given to it by Bergman in 1782.
Priestley also prepared it from slaked lime and sal-ammoniac,

Ca(OH)2+ 2NH4C1 -> CaCl2 + H2O + 2NH3,
Slaked + sal — > muriate + water + ammonia
lime ammoniac of lime

and obtained a very strong spirit by dissolving it in water,
which absorbed 336 volumes of the gas. It differed from acid
air in being lighter instead of heavier than common air.
A mixture of alkaline air and acid air produced sal-ammoniac,

NH3+HC1 -> NH4C1 (sal-ammoniac).

(Ammonia.) (Hydrogen (Ammonium
chloride.) chloride.)

A mixture of alkaline air and fixed air produced sal-volatile,

2NH3 + CO2 -> (NH3)2CO2 (sal- volatile.)

(Ammonia.) (Carbon (Ammonium
dioxide.) carbamate.)

CHAPTER VI

THE COMPOSITION OF FIXED AIR, CARBON, CARBONIC ACID,
AND THE CARBONATES

A. THE COMPOSITION OF FIXED AIR.

Fixed air a product of combustion. — Van Helmont

derived the name of his poisonous " gas sylvestre " from its
presence in the fumes of a fire of wood-charcoal (Latin
sy/va, a wood or forest). But as the only test which he
applied was that of a lighted candle, no importance attaches
to the fact that he gave the same name to the gases
liberated in fermentation, and by the action of acids upon
chalk.

The recognition by Black, in 1755, of fixed air as a
constituent of chalk provided for the first time a test by
which a gas differing from common air could be detected and
identified with certainty. By its power of reconverting lime
into chalk, he proved the presence of fixed air in substances
such as magnesia, sal-volatile, and the mild alkalis, as well
as in natural waters and in common air. Cavendish, in
1766, added to these tests the method of measuring the
physical constants (density and solubility) of the gas, and
proved that these were the same for fixed air prepared from
alkaline substances and by the fermentation of sugar.
These physical tests could not be applied to the crude gas

94

CH. vi THE COMPOSITION OF FIXED AIR 95

obtained by burning charcoal in air ; but Cavendish showed
that "a quantity of common air was reduced from 180 to
1 60 ounce measures, by passing through a red-hot iron
tube filled with the dust of charcoal " and " observed, that
there had been a generation of fixed air in this process, but
that it was absorbed by soap leys " (quoted by Priestley,
Experiments on Air, I. 129).

Fixed air is formed during the burning of charcoal
(Priestley, 1772) and the reduction of metallic calces
(Lavoisier, 1774). — Priestley, in 1772, by means of a
burning mirror, or lens, heated fragments of charcoal in air
confined over water. He found that the volume of the air
was diminished by one-fifth, that the residue extinguished
flame, was noxious to animals, and suffered no further
diminution of volume when exposed to the action of iron
filings and sulphur, or when mixed with nitrous air ; the air
had therefore been deprived of all its ^goodness" (the
discovery of oxygen was not made until two years later)
whilst the production of fixed air was shown by the
cloudiness which was produced when the gas was confined
over lime-water. When mercury was used instead of water,
the volume of the air was not diminished by the burning
charcoal until lime-water was added, when one-fifth of the
air was absorbed (" Of Air infected with the Fumes of
Burning Charcoal," Experiments on Air, 1774, I. 129-132).

About the same time, Lavoisier (Physical and Chemical
Essays, 1774, Chapter V; Works I, 598-613) found that
a large volume of " air " was produced when minium or red
lead was reduced by heating with charcoal in an iron retort
(Fig. 21.) A very large quantity of gas was set free, 560
cubic inches being liberated in the production of about
f cubic inch of lead, although the charcoal heated alone
gave only eight cubic inches of gas in the course of two
days. The gas dissolved in lime-water, and extinguished a
lighted candle; a rat died almost instantly in it. The

96 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

FIG. 21— LAVOTSIER'S APPARATUS FOR COLLECTING THE GAS PRODUCED BY
HEATING RED-LEAD AND CHARCOAL IN AN IRON RETORT.

The air could be drawn out from the bell-jar through the small aperture shown
at the top of the jar, or it could be removed by means of an air-pump which has
not been reproduced in the figure.

vi THE COMPOSITION OF FIXED AIR 97

production of fixed air during the reduction of the calx
was thus clearly proved ; but it was not until Priestley had
discovered oxygen, as described in Chapter III, that
Lavoisier was able to give a satisfactory explanation of
what had occurred in this reduction,

Lavoisier (1774) proves that fixed air is an oxide of
carbon. — The crucial experiments by which fixed air was
proved to be an oxide of carbon were made by Lavoisier in
November, 1774, after Priestley had demonstrated to him
the method of making oxygen from the red calx of mercury.

By heating an ounce of this calx with forty-eight grains of
charcoal in a tiny glass retort he was able to collect, in a
bell jar over water, 64 cubic inches of gas. This gas dis-
solved in water, communicating to it the properties of the
natural acidulated waters, destroyed animals brought into it,
at once extinguished candles and other burning substances,
precipitated lime-water, and combined readily with the
alkalis, removing their causticity and enabling them to
crystallise. " All these properties are exactly those of the
kind of air known as fixed air."

Lavoisier found that the red precipitate heated alone
gave 78 cubic inches of a gas which did not dissolve in
water, did not precipitate lime-water, did not unite with
alkalis or diminish their caustic qualities, but which could be
used again for the calcination of metals. " In conclusion,
it had none of the properties of fixed air ; far from being
fatal, like it, to animals, it seemed, on the contrary, more
proper for the purposes of respiration ; candles and burning
bodies not only were not extinguished by it, but burned
with an enlarged flame in a very remarkable manner ; the
light they gave was much greater and clearer than in
common air ; charcoal burned in it with a brilliancy
almost like that of phosphorus, and all combustible sub-
stances were consumed in it with surprising rapiaity. All
these circumstances convinced me that this air, far from

H

98 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

being fixed air, was even more respirable, more combustible,
and consequently more pure even than the air in which we
live."

The calx heated alone gave oxygen and mercury ; when
heated with charcoal it gave fixed air and mercury. Fixed
air was therefore evidently a compound of charcoal with
oxygen. Lavoisier argued that —

" Since carbon disappeared entirely in the revivification of
mercury from its calx, and since there result from this
operation only mercury and fixed air, one is forced to
conclude that the principle which has hitherto been known
as fixed air is the result of the combination of eminently
respirable air with charcoal" ("On the Nature of the
Principle which combines with metals during their Calcina-
tion and increases their Weight," Works, II. 122-128).

The charcoal (French charbon) used in experiments such
as these always left behind when burnt a larger or smaller
proportion of ash. Lavoisier and his colleagues, therefore,
introduced in 1787 "the modified name of CARBON/ which
indicates the pure and essential principle of charcoal," thus
"distinguishing it from charcoal according to the vulgar
acceptation " and isolating it "from the small quantity of
foreign matter which it generally contains and which
constitutes the ash" (Method of Chemical Nomenclature,
tr. 1 788, p. 32).

In their system of nomenclature, fixed air became an
OXIDE OF CARBON : its production during the burning of
charcoal was represented by the equation :
carbon + oxygen = fixed air

(oxide of

carbon)

whilst the reduction of a calx was shown by the equation :
carbon + calx —fixed air + metal.

(oxide of (oxide of

metal) carbon)

French carbone.

vi THE COMPOSITION OF FIXED AIR 99

Diamond and graphite.— Newton had suspected (Opticks,
1704, II. 75) that the diamond, with its high refractive
power, might be a combustible substance. This idea was
confirmed by several notable experiments in which the
combustion was effected both by powerful burning-glasses
and by means of furnaces. But as the experiments were
always made in the open, no idea could be formed of the
nature of the products ; indeed, it was generally believed
that the diamonds had merely been vaporised without
burning at all.

To Lavoisier belongs the credit of having shown, in 1772,
in conjunction with Macquer and Cadet, that if air be com-
pletely excluded, the diamond remains unaltered at the
highest temperature of the furnace ; the burning of the
diamond is therefore a true combustion. In order to deter-
mine the nature of the products of combustion, Lavoisier
employed a very large burning glass (compare Fig. 22) to heat
diamonds supported in air or oxygen contained in glass
jars inverted over water or over mercury. No water, smoke,
or soot was produced ; when mercury was used the volume
of gas was not altered, but in contact with water the volume
was somewhat diminished; in both cases the gas in which
the diamond had been burnt turned lime-water milky
("Destruction of the Diamond by Fire," 1772, Works, II.
38-88). Charcoal behaved in just the same way as the
diamond. There was, therefore, no doubt that each of
these substances gave rise to fixed air as the sole product of
combustion. A few years later, in 1797, Tennant (an
English chemist who had already shown that carbon could
be recovered from fixed air by the action of phosphorus
vapour on red hot chalk), burned diamonds by means of
melted saltpetre, and showed that they gave rise to precisely
the same quantity of fixed air as when charcoal was used. This
observation was confirmed by the combustions carried out
many years later by Dumas and Stas (Chapter VIII, p. 150) ;

H 2

ioo HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

vi CARBONIC ACID AND THE CARBONATES , 101

it showed clearly that the diamond, like charcoal, must be
regarded as a variety of carbon.

The mineral GRAPHITE, also known as PLUMBAGO or
BLACK LEAD, was shown by Scheele, in 1779, to give rise to
fixed air when fused with nitre. Other workers showed that
(if allowance were made for the ash which it always
contains) it gave the same proportion of gas as charcoal and
diamond, and must therefore be regarded as a third variety
of carbon,

Such varieties of an element as charcoal, diamond, and
graphite were described by Berzelius, in 1840 (Jahresbericht^
20, Chem. p. 13), as ALLOTROPES.

Bi CARBONIC ACID AND THE CARBONATES.

Bergman (1774) regards fixed air as an acid. — It is

uncertain to what extent Black recognised that fixed air had
the properties of a weak acid. He was aware that it was
attracted by the caustic alkalis and by lime, and that it
blunted and diminished their caustic and alkaline properties.
But the small quantity of liquid which could be condensed
when distilling chalk or magnesia differed so completely
from the mineral acids produced by distilling green vitriol
or nitre, that it is not surprising that Black hesitated to
describe fixed air as an acid.

The clear recognition of its acid properties was due to the
Swedish chemist Torbern Bergman (1735-1784), who in
1774 published a full description of its properties in a paper
"On the Aerial Acid " (Bergman's Essay 's, translated by
E. Cullen, 1784, pp. 1-90).

In support of this view he mentions :

(1) Its solubility in water, as measured by Cavendish
(Chapter V).

(2) The acid taste which it imparts to natural and
artificial aerated waters, as noticed by Brownrigg and by

io'2 HISTORICAL, INTRODUCTION TO CHEMISTRY CHAP.

Priestley. In a paragraph headed " Fixed Air has an Acid
Taste," Bergman writes :

" As this air is in form of an elastic vapour, it can hardly
be tasted by itself, at least distinctly ; but if it be united with
water, which is in itself void of flavour, being accumulated
and rendered less volatile by this union, it readily affects
the tongue with a weak but agreeable acidity. This is the
real spirit of the cold mineral waters, which undoubtedly
occasioned them to be called acidulous ; and by means of
which, together with a due proportion of suitable salts, we
may perfectly imitate the Seltzer, Spa, and Pyrmont waters.
Such artificial waters I have now been using for eight vears
with signal advantage" (JSssays, p. 12).

(3) Its power of imparting to litmus a transient red
coloration :

" Syrup of violets, and such other blue vegetable juices
as I have hitherto tried are not reddened by fixed air,
the tincture of tournsole [litmus] is of all known tinctures
most easily acted on by acids, therefore the slightest
vestiges, which cannot by any other means be discovered,
are by this tincture easily detected " (Essays, p. 16).

Thus if water " be tinged with tournsole to a perfect blue,
when fixed air sufficient to fill about the 1/50 of the vessel has
passed through it, it will be manifestly red ... In like
manner one part of water, saturated with fixed air, makes 50
parts of the above tincture distinctly red. This change of
colour, however occasioned by the fixed air, soon disappears
in an open vessel, particularly if it be exposed to heat, or
the rays of the sun ; a circumstance which indicates the
volatile nature of the acid that produces the change
(Essays, pp. 14-15).

(4) Its power of combining with mild alkalis to form
neutral crystalline compounds, as discovered by Black and
confirmed by Cavendish (pp. 69 and 108).

(5) Its power of dissolving chalk and magnesia, dis-
covered by Cavendish (p. 103).

(6) Its power of dissolving iron (giving rise to artificial

vi CARBONIC ACID AND THE CARBONATES 103

chalybeate waters, as discovered by the English apothecary %
Lane in 1769), and of dissolving zinc.

(7) Its power of precipitating substances dissolved in
alkalis, e.g. sulphur dissolved in lime-water.

On account of the acid properties of its solutions the
French chemists in 1787 gave to the gas the name CARBONIC
ACID.

" As fixed air has been perceived to be produced by the
direct combination of charcoal with vital air, by the assist-
ance of combustion, the name of this gaseous acid can no
longer be arbitrary, but necessarily must be derived from its
radical, which is the pure carbonic matter ; therefore it is
carbonic acid and its compositions with different bases are
carbonates" (Chemical Nomenclature, tr. 1788, p. 32).

But as Bergman recognised, this acidity belongs to the
solution rather than to the gas. For this reason the name
" carbonic acid " is now restricted to the solution of the gas
in water, whilst the gas itself1 is called CARBONIC ANHYDRIDE.
The recognition of carbonic acid as an acid necessitated a
change in the classification of the mild alkalis, which were
now regarded as salts under the name of CARBONATES. In
later years the term ' alkali ' became associated almost ex-
clusively with the caustic alkalis, instead of the mild effer-
vescent salts to which the name had been applied for a
thousand years previously.

Cavendish (1767) discovers that chalk and magnesia
are rendered soluble by fixed air. — To Cavendish belongs
the credit of discovering that chalk is rendered soluble in
water by the presence of fixed air. This discovery was made
in the course of his " Experiments on Rathbone-Place
Water" (Phil. Trans., 1767, 57, 92). He found that 494
ounces of a water, which on boiling liberated one-seventh of
its volume of fixed air,2 deposited 271 grains of a calcareous

1 Following Laurent (1854) ; see Chapter XII.

2 411 ounces of water liberated 66 ounce-measures of fixed air and
8J ounce-measures of common air.

104 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

earth consisting almost entirely of chalk. That chalk was
actually held in solution by fixed air was proved by mixing
lime-water with an excess of water saturated with the gas,
when a clear solution was produced, whereas the use of less
fixed air caused the lime to be precipitated in the form of
chalk. Cavendish writes :

" Calcareous earths, in their natural state, i.e. saturated
with fixed air, are totally insoluble in water ; but the same
earths, entirely deprived of their fixed air, i.e. converted
into lime, are in some measure soluble in it ; for lime-water
is nothing more than a solution of a small quantity of lime
in water. It is very remarkable, therefore, that calcareous
earths should also be rendered soluble in water, by furnish-
ing them with more than their natural proportion of fixed
air, i.e. that they should be rendered soluble, both by
depriving them of their fixed air, and by furnishing them
with more than their natural quantity of it. Yet strange as
this may appear, the following experiments, I think, show
plainly it is the real case."

" A bottle full of rain water was inverted into a vessel of
rain water, and some fixed air forced up into the bottle, at
different times, till the water had absorbed as much fixed air
as it could readily do ; 1 1 ounces of this water were mixed
with 6J of lime water. The mixture became turbid on first
mixing, but quickly recovered its transparency, on shaking,
and has remained so for upwards of a year."

" Lest it should be supposed, that the reason why the
earth was not precipitated in the foregoing experiment, was,
that it was not furnished with a sufficient quantity of fixed
air, the following mixture was made, which contains the
same proportion of earth as the former, but a less proportion
of fixed air: 4! ounces of the above-mentioned water,
containing fixed air, were diluted with 6J of rain water, and
then mixed with 6| ounces of lime-water. A precipitate was
immediately made on mixing, which could not be re-dissolved
on shaking" (Phil. Trans., 1767, 57, 101 and 104-105).

This observation was confirmed by Lavoisier, and also by
Bergman, who noticed that if

vi CARBONIC ACID AND THE CARBONATES 105

"a small portion of lime-water be dropped into water
impregnated with fixed air, slight clouds are immediately
formed, occasioned by the saturation of the lime by the
fixed air; these clouds,, however, disappear upon gently
shaking the vessel, the lime being dissolved by the
superabundant fixed air " (Essays, p. 34).

Bergman also recorded the fact that crystals of calcareous
spar (Fig. 23) were dissolved by water containing fixed air, and
that minute crystals of similar shape were sometimes deposited

FIG. 23 — CRYSTALS OF CALC-SPAR. British Museum (Natural History).

when chalk was reprecipitated by the escape of the fixed air
which held it in solution. These observations were
important as affording a clue to the way in which crystalline
calc-spar may have been produced ; they also afforded a
satisfactory explanation of the formation of stalactites (Fig.
24) and stalagmites by the escape of fixed air from the
dripping water of caverns in limestone districts.

The hardness of water due to salts of lime and
magnesia. — The mineral matter held in solution by fixed
air makes a water " hard," i.e. the water gives a curd with
soap instead of a lather. As this hardness is removed

io6 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

when the water is boiled it is called TEMPORARY HARDNESS.
Cavendish showed that the whole of the chalk in such a
water could be " precipitated ... by the addition of a
proper quantity of lime-water" (loc, cit, p. 107). This
method of removing the fixed air and so precipitating the
chalk from a hard water is used on a large scale in order to
" soften " the water-supply in chalky districts such as that of
the Caterham Valley.

Magnesia, of which Cavendish detected a small quantity

mixed with the
chalk from Rath-
bone Place water,1
also owes its solu-
bility to fixed air.
It dissolves to a
larger extent than
chalk, and the
solution is used in
medicine as a
gentle alkali under
the name of " fluid
magnesia." Like
chalk, the mag-
nesia produces a
temporary hard-
ness, which can
be removed by
boiling, or by the

addition of lime ; but the precipitation of the magnesia, and
the softening of the water, takes place much more slowly
than in the case of chalk.

Gypsum or selenite (of which Cavendish collected 39

1 By acting on the earthy precipitate with oil of vitriol the chalk
was converted into sparingly-soluble selenite or gypsum, and the
magnesia into Epsom salts, of which 18 grains were extracted.

FIG. 24 — STALACTITE, i.e., CHALK DEPOSITED BY THE
ESCAPE OF FIXED AIR FROM DRIPPING WATER.
British Museum (Natural History).

vi CARBONIC ACID AND THE CARBONATES 107

grains by evaporating 494 ounces of the water to a bulk of
three ounces) shares with all the soluble salts of lime the
property of rendering water hard. As the gypsum is not
precipitated when the water is boiled, the hardness which it
produces is called PERMANENT HARDNESS. This permanent
hardness can be removed by adding to the water a mild
alkali such as soda or potash, which precipitates the lime in
the form of chalk and so renders the water soft. This is
the method generally used for softening permanently hard
waters.

Soluble salts of magnesia (of which Cavendish precipitated
36 grains by the addition of fixed alkali) also contribute to
the permanent hardness of water. Sea water, which contains
a little gypsum, and a relatively large proportion of Epsom
Salts, is so hard that it is practically impossible to make it
lather, even by using a very large amount of soap. The
precipitation of soap by the mineral matter present in hard
waters was discussed by Bergman (".Of the Analysis of
Waters," Essays, pp. 90-192), who made use of a solution
of soap in alcohol as a means of detecting such substances
in water, thus,

"Soap is not soluble in every kind of water; this is
occasioned either by a disengaged acid, or by a large pro-
portion of ... salt with an earthy or metallic base . . . :
such waters are generally called hard waters, and are unfit
for washing cloths, as also for boiling pulse, and the harder
kinds of flesh."

" If there be present in a can of water but 8 grains of
alum, [muriate of] magnesia or [muriate of] lime, a single
drop of this water occasions a turbidness in a solution of
soap in alcohol, diluted with an equal bulk of distilled
water" (Essays, p. 139).

Lane (1769) discovers that fixed air will dissolve
iron. — Fixed air also possesses the power of dissolving iron,
giving rise to chalybeate waters. In these waters the iron

io8 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

is held in solution by fixed air in much the same way as
the chalk and magnesia in the water investigated by
Cavendish. This property of fixed air was discovered by
T. Lane, an English apothecary, who described his experi-
ments in the Philosophical Transactions for 1769, as follows :

"A wide-mouthed bottle, containing half a pint of
distilled water and sixty grains of steel filings, was suspended
forty-eight hours over some distiller's molasses, in brisk
fermentation ; so as to receive the fixed air escaping from
the fermenting liquor ; the surface of which was ten inches
below the mouth of the bottle. Immediately after its
removal, the clear water was decanted from the filings and
ochrous sediment."

" This liquor had a brisk and ferruginous taste, with a
flavour of the molasses. An infusion of galls, or green tea,
soon changed part of it to a colour like ink. The remainder,
being exposed to the open air, presently became turbid,
threw up a party coloured pellicle, and deposited a yellowish
sediment."

" The water now retained but little power of tinging with
galls; and in a few days lost this property entirely" ("On
the Solubility of Iron in Simple water, by the intervention
of Fixed Air," Phil. Trans., 1769, 59, 218).

These simple experiments, carried out 140 years ago, disclose
the two essential features of the modern theory of the rust-
ing of iron, namely, (i) that iron is dissolved by carbonic
acid to a colourless solution, and (2) that this solution
deposits a yellow rust on exposure to the air.

Black (1755) discovers a second series of salts derived
from fixed air. — Whilst there is some uncertainty as to
whether Black recognised fixed air as an acid, there is no
doubt as to his discovery of the existence of a salt containing a
larger proportion of fixed air than ordinary potash.

" That the fixed alkali, in its ordinary state, is seldom en-
tirely saturated with air, seems to be confirmed by the
following experiment. I exposed a small quantity of a pure

vi CARBONIC ACID AND THE CARBONATES 109

vegetable fixed alkali to the air, in a broad and shallow
vessel, for the space of two months ; after which I found a
number of solid crystals, which resembled a neutral salt so
much as to retain their form .pretty well in the air, and to
produce a considerable degree of cold when dissolved in
water. Their taste was much milder than that of ordinary
salt of tartar ; and yet they seemed to be composed only of
the alkali, and of a larger quantity of air than is usually con-
tained in that salt, and which had been attracted from the
atmosphere : for they still joined very readily with any acid,
but with a more violent effervescence than ordinary ; and
they could not be mixed with the smallest portion of
vinegar. . . . without emitting a sensible quantity of air "
(A.C.R. XVII. 42).

The same salt was prepared by Cavendish and by Berg-
man by the direct action of fixed air on a solution of the
alkali. Cavendish, who examined it in 1766, found that in
comparison with ordinary potash it contained twice as much
fixed air relatively to its power of neutralising acids.
Bergman also examined this salt, and made a rough
analysis, which showed that it contained both water and
fixed air in combination with the mild alkali (JSssays, p. 18).

Such salts, which contain a double portion of fixed air,
were described by Bergman as "aerated vegetable alkali,"
" aerated magnesia," and so forth, but are now known as
BICARBONATES. It is remarkable (as Bergman noticed) that
whilst the excess of fixed air converts the mild alkalis into
less soluble neutral salts, the solubility of chalk and magnesia
in water depends on their conversion into more soluble bi-
carbonates. Similar soluble bicarbonates are probably
formed when iron and zinc are dissolved by carbonic
acid, since the carbonates of these metals are insoluble in
water.

SUMMARY AND SUPPLEMENT.

Priestley (1/72) showed that fixed air is produced when
charcoal is burnt in air confined over water, or over mercury ;

CHAPTER VII

THE BURNING OF INFLAMMABLE AIR, AND THE COMPOSITION
OF WATER

Priestley (1781) notices the formation of water as a
product of combustion. — In ancient times the theory was
held that all matter was derived from four " elements, "-
earth, air, fire, and water. The discovery of the composite
nature of earths (such as the metallic oxides) and of air has
been described in Chapter III ; in the same chapter an
account has been given of the experiments which proved
that fire is not a material substance, and cannot of itself add
to, or subtract from, the weight of things. The composite
nature of water was not suspected1 until a later period,
although it was supposed to give birth to many material
substances, and even to vegetable and animal tissues.2

1 "There are, I believe, very few maxims in philosophy that have
laid firmer hold upon the mind, than that air, meaning atmospherical
air (free from various foreign matters, which were always supposed to
be dissolved, and intermixed with it), is a simple elementary substance,
indestructible and unalterable, at least, as much so as water is supposed
to be" (Priestley, Experiments on Air, II. 30 — 31).

2 Van Helmont (1577 — 1644) planted a willow tree weighing five
pounds in 200 pounds of dry earth, and watering it with rain or
distilled water, found that after five years it weighed 169 Ibs., although
the earth had only lost two ounces. He knew nothing of the air as a
source of the combustible part of a plant, and concluded that 164
pounds of wood, bark, roots, etc., had been produced from the water
alone.

CH. vii THE BURNING OF INFLAMMABLE AIR 113

The observation which led to the discovery of the com-
pound nature of water was made by Priestley, in 1781, in
the course of "a mere random experiment, made to
entertain a few philosophical friends " (Experiments and
Observations, 1781, V. Appendix, p. 398). This consisted
in exploding a mixture of common air and inflammable air by
means of an electric spark in a closed glass vessel. After
the explosion the sides of the vessel were seen to become
covered with dew. Further experiments made by Mr.
Warltire (ibid., pp. 395-397) with a copper vessel appeared
to indicate a slight loss of weight, owing, it was thought,
to the heat which had escaped. This experiment attracted
the attention of Cavendish, who repeated it in 1781, and
published an account of his work in the Philosophical
Transactions of the Royal Society for 1784, under the title
" Experiments on Air" (A.C.R. III).

Cavendish (1781) prepares water by burning inflam-
mable air with common air. — Using a large explosion vessel
capable of holding 24,000 grains of water, Priestley's experi-
ment " was repeated several times with different proportions
of common and inflammable air." The best conditions were
reached when 423 measures of inflammable air were exploded
with 1000 measures of common air, whereby the volume of
the air was reduced to 811 measures, and practically all the
oxygen was removed. But, whatever the proportions in
which the gases were mixed, Cavendish "could never perceive
a loss of weight of more than one-fifth of a grain, and com-
monly none at all." Although the supposed loss of weight
could not be confirmed, Cavendish found himself in agree-
ment with Priestley when he noticed that :

"In all the experiments, the inside of the glass globe
became dewy . . . ; but not the least sooty matter could be
perceived" (A. C. R. III. 13).

" The better to examine the nature of this dew, 500,000
grain measures of inflammable air were burnt with about 2\

I

H4 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

times that quantity of common air, and the burnt air made to
pass through a glass cylinder eight feet long and three-
quarters of an inch in diameter, in order to deposit the dew.
The two airs were conveyed slowly into this cylinder by
separate copper pipes, "passing through a glass plate which
stopped up the end of the cylinder ; and as neither inflam-
mable nor common air can burn by themselves, there was
no danger of the flame spreading into the magazines from
which they were conveyed. Each of these magazines con-
sisted of a large tin vessel, inverted into another vessel just
big enough to receive it. The inner vessel communicated
with the copper pipe, and the air was forced out of it by
pouring water into the outer vessel ; and in order that the
quantity of common air expelled should be 2\ times that of
the inflammable, the water was let into the outer vessels by
two holes in the bottom of the same tin pan, the hole which
conveyed the water into that vessel in which the common
air was confined being 2j times as big as the other."

" In trying the experiment, the magazines being first filled
with their respective airs, the glass cylinder was taken off,
and water let, by the two holes, into the outer vessels, till
the airs began to issue from the ends of the copper pipes ;
they were then set on fire by a candle, and the cylinder put
on again in its place. By this means upwards of 135 grains
of water were condensed in the cylinder, which had no taste
nor smell, and which left no sensible sediment when evapor-
ated to dryness ; neither did it yield any pungent smell
during the evaporation ; in short, it seemed pure water."

" By the experiments with the globe it appeared, that
when inflammable and common air are exploded in a proper
proportion, almost all the inflammable air, and near one-
fifth of the common air, lose their elasticity, and are con-
densed into dew. And by this experiment it appears, that
this dew is plain water, and consequently that almost all the
inflammable air, and about one- fifth of the common air, are
turned into pure water" (A. C. R. III. 14 — 15).

Cavendish prepares water by exploding inflammable
air with oxygen. — Having thus shown that only one-fifth
of the air was condensed by the explosion of ordinary air

vii THE BURNING OF INFLAMMABLE AIR 115

with inflammable air, Cavendish repeated his experiments
with oxygen in place of air.1

" In order to examine the nature of the matter condensed
on firing a mixture of dephlogisticated and inflammable air,
I took a glass globe, holding 8800 grain measures, furnished
with a brass cock and an apparatus for firing air by electricity.
This globe was well exhausted by an air-pump, and then
filled with a mixture of inflammable and dephlogisticated
air, by shutting the cock, fastening a bent glass tube to its
mouth, and letting up the end of it into a glass jar inverted
into water, and containing a mixture of 19500 grain measures
of dephlogisticated air, and 37000 of inflammable ; so that,
upon opening the cock, some of this mixed air rushed
through the bent tube, and filled the globe.2 The cock was
then shut, and the included air fired by electricity, by which
means almost all of it lost its elasticity. The cock was then
again opened, so as to let in more of the same air, to supply
the place of that destroyed by the explosion, which was
again fired, and the operation continued till almost the
whole of the mixture was let into the globe and exploded.,
By this means, though the globe held not more than the
sixth part of the mixture, almost the whole of it was exploded
therein, without any fresh exhaustion of the globe " (A. C. R.
III. 15—16).

The gas remaining in the globe was found to be 2,950
grain measures, of which about 1,000 consisted of unburnt
oxygen. The residue, which was usually not more than one-
fiftieth of the original mixture, Cavendish attributed to impuri-
ties in the gases used, and concluded that if these " could be
obtained perfectly pure the whole would be condensed."

The liquid condensed in the globe weighed thirty grains.

1 Cavendish does not give any illustration of his apparatus ; it was
evidently very similar to that of Monge (Fig. 26), but differed in that
the gases were not drawn from separate reservoirs but from an inverted
jar containing an explosive mixture of the two gases.

2 " In order to prevent any water from getting into this tube, while
dipped under water to let it up into the glass jar, a bit of wax was
stuck upon the end of it, which was rubbed off when raised above the
surface of the water."

I 2

H4 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

times that quantity of common air, and the burnt air made to
pass through a glass cylinder eight feet long and three-
quarters of an inch in diameter, in order to deposit the dew.
The two airs were conveyed slowly into this cylinder by
separate copper pipes, "passing through a glass plate which
stopped up the end of the cylinder ; and as neither inflam-
mable nor common air can burn by themselves, there was
no danger of the flame spreading into the magazines from
which they were conveyed. Each of these magazines con-
sisted of a large tin vessel, inverted into another vessel just
big enough to receive it. The inner vessel communicated
with the copper pipe, and the air was forced out of it by
pouring water into the outer vessel ; and in order that the
quantity of common air expelled should be 2 \ times that of
the inflammable, the water was let into the outer vessels by
two holes in the bottom of the same tin pan, the hole which
conveyed the water into that vessel in which the common
air was confined being 2\ times as big as the other."

" In trying the experiment, the magazines being first filled
with their respective airs, the glass cylinder was taken off,
and water let, by the two holes, into the outer vessels, till
the airs began to issue from the ends of the copper pipes ;
they were then set on fire by a candle, and the cylinder put
on again in its place. By this means upwards of 135 grains
of water were condensed in the cylinder, which had no taste
nor smell, and which left no sensible sediment when evapor-
ated to dryness ; neither did it yield any pungent smell
during the evaporation ; in short, it seemed pure water."

" By the experiments with the globe it appeared, that
when inflammable and common air are exploded in a proper
proportion, almost all the inflammable air, and near one-
fifth of the common air, lose their elasticity, and are con-
densed into dew. And by this experiment it appears, that
this dew is plain water, and consequently that almost all the
inflammable air, and about one- fifth of the common air, are
turned into pure water" (A. C. R. III. 14 — 15).

Cavendish prepares water by exploding inflammable
air with oxygen. — Having thus shown that only one-fifth
of the air was condensed by the explosion of ordinary air

vii THE BURNING OF INFLAMMABLE AIR 115

with inflammable air, Cavendish repeated his experiments
with oxygen in place of air.1

" In order to examine the nature of the matter condensed
on firing a mixture of dephlogisticated and inflammable air,
I took a glass globe, holding 8800 grain measures, furnished
with a brass cock and an apparatus for firing air by electricity.
This globe was well exhausted by an air-pump, and then
filled with a mixture of inflammable and dephlogisticated
air, by shutting the cock, fastening a bent glass tube to its
mouth, and letting up the end of it into a glass jar inverted
into water, and containing a mixture of 19500 grain measures
of dephlogisticated air, and 37000 of inflammable ; so that,
upon opening the cock, some of this mixed air rushed
through the bent tube, and filled the globe.2 The cock was
then shut, and the included air fired by electricity, by which
means almost all of it lost its elasticity. The cock was then
again opened, so as to let in more of the same air, to supply
the place of that destroyed by the explosion, which was
again fired, and the operation continued till almost the
whole of the mixture was let into the globe and exploded.;
By this means, though the globe held not more than the
sixth part of the mixture, almost the whole of it was exploded
therein, without any fresh exhaustion of the globe " (A. C. R.
III. 15—16).

The gas remaining in the globe was found to be 2,950
grain measures, of which about 1,000 consisted of unburnt
oxygen. The residue, which was usually not more than one-
fiftieth of the original mixture, Cavendish attributed to impuri-
ties in the gases used, and concluded that if these " could be
obtained perfectly pure the whole would be condensed."

The liquid condensed in the globe weighed thirty grains.

1 Cavendish does not give any illustration of his apparatus ; it was
evidently very similar to that of Monge (Fig. 26), but differed in that
the gases were not drawn from separate reservoirs but from an inverted
jar containing an explosive mixture of the two gases.

2 " In order to prevent any water from getting into this tube, while
dipped under water to let it up into the glass jar, a bit of wax was
stuck upon the end of it, which was rubbed off when raised above the
surface of the water."

I 2

ii6 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

Cavendish noticed that it "was sensibly acid to the taste,"
an observation which formed the starting-point of his
experiments on the composition of nitric acid, described in
Chapter X. The formation of acid was, however, only
incidental, since when no excess of oxygen was used the
liquid condensed after the explosion was pure water.

Lavoisier (1777 — 1782) studies the burning of inflam-
mable air, — The products resulting from the burning of
inflammable air were being investigated in France about
the same time that Cavendish was carrying out his
experiments in England.

Macquer, in 1776, "having presented a fragment of white
porcelain to the flame of inflammable air burning quietly at
the mouth of a bottle, observed that the flame was not
accompanied by any smoke ; he found the fragment merely
moistened with droplets of a colourless liquid . . . which
he recognised ... as pure water" (Lavoisier's Works,

II- 335)-.

Lavoisier, who had obtained acid products on burning
sulphur, phosphorus, and carbon, carried out in 1777 the
experiment of burning inflammable air in a flask containing
lime-water. He found that no fixed air was produced by
burning the gas prepared from oil of vitriol and iron. Four
years later, in 1781 — 1782, he repeated the experiment with
oxygen in place of air.

"We took a bottle of six pints, which we filled with
inflammable air ; we lighted it quickly, and poured into it
two ounces of lime-water ; we closed the bottle at once with
a wooden stopper, traversed by a pointed copper tube,
connected by a flexible pipe with a reservoir filled with vital
air. Contact between the inflammable air and the atmos-
pheric air having been interrupted by the stopper, the
surface of the inflammable air ceased to burn, but there
formed at the end of the copper tube, in the interior of the
bottle, a beautiful point of brilliant flame, and we saw with
great pleasure the vital air burning in the inflammable air,

vir THE BURNING OF INFLAMMABLE AIR 117

in the same manner and with the same circumstances as
inflammable air burns in vital air " (" Decomposition and
Recomposition of Water," Works, II. 334-359 ; p. 336).

The experiment was repeated twice, with distilled water
and with a diluted alkali in place of the lime-water ; the
water, after the combustion, was as pure as before ; it gave
no signs of acidity, and the alkaline liquor was precisely in
the same condition as before the experiment.

In June, 1783, the experiment was carried out on a larger
scale, inflammable air and vital air from two large reservoirs
being burnt together in a bell jar, inverted over mercury.
Half an ounce of water was collected.

"This water, submitted to all the tests that one could
imagine, appeared as pure as distilled water : it did not
redden tincture of tournesol ; it did not turn syrup of violets
green ; it did not precipitate lime-water ; finally, by all the
known tests, one could not discover the least sign of
any impurity " (ibid. p. 338).

Lavoisier was able to interpret his observations very
clearly and simply by means of his theory that burning
meant combination with oxygen and not loss of phlogiston.
In order to express his view that water was simply a compound
of inflammable air with oxygen, in the formation of which
"phlogiston played no part whatever," Lavoisier and his
colleagues suggested in 1787 the name HYDROGEN, or " water-
producer," for the inflammable gas. They say :

"It is the only substance which produces water by its
combustion with oxygen gas . . ., and we have therefore
called it hydrogen, from v'Swp, water, and yeiVo/xai, / beget •
experiments having proved that water is nothing but oxy-
genated hydrogen, or the immediate production of the
combustion of oxygen gas with hydrogen gas, deprived of
the light and caloric which disengage during the combustion"
(Chemical Nomenclature, tr. 1788, p. 24).

u8 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

Lavoisier (1784) decomposes water by means of iron. —

Having shown that water must be regarded as a compound
of inflammable air with oxygen, Lavoisier argued that it
should be possible to obtain inflammable air by the removal
of oxygen from water. This he succeeded in doing by
passing steam through an iron gun-barrel (Fig. 25) heated
in a charcoal furnace (Meusnier and Lavoisier, " Decom-
position of Water," 1784; Works, II. 360 — 373).

Lavoisier writes :

" If, when the gun-barrel is red and incandescent, water
is allowed to flow drop by drop in very small quantities, it
is decomposed completely, and none of it escapes from the
lower opening of the gun ; the oxygen of the water combines
with the iron and calcines it ; at the same time the
inflammable constituent of the water, set free, passes into
the gaseous state, with a specific weight about 2/25 of that
of common air. At the beginning of the experiment, the
production of inflammable air is very rapid ; it soon
becomes slower, and reaches a uniform condition, which lasts
several hours ; finally, at the end of eight to ten hours, more
or less, according to the thickness of the barrel, the passage
of inflammable air becomes slower, and the water ends by
escaping entirely from the gun, as it entered it, without
decomposing. If this operation has been pushed to the
end, all the substance of the iron which composed the
gun-barrel is converted into a black, brilliant substance,
crystallised in facets like specular iron ore . . . : this
substance occupies a much greater volume than the iron
from which it was produced ; the gun-barrel is consequently
increased in thickness, and its internal diameter considerably
diminished."

" The phenomena are very different if a metal is used for
which oxygen has less affinity than for inflammable air : if,
for example, one substitutes, in the preceding experiment, a
copper barrel for one of iron, the water is reduced to vapour
by passing through the incandescent part of the tube, but it
condenses by cooling in the worm ; the process is then only
an ample distillation without loss, and there is neither

VII

THE BURNING OF INFLAMMABLE AIR

119

120 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

calcination of the copper, nor production of inflammable
m» (Works, II. 352-353)-

Early quantitative experiments on the composition of
water. — Almost simultaneously with the work of Cavendish
and Lavoisier, the formation of pure water as the sole pro-
duct of burning hydrogen and oxygen was noticed by

FIG. 26 — MONGE'S APPARATUS FOR EXPLODING
HYDROGEN AND OXYGEN IN AN EXHAUSTED GLOBE.

The gases were admitted through syphon-tubes, pr,
PR, into the graduated measuring-cylinders, GH ; they
passed through the taps, IK, to the globe, M, which was
exhausted through a tap, L, leading to an air-pump at O.

Monge, who carried out a series of careful quantitative
experiments in the year 1781 (Mem. Acad. Sci , Paris, 1783,
78 — 88). Using the apparatus shown in Fig. 26, he
exploded measured volumes of hydrogen and oxygen in
an exhausted glass globe, and thus collected a large

vii THE BURNING OF INFLAMMABLE AIR 121

quantity of the liquid product. The method of procedure
was to admit T*Y volume of oxygen, fill up with hydrogen,
and explode; by admitting fresh quantities of oxygen to
the excess of hydrogen six further explosions could be
made. A fresh supply of hydrogen was then admitted
and the process repeated until 137 explosions had taken
place; the globe was then found to be filled with a
residue of non-explosive gas, which was drawn out by means
of a pump, collected and measured. At the end of three
series of experiments, in which 372 explosions were made,
the quantities of gas used were :

Hydrogen ... ... 145 TVi pints.

Oxygen *.-.?' ./ 74 A

The weight of gas, calculated from the measured densities,
was :

Hydrogen ... ... ... ... 442*03 grains.

Oxygen 1786*53 „

2228-56 ,,
Or corrected for changes of pressure 2187-56 ,,

The products were :

Liquid ... ' ... 1917-10 grains.

Unburnt gas (7 pints) ... ... 171*91 »

2089*01 „
Deficiency ... ... ... 98*55 grains.

" This liquid, perfectly transparent, reddened impercep-
tibly paper tinted blue by tournesol, much less than that
obtained in a previous experiment, less even than saliva.
This acidity can not be attributed to fixed air, because the
liquid did not precipitate lime-water, and because distilled
water, equally acidulated by fixed air, at once rendered lime-
water milky ; it scarcely whitened a solution of silver in
nitric acid, and a little more sensibly that of mercury in
the same acid. In addition to its slight acidity, it has also
the empyreumatic taste that water always acquires in distilla-

122 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP-

tion ; this product must then be regarded as pure water
charged with the small quantity of vitriolic acid which in-
flammable air carries when prepared by the dissolution of
iron."

" It follows from this experiment, that when inflammable
air and dephlogisticated air, both pure, are exploded, there
is no other product but pure water, heat, and light " (loc. cit.,
pp. 86 and 87).

Monge's experiments agreed with those of Cavendish and
with the later experiments of Humboldt and Gay Lussac
(Journ. de Physique, 1805, 60, 129 — 168), in showing that the
volume of hydrogen used in the explosion is about twice as
great as that of the oxygen. But, owing to the use of moist
gas, the density of the hydrogen and its proportion by weight
were greatly over-estimated. It is, however, noteworthy that
the method used by Monge in 1781 was essentially the same
as that employed 115 years later in the classical experiments
of Scott and of Morley, who determined the composition of
water with very great accuracy by measuring the densities
of the two gases and the ratio of their combining volumes.

Volta's eudiometer. — The fact that oxygen and hydrogen
(unlike nitrous air and oxygen, Chapter X, p. 201) combine
together in a definite ratio, was used by the Italian physicist
Alexandro Volta (1745 — 1827) (Annali di Chimtca, 1790,
I. 171 — 232; 1791, II. 261 — 286; 1791, III. 36 — 45) to
estimate these gases. In order to explode the gases together,
Volta introduced a metal cap, b, carrying an insulated wire,
c, into the upper part of the graduated tube which constituted
Priestley's " eudiometer." The instrument thus modified
was known as VOLTA'S EUDIOMETER (Fig. 27). In order to
estimate oxygen, for instance in atmospheric air, an excess
of hydrogen was added and the mixture exploded ; as
two volumes of hydrogen combined with one of oxygen to
form water, one-third of the decrease of volume was due to
oxygen, the volume of which could thus be estimated very

VII

THE BURNING OF INFLAMMABLE AIR

123

accurately. This method of analysis had two advantages (i)
the hydrogen need not be pure, and (2) the volume of
hydrogen added need not be adjusted accurately.

In the hands of Gay- Lussac, Berthollet, Cruikshank, and
others, Volta's eudiometer be-
came an important instrument,
not only for the estimation of
oxygen and hydrogen, but for
the examination and analysis
of other combustible gases.

Water is produced by the
action of inflammable air on
metallic calces. — The revivifi-
cation of metallic calces by in-
flammable air was discovered in
1785 by Priestley (Experiments
and Observations, 1786, VI. 5).

By means of a burning
glass, he heated red lead in a
bell -jar inverted over water
and filled with the inflammable
gas. As soon as the red lead
became dry the inflammable
air was absorbed l and metallic
lead was produced ; no other
product was seen and Priestley
concluded that :

red lead '+ inflammable air
(phlogiston) = lead,

the inflammable air being re-
garded by him as pure phlo-
giston.

1 101 volumes of the gas were
reduced to 2, which still consisted
of inflammable air (loc. cit. p. 7).

FIG. 27 — VOLTA'S EUDIOMETER.

The funnel at the base was used
to collect the gas over water con-
tained in the trough C. The ring
AD was used to read off the posi-
tion of the meniscus, e, on the
graduated scales which enveloped
the greater part of the tube. Details
of the sparking-device are shown at
abc.

124. HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

Lavoisier called attention to the fact that the calx was
diminished in weight by its combination with inflammable
air, and concluded that some other product, namely water,
must be formed in addition to lead, thus :

red lead '+ hydrogen = lead+ water.

Water was also produced in small quantities when red lead
was revivified by means of charcoal, showing that the
charcoal contained hydrogen as well as carbon (Lavoisier's
Works, II. 344—348).

Oxidation and reduction. — The oxides of copper and iron
can be reduced like red lead by heating them in a current
of hydrogen, but the refractory oxide of zinc is not acted
on by the gas. The case of iron is of interest because
red-hot iron is oxidised by heating in a current of steam
(p. 1 1 8) but the oxide is reduced when a current of hydrogen
is passed over it. The action is therefore reversible, pro-
ceeding in one direction or the other, as shown by the
two arrows, according as hydrogen or steam is present in
excess.

iron + steam^iron oxide -f hydrogen.

It should be noticed that the term OXIDATION is often
used to describe the removal of hydrogen as well as the
addition of oxygen, whilst the converse process of REDUCTION
may result either in the removal of oxygen or the addition
of hydrogen.

Dumas's experiments on the composition of water,
(1842). — The reduction of copper oxide by hydrogen was
used by Berzelius and Dulong in 1820 (Ann. de Chimie, 15,
386-395) to determine the composition of water. Instead
of weighing the two gases of which it is composed, they
weighed the water and calculated the weight of the oxygen
from the loss in weight of the copper oxide ; the difference
between these weights gave the weight of the hydrogen.

vii THE BURNING OF INFLAMMABLE AIR .125

They found the ratio :

oxygen : hydrogen — 88 '9 : 1 1 ' i
= 8-01: 1

This experiment was repeated on a very large scale by
the French chemist Jean Baptiste Dumas (1800-1884) in
1842 ( " Researches on the Composition of Water,"
Ann. de Chimie, 1843, 8,189-207). In Dumas' experiments
(Fig. 28), the hydrogen, prepared by the action of zinc on dilute
oil of vitriol, was purified and dried by passing through seven
U -tubes, each a metre in height ; the purifying agents were
(i) lead nitrate solution to remove sulphuretted hydrogen, (2)
silver sulphate solution to remove arseniuretted hydrogen,
(3) potash (three tubes) to remove acid vapours. The dry-
ing agent was (4) sulphuric acid cooled in ice, or phosphoric
anhydride (two tubes). The purifying agents were distributed
over broken glass or pumice in order to secure proper con-
tact with the gas. After the purification no odour could be
detected even when a hundred litres of the gas were allowed
to escape into the air.

The copper oxide was contained in a large bulb of hard
glass provided with a beak a metre long to condense the
water formed on reduction. The bulb was weighed, after
removing the air by means of a pump. Before heating, the
air in the apparatus was displaced by hydrogen, but a cor-
rection was needed for the fact that the sulphuric acid used
in making the gas contained a little oxygen, which con-
tributed something to the weight of water produced.1
During the reduction, which occupied 10 to 12 hours,
the bulb was heated by a large spirit lamp. After
allowing the copper to cool, the hydrogen in the apparatus
was displaced by air; the bulb was again exhausted by
means of a pump and the loss of weight was determined by
a second weighing of the bulb.

1 In later experiments this was got rid of by passing the gas over hot
copper before drying it.

126 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

an-a

•C <o

^f||8|l|

I iSifli

o i^trii'siLg

c3 ^i?l*E[S'^

, g 3 O.H = s M^
« -5 « 8 g-s &|^

i 1 l^l'Ssi

s b

fa 9-

ri.so^

vii THE BURNING OF INFLAMMABLE AIR 127

The water was collected in a smaller bulb, followed by a
series of drying tubes, similar to those used to dry the
hydrogen. The water could not be weighed in a vacuum,
but the observed weight was corrected for the buoyancy of
the air which it displaced. In all, more than a kilogram of
water was collected.

In a typical experiment, using phosphoric anhydride as
the drying agent, the figures were :

Exhausted bulb and copper oxide ... ... 673*280 grams.

,, ,, copper 613-492 ,,

Oxygen = 59788 „

Bulb and drying tubes 93i'487 ,,

,, ,, ,, with water 998700 ,,

67-213
Air correction ... ... 0*069

Water = 67-282

Hydrogen combined with I part of oxygen =0*12533 ,,
Ditto, corrected for air in the sulphuric acid = 0*12508 ,,

From nineteen experiments Dumas concluded that the
ratio of hydrogen to oxygen lay between 0*1250 and 0*1256.
This gives the ratio of oxygen to hydrogen as 8*00 to 7*96,
or taking a mean value,

oxygen : hydrogen = 7 '98 : 1.

Scott's determination of the composition of water. —

The figures obtained from the laborious experiments of
Dumas remained unchallenged for almost half a century.
In 1893, however, Alexander Scott (Phil. Trans., 1893, 184,
543-568) obtained an independent value for the composition
of water by reverting to the method of finding the composi-
tion by weight from the densities and combining volumes of
the gases. The densities had been determined with very
great care by Lord Rayleigh. Scott measured the com-
bining volumes by exploding the moist gases after carefully

128 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

comparing their volumes at atmospheric temperature and
pressure.

The hydrogen was prepared by the action of metallic
sodium on steam, with or without absorption by palladium
(see footnote, p. 130). The oxygen was prepared by heating
silver oxide. Both gases were of the highest degree of purity,
the residue of impurity left after explosion being usually less
than one part in 100,000. The ratio of the combining
volumes was found to be 2'00245 at 14° to 18° C. Lord
Rayleigh's determination of the densities of the gases had
given the ratio 15 '88 2 : i, the combining weights of the
two gases were therefore in the ratio,

oxygen : hydrogen = 7*931 : 1.

Morley's experiments on the composition of water. —

The redetermination of the composition of water by
E. W. Morley in the year 1895 (Smithsonian Contributions
to Knowledge, tr. Zeit. physikal. Chem.> 1895, 20, 68, 242,
417) was carried out with extraordinary care and accuracy,
and has already become one of the classics of scientific
literature.

In these experiments the use of copper oxide was again
discarded ; the hydrogen and oxygen were weighed directly,
as in the earliest experiments, but with all the added
resources which had accrued from the experimental work
of the intervening century.

The density of gxygen (prepared by heating potassium
chlorate or by electrolysing dilute sulphuric acid) was
determined by weighing it directly in globes of 20 litres
capacity, suspended in pairs x in a large desiccator-cupboard
from the pans of a balance (Fig. 29). In a typical experiment
the data were :

1 The method of weighing a glass globe against a counterpoise of
similar form was introduced by Regnault : it serves to eliminate most of
the errors caused by variations of barometric pressure and of the amount
of moisture in the atmosphere.

VII

THE BURNING OF INFLAMMABLE AIR

129

Volume... .. 2 1 568 '4 c.c. Weight ... 28*4308 grams.

Temperature ... 17-48° C. Corrected ... I7'4O°C.

Barometer .. 745 '91 nim. Corrected ... 746*29 mm.

Weight of I litre of oxygen at sea level in 45°

latitude, at o° C. and 760 mm. pressure ... 1*42883 grams.

The mean of three complete series of experiments, in which
different methods were used for measuring temperature and
pressure, gave the value 1 42900 ±0 '000034.

FIG. 29— MORLEY'S APPARATUS FOR MEASURING THE DENSITY
OF OXYGEN.

The density of hydrogen was determined in a different
way, both on account of the difficulty of weighing 1*8 grams
of gas in a globe 600 times heavier, and also because of the
risk that this small weight might be increased appreciably

by the presence of mercury vapour. In the final series of
experiments, therefore, the gas was weighed in a tube filled
with metallic palladium (Fig. 30) ; l on heating the tube the
gas was expelled into three large globes immersed in ice ; the
pressure produced in these globes by a known weight of gas
was measured. In a typical experiment the data were :

Volume of three globes... ... ... ... 43*2574 litres.

,, ,, air space in manometer ... ... 0-0536 ,,

,, ,, leads 0*0365 ,,

Temperature, o° C. Total volume ...-43-3475 ,,

Pressure (corr.), 725-30 mm. Weight = 3 7 158 grams.
Density of hydrogen 0-089864 gram per litre.

Twenty-five experiments made in this way gave an average
value

0-089873 ±0-0000027.

In order to find the weights of the gases which combined
together to form water, it was necessary next to determine

FIG. 30— MORLEY'S PALLADIUM-TUBE FOR WEIGHING HYDROGEN.

a Is a bulb containing 600 grams of palladium ; b contains phosphoric
anhydride to dry the gas ; c is a seal of fusible metal.

the ratio of their combining volumes. For this purpose
Morley prepared a mixture of the two gases by passing an
electric current through dilute potash cooled in ice. The
apparatus was provided with drying tubes and a tap, and
was arranged so that it could be weighed 'before and after

1 Morley used 600 grams of palladium ; he found that this quantity
absorbed 3*8 grams of hydrogen and liberated 37 grams, occupying
over 40 litres. This method of purifying and weighing hydrogen was
introduced by Chirikoff in 1882 ; it forms one of the most important
of the advances that have been made in the methods of determining
the composition of water since the work of Dumas.

THE BURNING OF INFLAMMABLE AIR

delivering the gas. The density of the gas was determined

in the same way as in the case of hydrogen, by finding the

pressure produced in the three large

globes by a known weight of the gas.

In a typical experiment 23*0269

grams of the gas produced a pressure

°f 753'97 mm- m a volume of

43-3628 litres; density 0-535513;

mean of 10 experiments

0*5355 10 ±0*000010. From this

density the proportion of hydrogen

to oxygen was calculated, after

making various corrections, to be

2*00357. By exploding the gas in

a eudiometer it was found that

there was a slight excess of hydrogen,

amounting to 0-000293 °f trie total

volume or 0*00088 of the volume of

the oxygen. The combining volumes

were therefore in the proportion

hydrogen : oxygen = 2 '00269 : 1.

A second method of determining
the composition of water was by
weighing the water produced by the
combination of known weights of
hydrogen and oxygen. The hydro-
gen (21 litres) was drawn from a
palladium bulb a (Fig. 30) provided
with a drying tube of phosphoric
anhydride b, and a tiny plug of
fusible metal c in place of a tap. The
oxygen (42 litres) was drawn from two large globes which
were weighed in the same way as in the density experiments.

The combustion was carried out in a cylindrical glass
tube (Fig. 31) surrounded with cold water (Fig. 32) and

K 2

FIG. 31— MOKLEY'S COM-
BUSTION-TUBE.

The gases were admitted
by jets aa after passing
through drying tubes bb ;
they were ignited by
sparking at ff.

132 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

provided with two jets a a and a pair of wires //for a spark to
ignite the gas. The tube was exhausted before and after the
combustion ; the water produced was frozen before exhaust-
ing, and drying tubes b b were provided on either side in order
that none of the water might be lost. The gas remaining

FIG. 32— MORLEY'S APPARATUS FOR THE COMBUSTION OF HYDROGEN
AND OXYGEN.

The leads for the hydrogen (on the left) and for the oxygen (on the right), with
the regulating taps are hidden behind the wooden frame. _ They are joined to
the combustion-tube by long, flexible, glass leads, terminating in drying tubes.
The upper part of the combustion-chamber is surrounded by water, the lower
part by a freezing-mixture.

in the chamber at the end of the experiment was collected
and analysed. In a typical experiment the weights were :

Hydrogen
Oxygen ..

3*8223 -O'ooi2= 3 '82 1 1 grams.
30'3775-°'°346 =

Total

= 34-1640

Water

Whence oxygen to hydrogen
•water to hydrogen

= 34-1559

= 7W
= 8'939

vii THE BURNING OF INFLAMMABLE AIR 133

The mean of twelve experiments in which 400 grams of
water were produced gave the ratios

oxygen to hydrogen = 7 '%%§§
water to hydrogen = 8*9392

From the ratio 15*9002 of the densities and 2*00269 °f
the conjoining volumes, the two constituents are in the
ratio

oxygen : hydrogen = 7*9395 : 1.

The high order of accuracy reached in these modern
measurements on the composition of water is shown by the
following comparison of the values for o° C. and 760 mm.
pressure.

Ratio of densities Rayleigh 15*882

,, (reduced to o°C. ) 15*900
Morley ... ... 15*9002

Ratio of volumes ... ... Scott ... ... ... 2*00245

,, (reduced to o°C.) 2*00285

Morley 2*00269

Ratio of weights Rayleigh & Scott ... 7*931

Morley (a) 7 -9396

(from oxygen : hydrogen)

Morley (b) 7 -9392

(from water : hydrogen)
Morley (c) •••_••• 7*9395
(from combining volumes)

SUMMARY AND SUPPLEMENT.
A. Water an Oxide.

Cavendish, in 1781 (" Experiments^ Air," Part I, published
in 1784), showed that pure water, without taste or smell, was
produced by exploding or burning a mixture of inflammable
air with 2| times its volume of common air. Water was also
formed when that gas was exploded with \ volume of oxygen,
but in this case the product was slightly acid.

Lavoisier, in 1783, prepared half an ounce of pure water by
burning oxygen and inflammable air. He also proved the
presence of oxygen in water by passing steam through a red-
hot iron gun-barrel, when inflammable air and magnetic oxide
of iron were produced. These experiments showed that water

134 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

was not an element, as had been supposed previously, but a
compound of inflammable air with oxygen. Lavoisier and
his colleagues in 1787 therefore gave to the inflammable air
the name "hydrogen'' or water-producer (compare German
ivasserstoff}. Lavoisier's "Decomposition and Recomposition
of Water" may be represented by the equations :

3Fe + 4H2O-> Fe3O4 + 4H2
Iron + steam -> oxide of iron + hydrogen
2H2 + O2 -> 2H2O
Hydrogen + oxygen -> water-vapour.

Priestley, in 1785, showed that many metallic oxides could
be reduced to metals by heating them in hydrogen ; Lavoisier
showed that water is also produced. The action may be
represented by equations such as the following :

PbO + H2 -> Pb 4- OH2
Litharge + hydrogen — > lead + water-vapour
CuO + H2 -> Cu + OH2
Copper oxide 4- hydrogen -> copper 4- water- vapour.

Iron oxide is also reduced by hydrogen ; this action is
therefore reversible :

the action proceeding in one direction or the other according
as hydrogen or steam predominates in the gas.

B. The Composition of Water.

Cavendish, in 1781, showed that water was formed by the
combination of two volumes of hydrogen with one volume of
oxygen. This proportion was confirmed by Lavoisier, Monge,
and others, who burnt or exploded together large volumes of
the two gases, as well as by the careful experiments on the
explosion of hydrogen and oxygen which Gay-Lussac and
Humboldt carried out in 1805 with the help of Volta's
" eudiometer." Monge used the measured densities of the two
gases to calculate the composition of water by weight, but
obtained very erroneous results as the hydrogen was not dried.

Berzelius and Dulong, in 1820, made the first successful
determination of the composition of water by using copper

vii THE BURNING OF INFLAMMABLE AIR 135

oxide to oxidise the hydrogen and measuring the weight of
water produced and the loss in weight of the copper oxide

They found the ratio

Oxygen : hydrogen = W\ : 1.

This method was used on a large scale by Dumas in 1842. He

purified the hydrogen

from sulphuretted hydrogen by means of lead nitrate solution ;

from arseniuretted hydrogen by means of silver sulphate solution ;

from acid vapours by means of potash ;

from water-vapour by means of oil of vitriol cooled in ice, or

by means of phosphoric oxide ;

these purifying agents were contained in seven U-tubes, each
i metre in height. The copper oxide and copper were
weighed in a large vacuous bulb provided with a beak one
metre in length ; during the reduction the bulb was heated by
a large spirit lamp. In 19 experiments Dumas prepared one
kilogram of water and found the ratio

Oxygen : hydrogen = 7 '98 : 1.

Scott, in 1893, reverted to the method used by Monge. He
exploded pure hydrogen and oxygen and found the combining
volumes to be in the ratio 2*00245 : 1. When combined with
densities measured by Rayleigh, this gave the composition

Oxygen : hydrogen = 7 '931 : 1.

Morley, in 1895 (i), by weighing the gas in large globes
containing 22 litres found the density of oxygen to be

1*42900 grams per litre at o° C. and 760 mm. pressure.

(2) By measuring the pressure produced in three globes of
43 litres capacity when a known weight of hydrogen was
expelled from palladium, he found the density of hydrogen to be

0*089873 gram per litre at o° C. and 760 mm. pressure.

(3) By exploding a mixture of hydrogen and oxygen he found
the combining volumes to be in the ratio 2*00269 : 1 at o° C.
Whence the composition by weight is found to be

Oxygen : hydrogen = 1'^^ : 1.

(4) By burning together in a special combustion - chamber
known weights of hydrogen and oxygen (weighed as in the

136 HISTORICAL INTRODUCTION TO CHEMISTRY CH. vn

density measurements), Morley found the composition of water
by weight to be

Oxygen : hydrogen = 7 '9396 : 1 ;

whilst the weights of the oxygen and water gave the ratio

Oxygen : hydrogen = 1'%3Q2 : 1 ;
mean,

Oxygen : hydrogen = 7 '9394 : 1.

CHAPTER VIII

THE BURNING OF INFLAMMABLE GASES, LIQUIDS, AND
SOLIDS

A. AN INFLAMMABLE OXIDE OF CARBON.

Lassone (1776) prepares an inflammable gas by
reducing the calx of zinc with charcoal. — The loss of
weight, which takes place when a calx such as litharge is
reduced to a metal by heating it with charcoal, was proved
by Lavoisier to be due to the escape of a very large volume
of fixed air (carbonic anhydride). A different, and some-
what perplexing, result was observed when the white calx of
zinc was reduced in this way. This experiment, which was
carried out by Lassone in 1776, gave rise to an inflammable
gas, but one that differed in a very marked way from the
inflammable air (or hydrogen) which he obtained by the
action of alkalis, as well as of acids, upon the metal.

" From a mixture of a half-ounce [288 grains] of zinc
calx and a gros [72 grains] of charcoal powder put into a
pistol barrel, I extracted in the fire of a forge 96 cubic
inches of a gas, which burns rapidly without detonating ;
the flame is blue : this kind of air at first permanent
mixes afterwards little by little with water, it does not
redden nitrous air."

" Two gros [144 grains] of Prussian blue submitted in a
pistol barrel to the action of a forge fire, gave more than

1.37

138 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

34 cubic inches of air which burns without detonating,
producing a very beautiful blue flame."

" These two last experiments give an inflammable air of
quite a peculiar character ; for it does not make the
smallest explosion, nor the least noise when inflamed,
after mixing with common air or with dephlogisticated air ;l
from which it follows that there appear to. exist up to the
present two distinct kinds of inflammable gas ; one which
inflames rapidly with explosion and great noise when mixed
with common ^atmospheric air, and which detonates still
more strongly *when mixed with dephlogisticated, or pure
air ; and the second, which although mixed with the two
preceding gases, always burns quietly without noise."
(Mem.Acad. Sci., Paris, 1776, 90, 686-696 ; p. 691.)

Priestley (1785) prepares an inflammable gas from
smithy scale. — A few years later, in 1785, Priestley noticed
the production of a similar inflammable gas during the
reduction of SMITHY SCALE, or FINERY CINDER, the bluish-
black calx formed when iron is burnt in air.

" Having made the scales of iron, and also the powder of
charcoal very hot, previous to the experiment, so that I was
satisfied that no air could be extracted from either of them
separately by any degree of heat, and having mixed them
together while they were hot, I put them into an earthen
retort, glazed within and without, which was quite impervious
to air. This I placed in a furnace, in which I could give it
a very strong heat ; and connected with its proper vessels to
condense and collect the water which I expected to receive
in the course of the process. But, to my great surprise, not
one particle of moisture came over, but a prodigious quantity
of air, and the rapidity of its production astonished me ; so
that I had no doubt but that the weight of the air would
have been equal to the loss of weight both in the scales and
in the charcoal; and when I examined the air, which I
repeatedly did, I found it to contain one-tenth of fixed air,
and the inflammable air, which remained when the fixed air

1 i.e., oxygen.

vin THE BURNING OF INFLAMMABLE GASES 139

was separated from it, was of a very remarkable kind, being
quite as heavy as common air" (Experiments and Observa-
tions, 1786, VI. 109).

This observation attracted a great deal of attention.
According to Lavoisier's theory the reduction of an oxide
with charcoal should give rise to the metal and fixed air ; in
this experiment hydrogen appeared to be produced instead.
The only explanation that the French chemists were able to
give of the production of the inflammable gas was that the
charcoal, even when strongly heated, retained a certain
amount of hydrogen, which it released only when burnt
with the help of a suitable oxide.

Cruikshank (1801) recognises an inflammable oxide
of carbon. — The real nature of this inflammable gas was
discovered in 1801 by W. Cruikshank of Woolwich
(Nicholson1 s Journal, April 1801, 5, 1-9), who proved it
to be an oxide of carbon. This was generally regarded as
an impossible view, on account of the low density of the
gas. The French chemist, C. L. Berthollet (1748—1822),
for instance, in 1809 (Mem. Soc. Arcueil, 1809, II., 68 — 93,
p. 85), pointed out that this oxide would be " the only
example that one can quote, in which the gaseous compound
would be lighter than the lightest of its component
elements " and argued that the gas must contain hydrogen.

Cruikshank examined the gases formed by reducing the
black oxide of iron and the white oxide of zinc, and then
proceeded to study the reduction of the red oxide of copper,
of litharge, and of the black oxide of manganese described
by Scheele (p. 210). He found :

(1) "That all metallic oxides capable of enduring a red
heat, will, when mixed with charcoal, not only yield carbonic
acid, but also a very considerable quantity of an inflammable
gas.

(2) " That those oxides which retain their oxygen most
obstinately, yield the greatest quantity of inflammable gas ;

140 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

and on the contrary, that those which part with it readily,
afford the greatest proportion of carbonic acid.

(3) "That the carbonic acid gas was chiefly disengaged
at the beginning of the process, and the purest and greatest
quantity of inflammable gas towards the conclusion of it "
(loc. tit., p. 3).

The gas, after washing with lime water to remove the
carbonic anhydride, had a density 22/23 of that of a^r-
When exploded with oxygen in a Volta eudiometer it was
converted into fixed air. But it required for its combustion
less oxygen than was present in the fixed air produced
from it. The gas therefore contained carbon already
combined with oxygen, the proportion of carbon being
greater and the proportion of oxygen less than in fixed air.
Cruikshank described the gas as the GASEOUS OXIDE OF
CARBON, but it soon became known as CARBONIC OXIDE.

Cruikshank prepares carbonic oxide by reducing
carbonates with iron, — Cruikshank showed that many
gallons of the gas could be prepared by heating carbonated
baryta or common chalk with iron plates. In this experi-
ment fixed air combined with baryta or with lime was
reduced to carbonic oxide by the iron, which removed from
it a part of the oxygen which it contained. This method
of preparation had the advantage of giving a gas which was
free from hydrogen ; the gas prepared from charcoal always
contained hydrogen, thus affording some justification for
Berthollet's view that the gas was a hydrogen-compound.

Desormes and Clement (1801) investigate the inflam-
mable oxide of carbon,— The composition of this gas was
also investigated in France, by Desormes and Clement.
They found, in 1801, that the inflammable gas prepared by
heating zinc oxide with charcoal gave no deposit of dew
when burnt under a cold bell-jar, and that no drops of water
were formed when it was exploded with oxygen over
oil. They measured the volume of oxygen used and

viii THE BURNING OF INFLAMMABLE GASES 141

the volume of carbonic anhydride produced in the
explosion, and concluded that, whilst carbonic anhydride
contained 28% of carbon, the inflammable oxide contained
53% of carbon and 47% of oxygen (" On the Reduction of the
White Oxide of Zinc by Carbon, and on the Gaseous Oxide of
Carbon which is produced." Ann.de Chimie, 1801, 39, 26-64.)

Desormes and Clement prepare carbonic oxide by the
reduction of carbonates. — Having established the nature of
the gaseous oxide, De'sormes and Clement proved that the
same gas is formed during the reduction with charcoal of
refractory carbonates, such as that of baryta. This carbon-
ate resembles chalk in that it effervesces when acids are
added to it ; but it differs from chalk in that it is not readily
decomposed by heat. Whereas carbonated magnesia can
be decomposed in glass vessels, and chalk in an earthenware
retort (Ch. IV., p. 49), carbonated baryta must be decom-
posed by heating with charcoal-, the fixed air is then
reduced and escapes in the form of carbonic oxide. It
should be noted that Cruikshank prepared carbonic oxide
from the same carbonate by the action of iron (p. 140).

Desormes and Clement describe their experiments as
follows :

" The preceding experiments show that the gas generated
by the reduction of oxide of zinc, was only oxygen and
carbon which, meeting at a high temperature, combined in
different proportions from those of carbonic acid. It
seemed possible to form this combination by uniting carbon
and carbonic acid.

" One recalled the experiment of Pelletier who, by the
addition of charcoal to carbonate of baryta, succeeded in
driving out the carbonic acid from it by simple heat, and
one believed that in this experiment the carbonic acid
combining with the carbon, had formed a gas similar to
that obtained in the reduction of oxide of zinc, and was
separated easily from the baryta on account of its greater
elasticity.

142 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

" The experiments were made and the results agreed per-
fectly with our view.

"To 16 grams of carbonate of baryta made directly, 5
grams of charcoal were added and the mixture was heated
in an apparatus similar to that used for the reduction of the
oxide of zinc.

" After three hours' heating, a hole was formed in the
retort, and the heating was stopped : there had been collected
three litres of a gas of which the first portions contained 1/6
of carbonic acid and 5/6 of an inflammable gas similar in
every respect to carbonic oxide. The last portions contained
hardly any carbonic acid.

"In the residue were found 1*64 gram of pure baryta."
(loc. tit., p. 45.)

Desormes and Clement prepare carbonic oxide by re-
ducing carbonic anhydride with charcoal. — Desormes
and Clement made many attempts to prepare carbonic
oxide by the direct combination of carbon with air or
oxygen. They found that carbonic anhydride was in-
variably formed as the first product of combustion, but
that this gas could be reduced to carbonic oxide by passing
it through red-hot iron tubes packed with charcoal (loc. tit.,
p. 46). The gas was passed to and fro until no further
increase of volume was caused by the reduction. At the
present time large quantities of inflammable gas are manu-
factured in this way, by passing air through hot coal or coke ;
the oxygen first forms carbonic anhydride, which is then con-
verted into carbonic oxide by the excess of hot fuel ; the
product, known as PRODUCER GAS, consists mainly of
carbonic oxide and carbonic anhydride, diluted with large
volumes of nitrogen.

Desormes and Clement decompose water by means of
charcoal. — It was well known that large quantities of
inflammable gas could be produced by heating moistened
charcoal. The nature of this process was elucidated by
De'sormes and Clement, who showed that the charcoal

vin THE BURNING OF INFLAMMABLE GASES 143

reduces the water, liberating hydrogen from it, and is itself
converted into carbonic oxide (loc. tit., p. 52). This gas is
now prepared by blowing steam into fuel that has been
made hot by an air-blast, and is known as WATER GAS ; it
consists mainly of hydrogen, carbonic oxide, and carbonic
anhydride. As the steam cools the fuel, water-gas can
only be prepared alternately with producer gas ; a mixture
prepared by blowing in steam and air together is known as

SEMI-WATER GAS.

B. QUANTITATIVE EXPERIMENTS ON THE COMPOSITION
OF THE OXIDES OF CARBON.

Composition by volume. — It was noticed by Priestley and
by Lavoisier that charcoal burnt in air or oxygen does not
cause any change of volume, until the fixed air produced is
absorbed by water, by lime-water, or by an alkali. Oxygen
thus suffers no change of volume when it combines with
carbon to form carbonic anhydride ; at the same time
carbonic anhydride is shown to contain its own volume of
oxygen, thus :

carbon + oxygen — >• carbonic anhydride
i vol. i vol.

The composition of carbonic oxide .is less easily deter-
mined, because it is necessary to use a pure gas for the
analysis. Desormes and Clement found that 100 parts of
the gas exploded with about 40 parts of oxygen, and gave
about 80 parts of carbonic anhydride. Cruikshank
obtained similar numbers for gas prepared by means of
charcoal; but gas prepared by means of iron gave 19/20 of
its volume of carbonic anhydride when exploded with
oxygen. Berthollet (quoted by Gay-Lussac, A.C.R. IV.
15) found that 100 parts of the gas combined with 50 parts
of oxygen to give 100 parts of carbonic anhydride. From
Berthollet's values it follows that carbonic oxide takes up

144 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

one half its volume of oxygen in forming carbonic anhydride,
but does not suffer any change of volume in the process,
thus:

carbonic oxide 4- oxygen --> carbonic anhydride
i vol. \ vol. i vol.

Carbonic anhydride, on the other hand, parts with half its
oxygen and doubles in volume when it combines with
carbon to form carbonic oxide, thus :

carbonic anhydride + carbon — > carbonic oxide.

I VOl. 2 VOh.

Dalton decomposes carbonic anhydride by sparking. —

An independent proof of these relationships was obtained
by John Dalton of Manchester (1766-1844), who found
that carbonic anhydride (18 vols.) was decomposed by the
electric spark into carbonic oxide (16 vols.) and oxygen (9
vols.). The combustion of carbonic oxide is therefore a
reversible process, as shown by the equation :

Carbonic oxide + oxygen ^ carbonic anhydride
i vol. \ vol. i vol.

Dalton writes :

"Carbonic acid is decomposed by electricity into car-
bonic oxide and oxygen. I assisted Dr. Henry in an
experiment in which 52 measures of carbonic acid were
made 59 measures by 750 shocks; the gas after being
washed became 25 measures, whence these had arisen from
the decomposition of 18 measures of acid1; these 25
.measures consisted of 16 carbonic oxide and 9 oxygen ; for, a
portion being subjected to nitrous gas, manifested 1/3 of its
bulk to be oxygen ; and the rest was fired by an electric

1 59 ~ 25 ~ 34 measures of carbonic acid gas remained after sparking,
out of the 52 measures originally taken.

vin THE BURNING OF INFLAMMABLE GASES 145

spark, and appeared to be almost wholly converted into
carbonic acid. (New System of Chemical Philosophy, 1810,
II. 382.)

Berzelius and Dulong- (1820) calculate the composition
by weight of carbonic anhydride. — The earliest estimations
of the proportion by weight of carbon and of oxygen in
carbonic anhydride, were (like the early numbers for the
composition of water) dependent on a knowledge of the
densities of the gases concerned in the combustion.

As the volume of carbonic anhydride produced was equal to
the volume of oxygen used, the composition of the anhydride
could be deduced directly from a knowledge of the densities
of these two gases. This method was employed by Berzelius
and Dulong, working in Berthollet's laboratory at Arcueil
(Ann. de Chimie, 1820, 15, 393), in deducing the value
which was accepted as authoritative from 1820 to 1840.
Taking the density of oxygen as 1*1026 times that of air,
and that of carbonic anhydride as i'524, it followed from
the equality of the volumes of the two gases that 1*524
parts by weight of carbonic anhydride were made up of
1*1026 parts of oxygen and 1*524- 1*1026 = 0*42 14 part of
carbon. The percentage composition of the anhydride was
therefore :

Oxygen = 72'35% .

Lavoisier's combustion of charcoal. — Berzelius's numbers
were in close agreement with those which had been given
nearly 40 years earlier by Lavoisier ("Decomposition and
Recomposition of Water," 1783; Works, II. 345) as a
result of a careful quantitative investigation of the reduction
of red lead by charcoal. In this experiment the weight of
oxygen was deduced from the loss in weight of the oxide ;
the weight of fixed air or carbonic anhydride was calculated
from its volume and density. Lavoisier found that :

146 HISTORICAL INTRODUCTION TO CHEMISTRY CHAI-.

Grains. Grains.

Red lead 3>456 gave: Lead 3,108

Charcoal 432 Unburnt charcoal... 342

Fixed air 379*

3,829*
Los, 58!

Total 3,888

The loss was due to the production of water, the com-
position of which was reckoned to be :

Grains.

Oxygen Sl'°5

Inflammable air 770

5875

By subtracting the weight of hydrogen in the charcoaA and
the weight of oxygen used in burning it, these numbers give •

Grains.

Weight of carbon burnt 432 - 342 - 77 = 82*30

Weight of oxygen used to burn carbon 3,456 - 3,108 - 5 1*05 = 296*95

Fixed air (carbonic anhydride) =379 '25

The composition of carbonic anhydride is, therefore :
Oxygen = j 2-12 5%, say 72'1%
Carbon = 27-875% „ 27'9%

The burning of spirit of wine, of olive oil, and of a
candle. — Having established the composition of carbonic
anhydride, Lavoisier made use of his numbers to determine
the proportions of carbon and hydrogen contained in
various liquid and solid fuels. (" Combination of Oxygen
with Spirit of Wine," 1784; Works, II. 586-600.)

The following numbers were obtained by burning a small
spirit-lamp over mercury in a bell-jar filled with air l (Fig. 33)
and supplied with oxygen from a second bell-jar inverted
over water.

1 An explosion was always produced when the attempt was made to
burn the lamp in pure oxygen.

vin THE BURNING OF INFLAMMABLE GASES 147

Grains.

Spirit of Wine burnt
Oxygen used

Grains.
93*50
110-32

Total 203-82

Fixed air produced, 137*09
cubic inches, weighing 95*28

Water produced (by

difference) 108-54*

The fixed air contained :

Oxygen
Carbon

Grains.
68-60
26-68

95-28

The water contained

Oxygen
I lydrogen

Grains.
92-26
16-28

108-54

FIG. 33. — LAVOISIER'S APPARATUS FOR THE COMBUSTION OF SPIRIT OF WINE.

The oxygen in these compounds amounted to 160-86 grains,
whereas only 110-32 grains had been supplied, the spirit of
wine must therefore have contained :

1 This number agreed closely with the observation that 18 ounces of
water could be condensed from the products of combustion of 16 ounces
of spirit.

L 2

148 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

Oxygen 50-54 grains =54-1% J
Carbon 26 '68 ,, =28-5%
Hydrogen 16*28 ,, 3*17*4%

or, as Lavoisier preferred to express it :

Carbon 26 '68 grains =28-5%
Hydrogen 7-36 „ = 7-9%
Water 59-46 „ -63 '6%

Similar combustions of olive oil and of candle-wax gave

Oil. Wax.

Carbon 79% 82-3%

Hydrogen 21% 177%

Both substances contain oxygen, out the quantity was not
sufficient to be detected by Lavoisier's method of analysis.

Dumas and Stas (1841) on the composition of carbonic
anhydride. — About the year 1840 it became evident that
the numbers for the composition of carbonic anhydride, given
twenty years previously by Berzelius and Dulong, must be
erroneous, since careful analyses of compounds rich in
carbon often gave totals exceeding 100%. Thus two
typical HYDROCARBONS (i.e. compounds of hydrogen and
carbon) extracted from coal-tar gave the following figures
(Stas's Works, I. 270 and 275) :

Naphthalene. Benzene

Hydrogen 6'24% 770%

Carbon 95-40% 93 '53%

io i -64 101-23

"
In these substances the proportion of hydrogen was so

small that no large error could be produced by uncertainty
as to the weight of water produced, or the weight of hydrogen
which it contained. The excess above 100 per cent, could
only be due to over-estimating the proportion of carbon in
the carbonic anhydride formed on combustion. To test

1 The numbers now accepted for pure alcohol are : Oxygen 34 '8%,
Carbon 52-2%, Hydrogen 13-0% ; Lavoisier's spirit of wine evidently
contained about half its weight of water.

VIII

THE BURNING OF INFLAMMABLE GASES

149

this, Dumas and Stas
carried out in 1841 an
elaborate investigation of
the weight of carbonic
anhydride produced by the
combustion of different
forms of carbon (Ann. de
Chiniie, 1841, 1, 5 — in ;
Stas's Works, I. 235 — 287).
For their combustion
they used (i) Natural
graphite, freed from mineral
matter by igniting with
potash, extracting with
acids, and finally heating
to a white heat in a current
of chlorine gas (Chap. XI)
during twelve to fifteen
hours. (2) Arti ficial
graphite separated from
cast iron and purified in
the same way. (3) Dia-
monds, of which 10 to 12
grams were burnt ; these
were used mainly because
the purified graphite had
become so porous that it
was difficult to weigh it free
from condensed air and
moisture.1

1 The property, which char-
coal possesses in a remarkable
degree, of condensing gases in
its pores, was discovered by
Fontana in 1777 and has been
used largely in recent work.

a 2

150 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

The combustion was carried out in a current of oxygen,
which was freed from carbonic anhydride by means of
potash and dried by means of sulphuric acid. The carbon,
contained in a platinum boat, was burnt in a porcelain tube
heated strongly in a furnace (Fig. 34). The carbonic
anhydride was absorbed by potash and weighed directly,
with the help of absorption-apparatus devised by Liebig.
A small tube containing sulphuric acid was provided to
collect any water that might be produced, but in none of the
experiments was any gain in weight observed. The numbers
obtained were as follows ( Works, I. p. 252) :

Carbon. Carbonic Anhydride. Percentage of Carbon.

i -ooo 3-671 27-24

Natural 0-998 3 '660 27-27

Graphite 0-994 3-645 27-27

i -216 4'46i 27-26

i-47i 5*395 27 '27

0-992 3-642 27-24

Artificial 0*998 3-662 27-26

Graphite i'66o 6*085 27-28

1-465 5-365 27-31

0708 2-598 27-25

0-864 3*1675 27-27

Diamond 1-219 4*4^5 27-30

1-232 4-517 27-26

1-375 5'04i 27-28

Mean ...... 27-27

The composition of carbonic anhydride was therefore
Oxygen = 72'73%

The reduction in the percentage of carbon from 27-67 to
27-27 provided the correction that was needed in the
analysis of the hydrocarbons, which now gave :

Naphthalene H 6-24 C 94*02 Total 100-26

H 6*30 C 93*80 ,, roo'io

H 6-29 C 93-86 ,, 100-15

H 6-31 C 93*84 ,, 100-15

Benzene H 77 C 92*2 ,, 99-9

vin THE BURNING OF INFLAMMABLE GASES 151

Stas's analysis of carbonic oxide (1849). — Eight years
later, in 1849, Stas (Works, I. 287-308), carried out a
masterly analysis of carbonic oxide, in which 128-367
grams of oxygen were used in eight experiments to burn
224-683 grams of the gas to 353*050 grams of carbonic
anhydride. The experimental methods were very similar
to those used by Dumas to determine the composition of
water. Thus, instead of using free oxygen to burn the gas,
the oxygen was supplied by copper oxide which was reduced
to copper by heating it in a current of carbonic oxide. The
gas was not weighed directly : its weight was calculated from
the difference between the weight of oxygen used and the
weight of carbonic anhydride produced. The carbonic
anhydride was absorbed and weighed in combination with
potash.

The chief difficulty of the experiment was to provide a
supply of carbonic oxide free from carbonic anhydride,
oxygen, water, hydrogen and other inflammable gases. The
gas was prepared by heating oxalic acid with oil of vitriol :
the product, a mixture of carbonic oxide and carbonic
anhydride, was passed through potash into a reservoir
holding eighty to ninety litres (Fig. 35). From this
reservoir the gas was displaced, not by water, but by an
alkaline solution capable of absorbing both oxygen and
carbonic anhydride.1 The chief means of removing oxygen,
consisted, however, in passing the carbonic oxide through a
heated tube containing copper, in which the free oxygen
would have every opportunity of being converted into
carbonic anhydride. Both before and after passing through
this tube the gas was freed from carbonic anhydride and
from water by contact with potash and sulphuric acid. The
gas prepared in this way contained a little nitrogen but was
practically free from all other impurities.

1 A 5% solution of stannous oxide in potash, made with boiled
water, was used.

152 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

?

— ST

••*•"• *-^£) >^ -T3

o B9"g-o'y

p'S a-0 * rt

cc^s-s-a

'gl/a'S

-5-S'S^^^

I S 2«o

£3a§^

«o§<§2
•- jgrf^-J

.§&»« --^ ;t
li*-gf«8

•SliS'ia

„, 5'S.g 3 s

a a-^ ^ aa
o ^— ' c c

ilrfJ-S

5 §.:= «l s

•s<H,£.S
§W3,5S1

IB!^:|

.ri^lfi

•325.».M'|

(wei
wei
ontai

carbonic oxide over an al
ube containing hot coppe
n pumice ; J, Guard-tube
) containing copper oxide
) ; O, Potash on pumice (
eighed) ; R, Guard-tube

le as an inset.

litres) containing
d on pumice ; D, Glass tu
tash ; I, Sulphuric acid on
cm. long, 2^ cm. diameter)
ebig potash-bulbs (weighed)
lphuric acid on pumice (we
moisture from air.

m
la

^."2 o o.S'S'i
S S °^*> i
^.y"2^^C>o

25^5^:,,

OO
"3

o a * ££ o

vin THE BURNING OF INFLAMMABLE GASES 153

In these experiments, as in those of Dumas and of Dumas
and Stas, the purifying agents were distributed over pumice
stone, or broken glass, or were used in a granular form in
order to secure ample contact with the gas. The air in the
pumice stone was removed by repeatedly exhausting the
apparatus and filling it again with pure carbonic oxide.

The copper oxide, prepared by calcination, was contained
in a glass tube drawn out at each end. Taps were pro-
vided so that the tube could be exhausted before weighing.
During the reduction the tube was heated quite gently over
charcoal. In a series of blank experiments the tube was
found to vary in weight by only a milligram on a total of
540 grams when heated, provided that some hours were
allowed to elapse before re-weighing.

The gas escaping from the copper oxide tube was passed
through sulphuric acid to absorb any water that might be
formed ; the quantity collected was usually about ten milli-
grams. It then passed into a bulb containing potash,
where most of the carbonic anhydride was absorbed. A set
of Liebig bulbs filled with potash and three U-tubes, the
last filled with solid potash,1 were used to absorb the
remainder of the gas. The weight of carbonic anhydride
was corrected for the buoyancy of the air displaced by the
increased bulk of the potash after absorbing the gas.

The composition of carbonic oxide was shown by these
experiments to be :

Oxygen 57 '16%
Carbon 42'84%

C. MARSH GAS AND OLEFIANT GAS.

The burning of inflammable gases.— The existence of
inflammable gases other than those prepared by the action
of metals on the acids had been known from very early

1 The solid potash served to prevent the loss of moisture from the
absorbing apparatus.

154 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

times and was clearly recognised in van Helmont's work
on " gas pingue." But no definite chemical investigation
of these gases was made until the composition of water was
determined in 1781. It was then proved clearly by the
work of Cavendish, of Lavoisier, and of Monge that the gas
from metals and acids gave rise to water as the only product
of combustion, and in particular that no fixed air was pro-
duced. Every other inflammable gas that was examined
differed from hydrogen in giving rise to larger or smaller
quantities of fixed air. Water was also produced in nearly
every case, but Cruikshank in 1801, and De"sormes and
Clement in the same year, succeeded in preparing carbonic
oxide so far free from hydrogen that no water was produced
when it was burned.

Apart from hydrogen and carbonic oxide, every inflam-
mable gas gave both water and fixed air, and was therefore
composed of carbon and hydrogen, with or without oxygen.
Gases of this kind were prepared by heating wood, charcoal,
and oil, by passing spirit of wine through a red hot tube, and
so forth. They were examined by Cruikshank, by Berthollet,
and by a number of other workers, who determined their
relative densities, and their behaviour when exploded with
oxygen in Volta's eudiometer, as regards (i) the volume of
oxygen used, (2) the contraction after explosion, and (3) the
volume of carbonic anhydride produced. But as all the
gases examined were mixtures, no concordant results could
be obtained.

Olefiant gas or heavy carburetted hydrogen. — The first
gaseous compound of carbon and hydrogen to be prepared
and examined in a pure state was OLEFIANT GAS. This gas
was investigated in 1794 by a group of Dutch chemists
(" Researches on the different kinds of gas which are
obtained by mixing concentrated sulphuric acid with
alcohol," by J. R. Deiman, A. Paets van Troostwyck,
N. Bondt and A. Lauwerenburgh, Jour, de Physique, 1794,

vin THE BURNING OF INFLAMMABLE GASES 155

45, 178-191). They proved that it was a HYDROCARBON,
i.e., a compound of carbon and hydrogen only. They
distinguished it as " oily carburetted hydrogen," now
OLEFIANT GAS, from the property which it possessed of
combining with chlorine gas (Chapter XI) to form an oil
known as DUTCH LIQUID. The gas is now called ETHYLENE,
and the liquid ETHYLENE CHLORIDE.

The gas was also investigated by Dalton, who describes
it as follows :

"Olefiant gas may be produced by mixing 2 measures
of sulphuric acid with i measure of alcohol ; this mixture
in a gas bottle must be heated to about 300° [Fahrenheit] by
a lamp, when the liquid exhibits the appearance of ebullition,
and the gas comes over : it should be passed through water,
to absorb any sulphurous acid which may be generated."

" This gas is unfit for respiration, and extinguishes flame,
but it is highly combustible. . . . Olefiant gas burns with a
dense, white flame. It explodes with uncommon violence
when mixed with oxygen and electrified. . . . : unless a great
excess of oxygen be used, the charcoal is partly thrown down,
and it makes the gas turbid after explosion ; the result
in this case affords less carbonic acid than is due " (New
System, 1810, II. 438 and 440).

Composition of olefiant gas. — Dalton investigated the
volumetric relationships of the gas. He showed that it
combined with its own volume of chlorine (loc. cit., p. 439).
When exploded with an excess of oxygen, some carbon
remained unburnt, but a large contraction occurred, the
proportional volumes being (loc. tit., p. 440) :

olefiant gas + oxygen — -> carbonic anhydride + water ;

100 vols. 270 to 300 vols 185 or 190 vols.
or roughly

i vol. <3 vols. <2 vols.

When mixed with only its own volume of oxygen and
exploded, an expansion was produced instead of a contrac-

156 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

tion (loc. cit., p. 442), the carbon being burnt to carbonic
oxide, and the hydrogen liberated, thus :

olefiant gas + oxygen — >• carbonic oxide + hydrogen

I VOl. I VOl. <2 Vols. <2 Vols.

When sparked alone the gas was decomposed into carbon
and twice its volume of hydrogen, thus :

olefiant gas — > carbon + hydrogen
i vol. 2 vols.

Olefiant gas was thus shown to contain twice its volume of
hydrogen, and enough ca'rbon to give twice its volume of
carbonic oxide, or of carbonic anhydride.

Marsh gas or light carburetted hydrogen. — The forma-
tion of an inflammable gas during the decay of vegetable and
animal "matter was known from classical times, and was
noted by van Helmont, who described all such gases as
" gas pingue."

The first accurate work on the subject was that of
Volta who showed in 1776 (Letters on the Inflammable Air
of Marshes), that whilst the gas from metals and acids
required for its combustion twice its volume of air, the
MARSH GAS collected from putrid marshes required from ten
to twelve volumes. The gas was also examined about the
same time by Priestley in England and by Franklin in
America. It was investigated very carefully in 1804, by
Dalton, who gave to it the name CARBURETTED HYDROGEN.
Dalton describes its preparation and properties as follows :

" It was in the summer of 1804, that I collected at various
times, and in various places, the inflammable gas from ponds ;
this gas I found always contained some traces of carbonic
acid and a portion of azote ; but when cleared of these, it
was of a uniform constitution. After due examination, I was
convinced that just one half of the oxygen expended in its
combustion, in Volta's eudiometer, was applied to the

vni THE BURNING OF INFLAMMABLE GASES 157

hydrogen, and the other half to the charcoal. This leading
fact afforded a clue to its constitution."

" The properties of carburetted hydrogen are :

" It is unfit for respiration and for the support of combus-
tion.

" Its specific gravity when pure, from my experience, is
very near o'6.

" Water absorbs ^V of its bulk of this gas." (New System,
1810, II. 445-446.)

Olefiant gas, by contrast, had a specific gravity 0-9 relatively
to air, and was absorbed by water to the extent of \ of its
bulk.

Composition of marsh gas. — Dalton showed that of the
two volumes of oxygen required for the complete combustion
of marsh gas, one was required to burn the carbon, giving
rise to an equal volume of carbonic anhydride, whilst the
other was used for the combustion of the hydrogen. Marsh
gas, therefore, contained, like olefiant gas, twice its volume
of hydrogen ; but the carbon was only sufficient to produce
an equal volume (instead of two volumes) of carbonic
anhydride or of carbonic oxide. This view of the composi-
tion of the gas was confirmed by sparking it, when the
carbon was deposited and two volumes of hydrogen were
set free. Dalton writes :

" If 100 measures of carburetted hydrogen be put to up-
wards of 200 of oxygen, and fired over mercury, the result
will be a diminution of near 200 measures, and the residuary
100 measures will be found to be carbonic acid."

" When a portion of carburetted hydrogen gas is electri-
fied for some time, it increases in volume, in the end almost
exactly doubling itself; at the same time a quantity of
charcoal is deposited. The whole of the gas is then found
to be pure hydrogen" (ibid. p. 447).

When exploded with an equal volume of oxygen the carbon
was burnt to carbonic oxide and half the hydrogen was

158 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

burnt to water, leaving a gaseous residue consisting of
hydrogen and carbonic oxide in equal proportions.

The contrast between marsh gas and olefiant gas is shown
by the equations :

marsh gas (sparked) — >• carbon + hydrogen
i vol. 2 vols.

olefiant gas (sparked) — ^ carbon + hydrogen
i vol. 2 vols.

I marsh gas + oxygen — >• carbonic anhydride + water
i vol. 2 vols. i vol.

olefiant gas + oxygen — > carbonic anhydride + water
i vol. 3 vols. 2 vols.

Goal gas. — The inflammable COAL GAS obtained by distil-
ling coal consists largely of marsh gas. Dalton, in describing
the properties of marsh gas, writes :

" This gas is obtained nearly pure also^by distilling pit-coal
with a moderate red heat. It is now largely used as a
substitute for lamps and candles, under the name of coal gas.
According to Dr. Henry's analysis, coal gas does not usually
contain more than 4 or 5 per cent, of carbonic acid, sulphur-
etted hydrogen, and olefiant gas. The rest is principally
carburetted hydrogen, but mixed with some atoms of
carbonic oxide and hydrogen. The last portion of gas
driven off from pit-coal, seems to be entirely carbonic oxide
and hydrogen. The distillation of wood and of moist char-
coal, and many other vegetable substances, produces
carburetted hydrogen, but highly charged with carbonic acid,
carbonic oxide and hydrogen ; the two last gases always
appear exclusively at the end of the process " (ibid. pp. 445-

446).

SUMMARY AND SUPPLEMENT.
A. CARBONIC OXIDE.

Lassone, in 1776, prepared an inflammable oxide of carbon
by heating Prussian blue, and by heating zinc oxide with
charcoal,

ZnO -I- C _*> Zn + CO.

viii THE BURNING OF INFLAMMABLE GASES 159

It differed from hydrogen in that it burned with a beautiful
blue flame, and could not be detonated when mixed with air or
oxygen. The same gas was prepared by Priestley in 1783 by
heating smithy scale with charcoal

Fe3O4 + 4C -> 3Fe + 4CO.

Cruikshank, in 1801, prepared the gas, free from hydrogen,
by heating carbonated baryta (barium carbonate), or chalk
(calcium carbonate), with iron plates,

4BaCO3 + 3Fe -> 4BaO + 4CO + Fe3O4,
4CaCO3 + 3Fe -> 4CaO + 4CO + Fe3O4.

He showed that when exploded with oxygen in Volta's eudio-
meter it combined with half its volume of oxygen and gave an
equal volume of carbonic anhydride ; he therefore regarded it
as a lower oxide of carbon and called it "gaseous oxide of
carbon," now carbonic oxide,

2CO + O2 -> 2CO2

Carbonic oxide Oxygen Carbonic anhydride.
i i>oL \ vol. i vol.

De*sormes and Clement, in 1801, proved by similar experi-
ments that Lassone's inflammable gas was an oxide of carbon
which gave an equal volume of carbonic anhydride but no water
when burnt. They prepared it by the action of charcoal on
carbonated baryta,

BaCO3 + C-> BaO + 2CO,

a refractory carbonate which could not be decomposed by heat
alone. They could not prepare it by direct combustion of char-
coal in air or oxygen, but found that fixed air could be reduced
to carbonic oxide bv red-hot charcoal,

CO2 + C -> 2CO.

They also prepared a mixture of carbonic oxide and hydrogen
by the action of red-hot charcoal on steam,

H2O + C -> H2 -j- CO.

160 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

B. COMPOSITION OF THE Two OXIDES OF CARBON.

Priestley (1772), and Lavoisier (1772), showed that carbon
burns in air or oxygen without altering the volume of the
gas,

C + O2 -> CO2
i vol. i vol.

This fact was used by Berzelius and Dulong (1820) to
determine the composition of carbonic anhydride by comparing
its density with that of oxygen. The percentage of carbon
(27*65%) agreed with an early determination by Lavoisier (27*9%),
but was reduced to 27*27% by the classical experiments of
Dumas and Stas (1841), the details of which should be noted
carefully. Cruikshank (1801), De*sormes and Clement, and
Berthollet measured the changes of volumes produced byexplod-
ing carbonic oxide with oxygen ; they found that \ vol. of
oxygen was used and gave i vol. of carbonic anhydride.

CO + ^O2-> CO2
I vol. \ vol. i vol.

Since carbonic anhydride contains an equal volume of oxygen
carbonic oxide contains half this volume ; the volume of gas is
therefore doubled when carbonic anhydride is reduced by carbon
to carbonic oxide,

CO2 + C -> 2CO

I VOl. 2 VOls.

Stas, in 1849, determined the gravimetric composition of
carbonic oxide by passing the gas over weighed copper oxide
and weighing the carbonic anhydride produced,

CO + CuO -> Cu + CO2

The precautions taken to prepare a pure gas should be notea.
The proportion of carbon was found to be 57'16%.

C. MARSH GAS AND OLEFIANT GAS.

Marsh gas (methane, CH4) was discovered by Volta in
1776.

Olefiant gas (ethylene, C2H4) was prepared during the alche-
mistic period by the action of oil of vitriol on alcohol,

viri THE BURNING OF INFLAMMABLE GASES 161

C2H60 H20 -> C2H4

alcohol - water — > ethylene

It was shown to be a hydrocarbon by a group of Dutch chemists
who called it "oily carburetted hydrogen" (olefiant gas), from the
property which it possesses of combining with chlorine to form
an oily liquid,

C2H4 + C12 -> C2H4C12
ethylene + chlorine -» ethylene chloride

The volumetric relationships of these two gases were studied
by Dalton. They are as follows :

By sparking, carbon deposited, hydrogen (2 vols.) set free.
CH4 -» C + 2H2

I VOL 2

C2H4 -> 2C + 2H2

I VOl. 2 VOls.

By explosion with excess of oxygen

CH4 + 2O2 -> CO2 + 2H2O

marsh gas -f oxygen — > carbonic anhydride + water
i vol. 2 vols. \ vol. Contraction 2 vols.

C2H4 + 3O2 -> 2CO2 + 2H2O

ethylene + oxygen -> carbonic anhydride + water
i vol. 3 vols. 2 vols. Contraction 2 vols.

When exploded with less oxygen

CH4 + O2 -> CO +' H2 + H2O

marsh gas + oxygen —> carbonic oxide + hydrogen + water
i voL i vol. i vol. i vol.

C2H4 + O2 -> 2CO + 2H2

olefiant + oxygen — $> carbonic + hydrog
gas oxide

I VOl. I VOl. 2 VOls. 2 VOls.

Expansion 2 vols.

CHAPTER IX

SULPHUR AND PHOSPHORUS

A. SULPHUR, SULPHURIC ACID, AND THE SULPHATES.

Native sulphur. — The inflammable mineral known as
sulphur, or brimstone (Fig. 36), has been a familiar substance
from very early times. Its occurrence in the neighbourhood
of volcanoes, and the pungent smell which is produced when
it is burnt, caused it to be regarded during many centuries as
a symbol of the powers of evil. In alchemistic times it
acquired a n'ew significance as one of three principles,
mercury, sulphur, and salt, which took the place of the
four elements, earth, air, fire, and water, of the classical
writers. As representing the principle of inflammability,
sulphur was supposed to be present in all combustible sub-
stances, just as mercury was supposed to be present in all
metals ; much of the early work of Robert Boyle was de-
voted to proving the falseness of this assumption.

Sulphur from pyrites. — Sulphur is present in combination
with iron in the mineral pyrites or marcasite (Figs. 6, 7, 8,
pp. 8—9), and can be separated from it in part by
the action of heat. This method was used in the
middle ages for the artificial preparation of sulphur;
thus Mayow states that "Vitriols are produced from the
stone . . . called Marchasite, and from it on the application
of fire the flowers of common sulphur are elicited in con-
siderable abundance" (A.C.R. XVII. 28). The presence

162

CH. IX

SULPHUR AND PHOSPHORUS

163

of sulphur in the mineral is also shown by the pungent
odour of burning sulphur which is produced when pyrites
is burnt in a current of air ; at the same time the brilliant
metallic nodules of the mineral are converted into a red
rust-like ash or calx.

Sulphur present in oil of vitriol. — The presence of
sulphur in oil of vitriol was also recognised at an early
date. Basil Valentine described a method of preparing the

FIG. 36— RHOMBIC CRYSTALS OF SULPHUR. British Museum (Natural History).

acid from sulphur with the aid of saltpetre. Conversely,
Glauber was able to prepare " flowers of sulphur " from the
acid by heating its salts with coal, extracting with water, and
acidifying the solution. On account of this relationship the
French chemists in 1787 gave to oil of vitriol the name
SULPHURIC ACID, whilst its salts were described as

SULPHATES.

Stahl (1702) prepares a volatile acid from burning
sulphur. — The product formed by burning sulphur in air
is a pungent gas, which dissolves readily in water to a feebly

M 2

164 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

acid solution and combines with alkalis to form salts.
Stahl (1702) describes the process of collecting the acid
fumes with the help of linen or tow drenched in alkali ; he
states that the acid vapour is set free again by the action of
oil of vitriol, and that it is the weakest of the mineral acids,
though stronger than vinegar. In many respects the " sulphur-
ous acid " from burning sulphur resembles the " carbonic
acid " from burning charcoal ; but it differs in its obvious
acidity, in the greater solubility of the gas in water, and in
its power of bleaching coloured substances, such as red
rose leaves.

Stahl showed that the salts of the '• volatile " sulphurous
acid, prepared from burning sulphur, could be converted into
salts of the "fixed" oil of vitriol by exposing them to the
air. A similar experiment is described by Cavendish (A.C.R-
III. 9-11), who collected the fumes of burning sulphur
in milk of lime, obtaining a soluble salt which was converted
by exposure to the air into selenite or gypsum, the ordinary
sulphate of lime. In order to distinguish it from sulphuric
acid, the French chemists described Stahl's volatile acid as
SULPHUROUS ACID, and its salts as SULPHITES. The name
sulphurous acid is, however, now restricted to aqueous
solutions of the gas, the pungent gas itself being dis-
tinguished as SULPHUROUS ANHYDRIDE.

Priestley (1774) prepares a soluble gas from oil of
vitriol. — On account of its great solubility, the gas produced
by burning sulphur in air cannot be collected over water.
It was therefore not isolated until Priestley had developed
Cavendish's method of using mercury to confine those gases
which were too soluble to be stored over water.

Priestley prepared the gas in 1774 by heating oil of vitriol
with olive oil, and called it VITRIOLIC ACID AIR. He also
attempted, but without success, to separate it from oil of
vitriol by heat alone. In the course of this experiment
some mercury was drawn back into the hot acid,

ix SULPHUR AND PHOSPHORUS 165

thereby disclosing an excellent method of separating the
gas.

Priestley describes his observations as follows :

" When ... I got glass phials with ground stopples,
perforated, and drawn out into tubes, such as are represented
[in Fig 37], I found that heating the oil of vitriol in them
produced no air whatever."

" But though I got no air from the oil of vitriol by this
process, air was produced at the same time in a manner
that I little expected, and I paid pretty dearly for the
discovery it occasioned. Despairing to get any air from
the longer application of my candles, I withdrew them ;
but before I could disengage the phial from the vessel of

quicksilver, a little of it
passed through the tube
into the hot acid ; when,
instantly, it was all filled
with dense white fumes, a
prodigious quantity of air
was generated, the tube

FIG. 37— STOPPERED FLASK USED BY through which it was

S5r™ c^fiS*?™"* transmitted was broken into

many pieces (I suppose by

the heat that was suddenly produced) and part of the hot
acid being spilled upon my hand, burned it terribly, so that
the effect of it is visible to this day."

" Not discouraged by the disagreeable accident above-
mentioned, the next day I put a little quicksilver into the
phial with the ground stopple and tube, along with the oil
of vitriol ; when, long before it was boiling hot, air issued
plentifully from it ; and being received in a vessel of quick-
silver, appeared to be genuine vitriolic air, exactly like that
which I had procured before ; being readily -imbibed by
water, and extinguishing a candle in the same manner as
the other had done."

" Copper, treated in the same manner, yielded air very
freely, with about the same degree of heat that quicksilver
had required, and the air continued to be generated with
very little application of more heat. The whole produce

1 66 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

was vitriolic acid air, and no part of it inflammable"
(Experiments on Air, II. 16-20).

On account of its methods of preparation, Priestley con-
cluded that the volatile sulphurous acid was a"phlogisticated,"
or reduced, oil of vitriol.

Priestley describes the properties of vitriolic acid
air. — Priestley found that " vitriolic acid air," prepared by
the action of heat on a mixture of mercury and oil of vitriol,
resembled closely the " marine acid air " which he had
prepared from muriatic acid by the action of heat alone.
Both gases were intensely soluble in water,1 and each con-
densed to a white salt when mixed with the " alkaline air "
ammonia. The solution of " vitriolic acid air " in water was,
however, only weakly acid and the gas could be expelled by
gentle heat, or merely by exposure to the air ; in these
respects the properties of the solution were in marked con-
trast with those of muriatic acid and of oil of vitriol.
Priestley calls attention to this contrast as follows :

" Water being admitted to the vitriolic acid air absorbed
it about as readily as the marine acid air ; and by its union
with it must have formed the volatile or sulphureous acid of
vitriol" (ibid. p. 7).

"Water impregnated with marine acid air is, in all
respects, the very same thing with the common spirit of
salt, except that this acid may be made considerably
stronger in this manner than any spirit of salt made in
the common way, and that it has generally less colour.
But water impregnated with vitriolic acid air differs
most remarkably from oil of vitriol. Its acidity is now
become trifling to what it was ; and from being the most
fixed, and the strongest, it is now become the weakest, and
the most volatile of all acids ; the smell of it being intolerably
pungent, and almost the whole of it evaporating when it is
exposed to the open air" (Experiments on Air, 1777,
III. 272).

1 " Water absorbs about 20 times its bulk of this gas at a mean tem-
perature, according to my experience " (Dalton, Ntw System, II. 389).

ix SULPHUR AND PHOSPHORUS 167

The marked difference between vitriolic acid air and oil
of vitriol was attributed by Priestley to the fact that the oil of
vitriol had "combined with phlogiston." Lavoisier, how-
ever, showed that the change was due to removal of oxygen
from the oil of vitriol.

Priestley converts " vitriolic acid air " into oil of
vitriol. — Having prepared solutions of " vitriolic acid air,"
Priestley showed that they resembled oil of vitriol in their
power of dissolving metallic zinc (ibid. p. 274). He also
found that the volatile or " phlogisticated " acid could be
reconverted into ordinary oil of vitriol by exposure to air,
the acid being held in solution during the experiment by
combination with alumina.

" The volatile vitriolic acid, though produced from the fixed
vitriolic acid, is very considerably different from it, especially
as it may be dislodged from its basis by the vitriolic acid,
just as other weaker acids are dislodged by those that are
thence called the stronger. But that volatile vitriolic acid is
capable, however, of being brought back to the state of the
common vitriolic acid, and becoming the same thing that it
originally was, several experiments shew. At the time of
my last publication I had found that it was capable of dis-
solving iron and zinc, and of producing inflammable air,
which is the property of oil of vitriol : but I had a more
decisive proof of the same thing when, to water saturated
with vitriolic acid air, I had, for another purpose, put
some earth of alum till it was saturated. For, after six
months, in which this solution had been exposed
in an open phial, and one third of it was evaporated,
I observed many transparent crystals formed at the
bottom of the phial, as well as an incrustation on the sides
of the phial above the surface of the liquor. These crystals
were all triangular, of a considerable thickness, connected
with each other, and when examined appeared to be alum,
which is known to be the saline substance formed by the
same earth, and the proper vitriolic acid " {Experiments
and Observations, 1779, IV. 122-123).

168 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

He also showed that sulphur could be prepared by heat-
ing the solution in a sealed tube during a period of several
months (ibid. 124-129).

Composition of sulphurous anhydride. — It was observed
by Priestley in 1772 (Experiments on Air, I. 46) and con-
firmed by Dalton (New System, II. 391-392) that when
sulphur is burnt in air over mercury " no material change of
bulk is effected in the gas by the combustion ; and this is
also remarked in the analogous combustion of charcoal."
The pungent gas, which is now described as SULPHUROUS
ANHYDRIDE, thus resembles carbonic anhydride, in that it
contains its own volume of oxygen.

From this observation Dalton was able to calculate
the composition of the gas by weight. He had found
its density relatively to air to be 2*3, that of oxygen
being i'i. The gas was therefore twice as heavy as oxygen
and contained approximately equal weights of the two
constituents.

The fact that sulphur gained in weight when burnt was
announced to the French Academy by Lavoisier in 1772
(Works, II. 103) as the first of his great discoveries in
reference to combustion.

Lavoisier's experiments on the composition of sulphuric
acid. — Lavoisier recognised that sulphuric acid must be
regarded as a compound of sulphurous anhydride with
oxygen and with water. He attempted in 1777 (Works,
II. 194) to determine the proportions of the two gases by
acting on the acid with mercury, and collecting the gas
liberated (i) during this action and (2) during the decom-
position of the sulphate of mercury into sulphurous an-
hydride, oxygen, and mercury. The experiment was not
successful because a part of the sulphate of mercury
sublimed without decomposition, but it gave clear qualitative
proof that " volatile sulphurous acid is a vitriolic acid par-
tially deprived of oxygen."

ix SULPHUR AND PHOSPHORUS 169

Dalton (New System, II. 389-390) also attempted to
determine the proportions in which the two gases combine,
by uniting them over mercury in presence of water or with
the help of electric sparks. In the former case the action
was incomplete at the end of twelve days ; in the latter
case he found that "the mercury becomes oxidised, and
consequently liable to form a union with either of the
acids."

Gay-Lussac's analysis of sulphuric acid (1807).—
A successful analysis of sulphuric acid was made by
Gay-Lussac in 1807 in the course of an investigation on
the " Decomposition of the Sulphates by Heat " (Mem.
Soc. <?Arcueil> I. 215 — 251). He found that many of the
sulphates could be decomposed ; some were decomposed
easily, giving as products the oxide of the metal and white
fumes of SULPHURIC ANHYDRIDE ; others, requiring a higher
temperature for decomposition, gave off a gaseous mixture
of sulphurous anhydride and oxygen, leaving behind as
before a residue of oxide or earth. The quantitative
analysis of sulphuric anhydride was effected by decomposing
alum, the gases from the decomposition being collected
over mercury and analysed by absorption with potash.

Gay-Lussac writes :

" The first sulphate that I heated was sulphate of copper.
It first liberated water ; but as soon as the retort was red-hot,
white vapours of sulphuric acid x were produced, which were
accompanied by a nebulous gas, smelling strongly of
sulphurous acid, and in which after being washed a match
inflamed several times. This gas was then a mixture of
sulphurous acid x gas and oxygen . . . The two gases were in
volume almost in the ratio 2 to i ; but I will return later to
the exact determination of this ratio, and the way in which
the sulphuric acid decomposes" (pp. 217 — 218).

" Sulphate of iron undergoes the same decomposition by

1 In this passage the names "sulphurous acid" and "sulphuric acid"
are used to describe the anhydrides.

1 70 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

heat as sulphate of copper. The results are only modified
by this circumstance, that since the metal may take up a
higher degree of oxidation, there is liberated relatively more
sulphurous acid1 than oxygen" (p. 219).

" I took advantage of this decomposition of the sulphates
to determine the quantity of oxygen which must be added to
sulphurous acid to convert it into sulphuric acid. For this
purpose, I distilled over mercury some burnt alum, the base
of which neither produces nor absorbs any gaseous principle.
I collected gas at different stages of the distillation ; and
after taking a carefully measured volume, I washed it with
caustic potash and measured the residues. I thus found
that 100 parts

Of the 1st portion contained 32-33 of oxygen.
„ 2nd ,, „ 33-23

„ 3rd „ „ 32-53

„ 4* „ ,, 32-64

Mean . . 32-68

Since the proportion of the two gases was the same through-
out the whole course of the operation, one must conclude
that the decomposition of sulphuric acid by heat always
takes place in the same way, and that sulphurous acid
absorbs almost 0*5 of oxygen in passing to the state of
sulphuric acid " (pp. 236 — 237).

It was thus shown that sulphuric' anhydride consists of
sulphurous anhydride combined with one half its volume of
oxygen. Since sulphurous anhydride contained its own
volume of oxygen, the proportion of oxygen in sulphuric
anhydride was greater in the ratio 3:2.

B. SULPHURETTED HYDROGEN AND THE SULPHIDES.

Liver of sulphur. — Amongst the earliest discoveries of the
alchemistic period was the fact that alkalis possessed solvent
properties comparable with those of the acids. Thus Geber

1 In this passage the names " sulphurous acid " and " sulphuric acid "
are used to describe the anhydrides.

ix SULPHUR AND PHOSPHORUS 171

describes how sulphur may be dissolved in potash made
caustic with lime, and states that the clear solution becomes
milky on the addition of vinegar, yielding a sediment of
finely divided sulphur. Glauber, too, records the fact that
flint and rock-crystal, which are insoluble in acids, may be
dissolved by means of alkalis and precipitated again by
adding an acid.

The solutions of sulphur in alkalis were described as
LIVER OF SULPHUR and were used extensively in medicine
and in the preparation of pure MILK OF SULPHUR, made
by dissolving the sulphur in an alkali and then precipitating
it again with acid.

Sulphuretted hydrogen. — When liver of sulphur is acted
on by acids, only a portion of the sulphur can be recovered ;
a considerable part is lost in the form of a " foetid gas
having the odour of bad eggs " ( Hoffmann, 1772). This gas
possesses the property of blackening silver vessels (Boyle,
1663), is inflammable (Meyer, 1764), dissolves readily in
water,1 and is present in natural sulphur- waters, from which,
on exposure to the air, sulphur is deposited. On account of
its method of preparation from liver of sulphur it was
generally described as HEPATIC GAS. Scheele (on
" Stinking sulphureous Air," Essays on Air and Fire,
pp. 186-193) showed that it could be prepared from
sulphur by heating it in hydrogen, and that the sulphur
could be recovered from it by the action of oxidising
agents (nitric acid and chlorine) ; he therefore regarded
it as produced by the union of sulphur with phlogiston.
This view was modified by Lavoisier and his colleagues,
who regarded it as a compound of sulphur and hydrogen,
and gave to it the name of SULPHURETTED HYDROGEN.

Composition of sulphuretted hydrogen. — When sparked,
sulphuretted hydrogen is decomposed, sulphur being de-

1 "Water absorbs just its bulk of this gas ; when, therefore, it is mixed
with hydrogen, this last will be left after washing in water, or what is
still better, in lime-water " (Dalton, New System, II. 452).

172 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

posited and hydrogen liberated, without any change of
volume. Dalton writes :

"From the experiments of Austin, Henry, etc., it has
been established, that sulphuretted hydrogen undergoes no
change of volume by electrification, but deposits sulphur.
I have repeated these experiments, and have not been able
to ascertain whether there was increase or diminution. The
residue of gas is pure hydrogen " (New System, II. 452).

The gas therefore contains its own bulk of hydrogen.
The proportion of sulphur as well as of hydrogen can be
deduced from its behaviour when exploded.

" When mixed with oxygen, in the ratio of 100 measures
to 50 of oxygen (which is the least effective quantity), it
explodes by an electric spark ; water is produced, sulphur
is deposited, and the gases disappear. If 150 or more
measures of oxygen are used, then after the explosion over
mercury, about 87 measures of sulphurous acid are found
in the tube, and 150 of oxygen disappear, or enter into
combination with both the elements of the gas " (New

System, II. 452).

'

In the first case the hydrogen present in the gas is found
to require (like pure hydrogen) half its volume of oxygen for
combustion ; in the second case an additional volume of
oxygen is required to burn the sulphur, and rather less
than an equal volume of sulphurous anhydride is pro-
duced. The quantity of sulphur in the gas is therefore
about the same as in an equal volume of sulphurous
anhydride.

Metallic sulphides. — The fact that sulphur combines with
the metals was known from the earliest period of alchemy.
The compounds which result are similar in many respects
to the oxides ; they were described by Lavoisier and
his colleagues as SULPHURETS, but this name has now
been abandoned in favour of the name SULPHIDES.

ix SULPHUR AND PHOSPHORUS 173

Compounds of this type are of frequent occurrence
amongst minerals, such as :

Pyrites (Fig. 8) (Iron or copper and sulphur)

Cinnabar (Mercury and sulphur)

Galena (Fig. 38) (Lead and sulphur)

Blende (Zinc and sulphur)

The presence of sulphur was probably detected first in
pyrites, a mineral which is very rich in sulphur, and yields

FIG. 38. — LARGE CUBE OF GALENA.
The triangular faces on the top corners are parts of an octahedron.

considerable quantities of this element when distilled : in
other cases its presence was suspected on account of the
similarity of the mineral to artificial compounds of the
metals with sulphur.

The burning of sulphides. — When a metallic sulphide
is heated in air, the metal usually burns to the calx or oxide,
whilst the sulphur escapes in the form of sulphurous
anhydride. In this way the presence of sulphur may be
detected, even when it cannot be separated in the free state.

The burning of sulphide to oxide was used by Berzelius

174 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

in 1811-1812 (Ostwald's Klassiker, XXXV. 31-34), to
determine the proportions of iron and of sulphur in natural
iron pyrites, in order to compare it with the artificial
sulphide of iron prepared by heating iron with sulphur.
He found that :

9*93 grams of iron pyrites x gave 6'6o grams of oxide,
whilst, by an indirect method,

2 grams of artificial sulphide gave 1-82 grams of oxide,

the oxide containing 69^34% of iron. The composition of
these two substances was therefore :

Pyrites Iron 46*08 Sulphur 53*92%

Artificial sulphide „ 63 „ 37%

Conversion of sulphides to sulphates, — The sulphides
also oxidise slowly when exposed to air and water, but in
this case the product is usually a sulphate, i.e., a compound
of the calx or oxide of the metal with sulphuric anhydride.
On account of the readiness with which they unite with
oxygen, the solutions prepared by dissolving sulphur in potash,
soda, lime and baryta, and mixtures of iron filings with sulphur,
were used by Hales, Priestley, Scheele and others for
absorbing the active part of the air, as well as for
"diminishing" nitrous air, i.e., converting nitric into nitrous
oxide. It was not, however, until Lavoisier had developed
his "oxygen" theory of combustion that these changes
were satisfactorily explained. In a paper "On the Vitriol-
isation of Pyrites," published in 1777 ( Works, II. 209),
Lavoisier showed that the formation of green vitriol only
takes place in presence of air and can be stopped by covering
the pyrites with oil, or by keeping it under water. On
exposing the pyrites in a bell-jar of air over water, he found
that the vitriolisation was accompanied by a diminution in
1 10 grams of pyrites containing 0*07 gram of silica.

ix SULPHUR AND PHOSPHORUS 175

the volume of the air, which ceased after eighteen to twenty
days ; the air remaining in the bell-jar was found to have
been deprived of its oxygen, since it no longer supported
combustion, and in all respects resembled the " atmospheric
mofette " or azote. Lavoisier concluded that :

" The pyrites is a compound of sulphur and iron ; the act of
vitriolisation is nothing else than an addition of [oxygen] . . . ,
an addition which converts the sulphur into vitriolic acid ;
but, this acid being in contact with the finely-divided iron,
attacks and dissolves it as rapidly as it is formed, and iron
vitriol is produced" (Works, II. 211).

In the case of iron pyrites the action is rendered complex,
by the fact that the mineral contains more sulphur than is
required to convert the iron into sulphate. But in the case
of the other metals, e.g. lead, Berzelius found that the
proportions of metal and sulphur in the sulphide are exactly
the same as in the sulphate : the change from sulphide to
sulphate is then a direct combination of the sulphide with
oxygen.

The converse change of sulphates into sulphides was
accomplished by Glauber, who heated the vitriols with
charcoal, and thus converted them into " livers " from which
part of the sulphur could be precipitated by the addition of
acids.

Sulphuretted hydrogen as an acid, — Sulphuretted
hydrogen is most conveniently prepared by the action of acids
on artificial sulphide of iron (Scheele, Air and Fire, p. 193).
Dalton describes an improved method of carrying out this
process, as follows :

" The best way I have found to obtain sulphuretted
hydrogen in a pure state, is to heat a piece of iron to a
white or welding heat in a smith's forge, then suddenly
drawing it from the fire, apply a roll of sulphur ; the two
being rubbed together, unite and run down in a liquid form,
which soon fixes and becomes brittle. This compound or

176 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

sulphuret of iron, is to be granulated and put into a gas
bottle, to which dilute sulphuric acid is to be added,
after which the gas comes over plentifully. When the
sulphuret of iron is made in a crucible from iron filings and
sulphur, it seldom answers well ; it often gives hydrogen
mixed with the sulphuretted hydrogen" (New System, II.
450-451).

This action is very similar to that whereby muriatic acid
is liberated from salt by oil of vitriol, and served to suggest
that sulphuretted hydrogen might be regarded as a
feeble acid. This view was put forward, in I796,1 by
Berthollet (Ann. de Chimie, 1798, 25, 233-272), who showed
that the salts of silver and lead could be converted into their
insoluble sulphides by the addition of sulphuretted hydrogen,
just as silver is converted into its insoluble muriate by the
addition of muriatic acid, and lead into its insoluble sulphate
by the addition of sulphuric acid. These facts were seen to
be important when Lavoisier's oxygen-theory of acids became
the subject of criticism, for in spite of many endeavours no
oxygen could be detected in sulphuretted hydrogen itself,
nor in the sulphides derived from it. Berthollet sums up
his views as follows :

" Sulphuretted hydrogen dissolved in water, reddens
tincture of litmus, litmus-paper and tincture of radish ; it
combines with the alkalis, baryta, lime and magnesia ; it
forms with these substances compounds which exchange
bases when mixed with metallic solutions ; it decomposes
soap, displacing the oil from the alkali . . . ."

"Sulphuretted hydrogen possesses then all the properties
which characterise the acids. If several other common
properties did not demand a separate class of hydrogen-
compounds, it would undoubtedly be ranged amongst the
acids."

" I will not recall here the observations which I have
opposed to the opinion of those who pretend that acidity is

1 The paper was read before the Paris Academy, 21 Ventose,
An. IV.

ix SULPHUR AND PHOSPHORUS 177

an attribute which belongs only to oxygen, I will only add
that sulphuretted hydrogen contains no oxygen, and that it
differs very little, in its acid properties, from carbonic acid,
which contains nearly 76 per cent, of oxygen " (loc. at.,
237-238)

Sulphides and polysulphides. — Whilst artificial sulphide of
iron is converted by oil of vitriol into sulphuretted hydrogen
and sulphate of iron, iron pyrites is found to liberate sulphur
as an additional product. Sulphides of this kind, which
liberate sulphur when acted on by acids, are distinguished as
POLYSULPHIDES. The preparation of " milk of sulphur "
from " liver of sulphur " by the action of acids depends in
part on the fact that soluble polysulphides are formed when
sulphur is dissolved in an alkali.

C. PHOSPHORUS.

The preparation of phosphorus. — The inflammable
element PHOSPHORUS was prepared during the i7th
and 1 8th centuries, by a number of chemists, including
Brand, Kunkel, and Robert Boyle, by evaporating urine to
dryness, mixing it with sand, and heating strongly. It was
afterwards prepared from bones.1 Dalton describes the
process as follows :

" Phosphorus is usually prepared from the bones

of animals, which contain one of its compounds, phosphate
of lime, by a laborious and complex process. The bones
are calcined in an open fire; when reduced to powder,
sulphuric acid diluted with water is added ; this acid takes
part of the lime, and forms an insoluble compound, but
detaches superphosphate of lime, which is soluble in water.
This solution is evaporated, and the salt obtained is in a
glacial state. The solid is reduced to powder, and mixed with
half its weight of charcoal ; then the mixture is put into an

1 A full description of the method is given by Pelletier, Journal dc
Physique, 1785, 27, 26-32.

N

178 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

earthenware retort, and distilled by a strong red heat, when
the phosphorus comes over, and is received in the water
into which the tube of the retort is immersed " (New System,
1810, II. 240-241).

Properties of phosphorus. — Phosphorus is a wax-like
solid, which melts about blood-heat, and is so extremely
inflammable that it must be preserved under water. When
exposed to the air it undergoes a slow combustion, and
exhibits the luminosity from which it derives its name.
When gently heated it takes fire and burns very vigorously,
emitting white fumes.

The burning of phosphorus. — The burning of phosphorus
was investigated in 1772 by Lavoisier, who showed that it
gained in weight, in just the same way as sulphur and the
metals.

His observations are described in a " Memoir on the
Combustion of Phosphorus and on the nature of the acid
which results from this combustion " (Mew.. Acad. Sri., 1777,
p. 65 ; Works, II. 139), as follows :

" If phosphorus is ignited, by means of a burning-glass,
under a bell-jar immersed in mercury, one observes :

(1) That only a given quantity of phosphorus can be
burned in a given quantity of air, and that this quantity is
about i grain for 1 6 to 1 8 cubic inches of air.

(2) That, when this quantity has been burnt, the phos-
phorus is extinguished, and cannot be relighted, unless
brought into contact with a quantity of fresh air, which has
not been used for combustion.

(3) That fresh phosphorus, introduced under the same
bell-jar, burns no better than the first.

(4) That, during the burning of the phosphorus, there is
formed a great abundance of white flowers or flakes, like
very fine snow, which attach themselves everywhere to the
interior of the bell-jar, and are nothing else than solid
phosphoric acid.

(5) That, at the first moment of combustion, the air in
the jar expands considerably, on account of the heat of the

ix SULPHUR AND PHOSPHORUS 179

combustion; but that the same air then diminishes con-
siderably in volume, until, when the vessels are cold, it only
occupies four-fifths or five-sixths of the space which it
occupied before the combustion. If the flowers or white
flakes, which are found during this operation, are collected
and weighed before they have come into contact with fresh
air, and without allowing them to absorb moisture, one
observes that they are two and a half times as heavy as the
phosphorus used to produce them, in other words, that
from a grain of phosphorus 2\ grains of solid phosphoric
acid have been produced."

" The air which has thus been diminished as much as
possible by the combustion of phosphorus, is not denser
than atmospheric air ; its specific weight is even diminished
rather than increased ; it is no longer fit for the respiration
of animals nor for the combustion, nor inflammation of
substances" {Works, II. 139-140).

Phosphoric acid and the phosphates. — The white solid
prepared by burning phosphorus in air is now called
PHOSPHORIC ANHYDRIDE, the name PHOSPHORIC ACID being
reserved for the product obtained by the action of water
upon it. Lavoisier showed that phosphoric acid could be
prepared by the action of warm nitric acid upon phosphorus
(Works^ II. 277), as well as by burning it in air. He also
prepared a number of salts of phosphoric acid, and
described them as PHOSPHATES.

Phosphorous acid and the phosphites. — A different acid is
produced when phosphprus is allowed to smoulder, e.g. by
putting small pieces of phosphorus on the sloping sides of a
glass funnel, and letting the liquid drop into a bottle as
it is formed. The acid was examined by Sage (Mem. Acad.
Set., 1777, 91, 435). He found that the salts prepared by
neutralising it with soda or potash, unlike those prepared by
Lavoisier from burnt phosphorus, were not deliquescent, i.e.
did not become liquid by absorbing moisture from the
air. Gengembre noticed that the freshly-prepared acid
" is still luminous in the dark, and retains a slight odour of

N 2

i8o HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

garlic " ; if " one heats it in an open vessel, small flames
arise from it from time to time " (Gengembre, Mem.
Math. Phys., Paris Academy, 1785, X. 652). The new
acid, which contains less oxygen than phosphoric acid,
was distinguished by the French chemists in 1787 as
PHOSPHOROUS ACID and its salts as PHOSPHITES.

Phosphoretted hydrogen. — An inflammable compound
of phosphorus with hydrogen was discovered in 1783 by
Gengembre, whilst attempting to prepare a " liver " of
phosphorus by the action of alkalis upon it. The discovery
is described as follows :

" I put some fixed caustic vegetable alkali to digest over
phosphorus; at the end of some hours, I saw a multitude
of minute bubbles, which adhered to the surface of the
phosphorus : then I exposed the whole to a heat of 35 to
40 degrees, to accelerate the action of the alkali. Scarcely
was the phosphorus melted, when there arose an unbear-
able odour of decayed fish, and a considerable quantity of a
peculiar gas, which took fire explosively of its own accord,
as soon as it came in contact with the air " (" Memoir on
a new gas obtained by the action of alkalis on Kunkel's
Phosphorus." Read at the Academy, May 3, 1783 ; Mem.
Math. Phys., 1785, X. 651-658; p. 652 ; compare CrelPs
Chemische Annalen, 1789, I. 450-457 ; p. 451).

The gas liberated by the action of the alkali at lower
temperatures did not take fire spontaneously (loc. tit. p. 655).
The preparation and properties of the gas were described
by Dalton as follows :

" Let an ounce or two of hydrate of lime (dry slacked
lime) be put into a gas bottle or retort, and then a few small
pieces of phosphorus, amounting to 40 or 50 grains. If the
materials are sufficient to fill the bottle, no precaution need
be used ; but if not, the bottle or retort should be previously
filled with azotic gas, or some gas not containing oxygen, in
order to prevent an explosion. The heat of a lamp is then
to be applied, and a gas comes which may be received

ix SULPHUR AND PHOSPHORUS 181

over water. This gas is phosphuretted hydrogen ; but some-
times mixed with hydrogen. — Liquid caustic potash may be
used instead of hydrate of lime, in order to prevent the
generation of hydrogen.

Phosphoretted hydrogen gas has the following pro-
perties: (i) When bubbles of it come into the atmosphere,
they instantly take fire ; an explosion is produced, and a ring
of white smoke ascends, which is phosphoric acid : (2) It
is unfit for respiration, and for supporting combustion : (3)
Its specific gravity is 0*85, common air being denoted by
unity : (4) Water absorbs -^th of its bulk of this gas :
(5) If the gas be electrified, the phosphorus is thrown
down, and there finally remains the bulk of the gas of pure
hydrogen " (New System, II. 456-457).

The same gaseous compound is formed when a solution
of phosphorous acid is heated.1 (Gengembre, 1783, he. tit.)
Water is evaporated, phosphoretted hydrogen escapes and
finally a residue of phosphoric acid remains. In this action
the phosphorous acid undergoes a process of simultaneous
oxidation and reduction, one portion being oxidised to
phosphoric acid by another portion, which loses all its
oxygen, and is converted into phosphoretted hydrogen.
A similar process occurs when the gas is made from phos-
phorus by the action of an alkali : in this case a part of the
phosphorus is oxidised with the help of oxygen derived
from the decomposition of water, whilst another portion
combines with the hydrogen of the water to form phosphor-
etted hydrogen.

Phosphoretted hydrogen not an acid. — Berthollet points
out in his memoir on " Sulphuretted hydrogen " that :

"Water in which phosphoretted hydrogen is dissolved,
shows no sign of acidity ; and solutions of potash, lime and
ammonia, do not appear to absorb more of the gas than
pure water."

1 Kopp attributes this discovery to Pelletier (1790), seven years after
Gengembre's paper was read at the Paris Academy, but I have not
been able to find any reference to it in Pelletier's papers.

182 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

" Phosphoretted hydrogen has then no acid properties ;
and this is its principal difference from sulphuretted hydro-
gen. It follows : (i) that the gas is liberated as fast as it is
formed, whilst sulphuretted hydrogen is held in combination
with the alkali and the water ; (2) that the alkaline phos-
phides are decomposed in contact with water" (Ann. de
Chimie, 1798, 25. 267).

SUMMARY AND SUPPLEMENT.
A. AND B. SULPHUR.

Sulphur or brimstone is found in the free state and also in
combination with metals in the mineral sulphides, such as :
Iron pyrites FeS2 (persulphide of iron)
Copper pyrites FeCuS2 (sulphide of iron and copper)
Cinnabar HgS (sulphide of mercury)

Galena PbS (sulphide of lead)

Blende ZnS (sulphide of zinc).

A part of the sulphur in iron pyrites can be distilled out :

3FeS2 -> Fe3S4 + S2.

The whole of the sulphur in the mineral sulphides can, however,
be removed by burning, whereby both metal and sulphur are
converted into oxide, e.g. :

3FeS2 + 8O2 -> Fe3O4 + 6SO2
2ZnS + 3O2 -> 2ZnO + 2SO2.

Sulphur burns in air, without producing any change of volume,
to a pungent gas, sulphurous anhydride (sulphur dioxide, SO2) :

S + O2 -> SO2.
i vol. i vol.

The gas dissolves freely in water, forming a weak volatile acid,
sulphurous acid, H2SO3, but is expelled by boiling and by
exposure to air. The acid can be fixed by means of alkalis
or bases, with which it combines to form salts known as
sulphites, e.g. :

Sulphite of soda or sodium sulphite NagSOg
Sulphite of potash or potassium sulphite K2SO3
Sulphite of lime or calcium sulphite CaSO3.

ix SULPHUR AND PHOSPHORUS 183

The volatile acid and its salts were prepared in 1702 by Stahl,
who showed that the salts were converted by exposure to air
into salts of the fixed oil of vitriol. This action may be
represented by the equation :

2K2SO3 + O2 -> 2K2SO4.

Potassium Potassium

sulphite sulphate

The gas was isolated in 1774 by Priestley, who obtained it from
oil of vitriol by the reducing action of olive oil or of mercury :
Hg + 2H2SO4-> HgSO4 + SO2 + 2H2O

Mercury -f Sulphuric — > Mercuric + Sulphurous + Water
acid sulphate anhydride

The gas is freely soluble in water, but could be collected and
preserved over mercury.

Lavoisier (1777) showed that oil of vitriol might be regarded
as a compound of sulphurous anhydride with oxygen and water.
He attempted to analyse it by heating it with mercury and
collecting all the products, including the mercury which was
finally recovered from the action :

Hg + 2H2SO4 -> HgSO4 + SO2 + 2H2O
HgS04 -> Hg + 02 + S02.

Gay-Lussac, in 1807, carried out a successful analysis by
heating the sulphate of alumina in the form of alum, whereby he
obtained a mixture of sulphurous anhydride with half its volume
of oxygen :

2A12(SO4)3 -> 2A12O3 + 6SO2 + 3O2.

Sulphate of — > Alumina + Sulphurous + Oxygen
alumina anhydride

As sulphurous anhydride contains its own volume of oxygen,
the proportion in sulphuric anhydride must be 50 per cent,
greater.

Sulphuretted hydrogen (SH2) is set free in the form of a
fcetid gas by the action of acids on sulphides, such as the
artificial sulphide of iron, which is produced by fusing the metal
with sulphur :

FeS + H2SO4 -> FeSO4 + H2S.

Sulphide of + Oil of — > Ferrous + Sulphuretted
iron vitriol sulphate hydrogen

1 84 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

The gas is a very weak acid, but is able to precipitate insoluble
sulphides from solutions of many metallic salts, thus :

Silver 2AgNO3 + H2S -> Ag2S + 2HNO3
Lead Pb(NO3)2+ H,S -» PbS + 2HNO3
Copper CuSO4 + H2S -> CuS + H2SO4.

In other cases, however, the action proceeds in the opposite
direction, the sulphide being dissolved by the mineral acid and
sulphuretted hydrogen set free, thus :

Iron FeS + H2SO4 -> FeSO4 + H2S
Zinc ZnS + H2SO4 -> ZnSO4 + H2S.

Liver of Sulphur is prepared by dissolving sulphur in an
alkali or alkaline earth. The primary action is probably the
formation of sulphite and sulphide, thus :

3Ca(9H)2 + 38 -> CaSO3 + 2CaS + 3H2O.

Calcium Calcium Calcium

hydroxide sulphite sulphide

By the addition of acid, sulphur dioxide and sulphuretted
hydrogen are produced :

CaSO3 + 2CaS + 6HC1 -> 3CaCl2 + SO2+ 2H2S + H2O ;

these then interact in the solution to produce sulphur and
water :

SO2 + 2H2S -» 38 + 2H2O.

In the formation of liver of sulphur the converse action takes
place, sulphur and water being converted into sulphur dioxide
and sulphuretted hydrogen ; this reversal of the usual action may
be attributed to the affinity of the alkali for the two gases, both
of which possess marked acid qualities.

Further quantities of sulphur may be dissolved owing to the
production of thiosulphate and poly sulphides, thus :

CaSO3 + S-> CaS2O3

Calcium Calcium

sulphite thiosulphate

CaS + 8* -> CaS^.,.

Calcium Calcium

sulphide polysulphide

ix SULPHUR AND PHOSPHORUS 185

These substances liberate their additional sulphur when acids
are added, giving an increased yield of milk of sulphur.

C. PHOSPHORUS

Phosphorus, first prepared by distilling urine with sand, burns
very readily to a white, snowy oxide, increasing greatly in weight
(Lavoisier, 1777) :

P4 + 502 -> P4010.

The white oxide, phosphoric anhydride, P4O10, dissolves in water
to phosphoric acid,

P4010 + 6H20 -> 4H3P04,
from which series of phosphates 'may be prepared, thus :

Sodium phosphates, NaH2PO4, Na2HPO4, Na3PO4
Potassium phosphates, KH2PO4, K2HPO4, K3PO4
Calcium phosphates, CaH4(PO4)2, Ca2H2(PO4)2, Ca3(PO4)2.

The smouldering of phosphorus gives rise to phosphorous
acid,

P.J + 3O2 + 6H2O -> 4H3PO3,

from which phosphites may be prepared, e.g. :

Sodium phosphite, Na3PO3
Potassium phosphite, K3PO3.

In attempting to prepare a liver of phosphorus, Gengembre,
in 1783, obtained an inflammable gaseous phosphoretted hydrogen
Q\: phosphine, PH3 :

P4 + 3KOH + 3H20 -> 3KH2P02 + PH3.

Potassium Phosphine
hypophosphite

The same gas is obtained by heating phosphorous acid :
4H3P03 -> H3P + 3H3P04.

Phosphorous Phosphine Phosphoric
acid acid

The gas prepared by Gengembre, by the action of hot alkali on
phosphorus, was spontaneously inflammable on account of the
presence of the vapour of a liquid phosphoretted hydrogen, P2H4.

CHAPTER X

NITRE, NITRIC ACID, AND NITROGEN

A. NITROGEN

Speculations as to the nature of nitrous air and of
azote. — In the preceding chapters the nature and properties
of some of the common gases have been described. Thus
Black's " fixed air" has been shown to be a product of combus-
tion, a compound substance containing carbon and oxygen.
Cavendish's "inflammable air," on the other hand, is a
simple substance, which unites with oxygen to form water.
The nature of Priestley's "nitrous air" still remains,
however, to be discussed.

The fact that nitrous air combines with oxygen to form
brown nitrous fumes, and with oxygen and water to form nitric
acid suggested that the gas might be the "elementary
principle " from which nitric acid was derived, just as
carbonic, sulphuric, and phosphoric acids were derived from
carbon, sulphur, and phosphorus respectively. On the other
hand, the fact that nitrous air could be " diminished " by iron
filings and sulphur certainly suggested that it contained
oxygen; the product, it is true, did not behave like a
deoxidised nitrous air, since it supported combustion almost
as well as oxygen itself; but it undoubtedly contained
oxygen, which could scarcely be derived from any source
but the nitrous air.

CH. x NITRE, NITRIC ACID, AND NITROGEN 187

Assuming then that nitrous air already contained oxygen,
what was the nature of the substance with which the oxygen
was combined ? and what was the fundamental principle of
nitric acid and the nitrates ? There was much vague
evidence to show that this principle might be identified
with " azote," the inactive residue left after depriving air of
its oxygen. Thus Priestley had recorded the fact that :

" Nitrous air, which had been confined above a year in
contact with iron, standing in water, was in all respects like
phlogisticated common air : it neither diminished common
air, nor was diminished by nitrous air, and extinguished a
candle" (Experiments o?i Air, 1775, ti» r77)'

Cavendish, too, had noticed that charcoal deflagrated with
nitre produced an air which " as far as I can perceive . . .
differs in no respect from common air phlogisticated "
(A.C.R. III. 21). But in neither case was the evidence
sufficiently definite to justify the view that nitrous air was
an oxide of azote, or nitric acid a compound of azote with
oxygen and water.

This inert gas had, indeed, defied almost all attempts to
change it or to ascertain its nature. Priestley, after many
experiments on diminishing the volume of gases, could not
find any substance which would diminish common air to a
larger extent than one-fifth, although nearly every other gas
that he had handled could by some means or other be
condensed almost entirely, e.g. :

acid air and alkaline air by water,
fixed air by lime-water or by potash,
nitrous air by oxygen and water,

and so forth. It was, however, one of his random observa-
tions, on the diminution of air by sparking (Experiments and
Observations, 1779, IV. 284-287), that enabled Cavendish
to solve the problem of the nature of azote, and of the various
nitrous compounds.

1 88 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

Cavendish (1781) finds- nitric acid in the water produced
by the explosion of hydrogen and oxygen. — By burning
hydrogen in air, Cavendish, in 1781, obtained 135 grains of
" pure water " " which had no taste nor smell " ; but when he
exploded hydrogen with an excess of oxygen the 30 grains
of water which he obtained " was sensibly acid to the taste,
and by saturation with fixed alkali and evaporation, yielded
near two grains of nitre ; so that it consisted of water united
to a small quantity of [nitric] acid" ^A.C.R. III. 16).

The explosion of hydrogen with an excess of oxygen always
led to the production of nitric acid, even when the oxygen
was prepared without using a nitrate, e.g., from red lead and
oil of vitriol, or from the leaves of plants by exposing them
to sunlight in a vessel filled with water. But when the
proportion of oxygen was reduced, so that only about 2% of
it remained unburnt, " the condensed liquor was then not
acid, but seemed pure water." Cavendish also found that
" when inflammable air was exploded with common air ...
the condensed liquor was not in the least acid" (A.C.R. III.
1 8), although a large excess of oxygen was provided. The
production of nitric acid was therefore dependent (i) on
the presence of an excess of oxygen, and (2) on the high
temperature reached in the explosion of hydrogen with
oxygen.

As it seemed improbable that the nitric acid could be
derived from the oxygen, Cavendish attributed its formation
to the oxidation of azote (present as an impurity) during
the explosion.

Cavendish (1784) prepares nitric acid by sparking air.—
The high temperature which Cavendish thought to be

1 Priestley, in 1788, discovered nitrate of copper in the moisture pro-
duced by exploding inflammable air with common air in a copper vessel ;
in one experiment 442 grains of the green liquor contained enough nitric,
acid (free and combined) to yield 23 grains of nitre (Experiments and
Observations on Acidity, Water ; and Phlogiston, Phil. Trans., 1788,
78, 147-157, 313~330)-

NITRE, NITRIC ACID, AND NITROGEN

t89

necessary for the combination of azote with oxygen can be
obtained without the addition of inflammable air or of any
other foreign substance by using electric sparks. Priestley
had noticed that these diminished the volume of air and
thought that fixed air was produced. Cavendish confirmed
the diminution of volume, but found that the production of
fixed air was due "to the burning of some inflammable
matter in the apparatus," and "that the real cause of the

FIG. 39.— CAVENDISH'S APPARATUS FOR SPARKING AIR
OVER MERCURY.

diminution " was the conversion of azote into nitric acid.
Cavendish writes :

" The apparatus used in making the experiments was as
follows. The air through which the spark was intended to
be passed was confined in a glass tube M, bent to an angle,
[as in Fig. 39], which, after being filled with quicksilver, was
inverted into two glasses of the same fluid, as in the figure.

By means of a small glass tube suitably bent :

" I was enabled to introduce the exact quantity I pleased
of any kind of air into the tube M "; and by the same means,

190 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

I could let up any quantity of soap-lees, or any other liquor
which I wanted to be in contact with the air."

" The bore of the tube M used in most of the following
experiments was about one-tenth of an inch ; and the
length of the column of air, occupying the upper part of the
tube, was in general from i J to f of an inch."

" When the electric spark was made to pass through
common air, included between short columns of a solution
of litmus, the solution acquired a red colour, and the air
was diminished, conformably to what was observed by
Dr. Priestley."

" When lime-water was used insteadof the solution of litmus,
and the spark was continued till the air could be no further
diminished, not the least cloud could be perceived in the
lime-water : but the air was reduced to two-thirds of its
original bulk ; which is a greater diminution than it could
have suffered by mere phlogistication, as that is very little
more than one-fifth of the whole."

" When the air is confined by soap-lees, the diminution
proceeds rather faster than when it is confined by lime-
water ; for which reason, as well as on account of their
containing so much more alkaline matter in proportion to
their bulk, soap-lees seemed better adapted for experiments
designed to investigate the nature of this acid, than lime-
water" (A.C.R. III. 39-43).

Cavendish found that the diminution of volume on
sparking either azote or pure oxygen separately over soap-
lees was very small ; but when five parts of oxygen " were
mixed with three parts of common air,1 almost the whole of
the air was made to disappear " (A.C.R. III. 44). The soap-
lees " being evaporated to dryness, yielded IT%- grains of salt,
which is pretty exactly equal in weight to the nitre which that
quantity of soap-lees would have afforded if saturated with
[nitric] acid. This salt was found, by the manner in which
paper dipped into a solution of it burned, to be true nitre "

1 In this experiment 5 + (i x 3) = 5 ^ volumes of oxygen combined
with 1x3 = 2-2; volumes of azote, i.e. 7 volumes of oxygen combined
with 3 of azote.

x NITRE, NITRIC ACID, AND NITROGEN 191

(A.C.R. III. 45). It was free from vitriolic and muriatic
acids, but gave a precipitate with a solution of silver owing
to the fact that a part of the nitre was in the " phlogisti-
cated" (i.e. reduced or deoxidised) condition, to which it
is brought when heated so as to drive off a part of its oxygen
(see below, p. 199).

The composition of air. — It has been seen that air contains
about one-fifth of oxygen and a trace (about 1/2500) of fixed
air or carbonic anhydride, as well as a variable quantity of
moisture. The inert residue, of " azote " is much less easy
to analyse. Cavendish writes :

" As far as the experiments hitherto published extend, we
scarcely know more of the nature of the phlogisticated part
of our atmosphere, than that it is not diminished by lime-
water, caustic alkalis, or nitrous air; that it is unfit to
support fire, or maintain life in animals ; and that its specific
gravity is not much less than that of common air" (A.C.R.
III. 49).

His experiments had shown that it consisted largely of
NITROGEN,1 i.e. of a gas which can be converted into nitre
by sparking with oxygen in presence of potash. "Yet it
might fairly be doubted whether the whole is of this kind,
or whether there are not in reality many different substances
confounded together by us under the name of phlogisticated
air.2 I therefore made an experiment to determine, whether
the whole of a given portion of the phlogisticated air of
the atmosphere could be reduced to [nitric] acid, or whether
there was not a part of a different nature from the rest,
which would refuse to undergo that change" (A.C.R. III.

49)-

For this purpose Cavendish took the usual mixture of air
and oxygen and diminished it as much as possible by sparking.

1 The name was invented by the French chemist and calico-
printer Chaptal about 1790 (see the Preliminary Discourse of his
Elements of Chemistry, tr. 1791, pp. xxxiv. to xxxvi.).

2 i.e. Azote.

192 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

He then added some more oxygen to the residue and
continued to spark till no further diminution took place.
The excess of oxygen was absorbed by means of " liver of
sulphur,"

"After which only a small bubble of air remained un-
absorbed, which certainly was not more than TJy- of the
bulk of the phlogisticated air l let up into the tube ; so that
if there is any part of the phlogisticated air l of our atmos-
phere which differs from the rest, and cannot be reduced
to [nitric] acid, we may safely conclude, that it is not
more than yi^- part of the whole" (A.C.R. III. 50).

For 1 10 years this bubble of gas was thought to be merely
a residue of unabsorbed nitrogen; but in 1895 Rayleigh
and Ramsay found that the atmosphere actually contained
about 0-93% of a gas, ARGON, even more inert than
nitrogen ; there is therefore no doubt that the bubble
which Cavendish noticed was actually a part of the
azote " which differs from the rest, and cannot be reduced
to nitrous acid." His cautious statement has, indeed, proved
to be entirely accurate, in contrast with the hasty assump-
tion of his successors that the azote consisted entirely of
nitrogen.

B. COMPOSITION OF THE GASEOUS OXIDES OF NITROGEN.

Oxides of nitrogen. — The experiments of Cavendish
showed that nitrogen, the chief constituent of azote, could
be converted into ordinary nitre by sparking it with oxygen
over potash. This gas was therefore the characteristic
constituent of nitric acid. The three gases which Priestley
had prepared from nitric acid probably contained the same
constituent in combination with oxygen. Experiments were
therefore made by different workers to ascertain their com-
position.

1 i.e. Azote.

x NITRE, NITRIC ACID, AND NITROGEN 193

Composition of nitrous oxide.— The first successful
analysis was carried out by Davy (circ. 1800) in the case of
the diminished nitrous air or " laughing gas," to which he
gave the name NITROUS OXIDE. The gas was analysed by
burning charcoal in it : the proportion of oxygen was
calculated from the volume of carbonic anhydride produced,
whilst the nitrogen left after the combustion was measured
directly. The gas, generated from nitrate of ammonia,
" was in its highest state of purity, as it left a residuum of
^§- only, when absorbed by boiled water."

" Ten cubic inches of it were inserted into a jar graduated
to OT cubic inch, containing dry mercury. Through this
mercury a piece of charcoal which had been deprived of its
hydrogen by long exposure to heat, weighing about a grain,
was introduced, while yet warm. No perceptible absorption
of the gas took place."

" Thermometer being 46°, the focus of a lens was thrown
on the charcoal, which instantly took fire, and burnt vividly
for about a minute, the gas being increased in volume.
After the vivid combustion had ceased, the focus was
again thrown on the charcoal ; it continued to burn for
near ten minutes, when the process stopped."

" The gas, when the original pressure and temperature
were restored, filled a space equal to 12*5 cubic inches"
(Davy's Works, III. 59-60).

An examination of the residue showed that 5*2 cubic inches
of the gas had been decomposed, giving rise to 2*4 cubic inches
of carbonic anhydride (condensed by the addition of
ammonia) and 5*1 cubic inches of nitrogen, whilst the other
4'8 cubic inches were unchanged and could be dissolved out
by means of boiled water. Knowing the proportion of
oxygen in the carbonic anhydride, and the densities of
nitrogen and of nitrous oxide, Davy was able to conclude :

" that 100 grains of nitrous oxide are composed of 37
oxygen and 63 nitrogen : existing in a much more con
densed form than when in their simple forms."

o

194 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

It is, however, a simpler matter to deduce the composition
of the gas by volume. Priestley and Lavoisier had shown
that charcoal burns in oxygen without producing any change
in the volume of the gas; 2-4 cubic inches of carbonic
anhydride would thus contain 2*4 cubic inches of oxygen
derived from nitrous oxide. The 5*2 cubic inches of nitrous
oxide, therefore, contained 5-1 cubic inches of nitrogen, and
2 -4 cubic inches of oxygen; or, speaking approximately,
nitrous oxide contains its own volume of nitrogen and half its
volume of oxygen.

Composition of nitrous air or nitrous gas (nitric oxide).
—Davy next turned his attention to the gas which Mayow
had prepared by the action of iron on nitric acid, and
Priestley had described under the name of " nitrous air."
This gas was described by Davy in 1800, and by Gay-
Lussac in 1809 and in 1816 as "nitrous gas." The modern
name of NITRIC OXIDE was introduced about the year 1818.

It had been shown by Priestley in 1786 (Experiments
and Observations, 1786, VI. 304) that iron increased in
weight when heated by means of a lens in a jar of nitrous
air ; at the same time the volume of gas was diminished
about one half, the residue consisting of an inactive gas
resembling atmospheric nitrogen. Charcoal heated in the
same way produced carbonic anhydride and nitrogen (loc.
cit. p. 434), but the results were less trustworthy. Davy
repeated the experiment with charcoal and found that,
after applying the burning glass for four hours, 15-4 cubic
inches of nitrous gas gave 8*7 cubic inches of carbonic
anhydride and 7-4 cubic inches of nitrogen. From
these figures he calculated "that 100 grains of nitrous
gas contain 56-5 oxygen and 43-5 nitrogen" (Davy's
Works, III. 77-79). The composition by volume may
be deduced by assuming that the carbonic anhydride
contained its own volume of oxygen derived from the nitrous
gas. The 1 5 '4 volumes of nitrous gas would then contain

x NITRE, NITRIC ACID, AND NITROGEN 195

87 volumes of oxygen and 7*4 volumes of nitrogen, i.e. one
volume of nitrous gas contained rather more than half its
volume of oxygen and rather less than half its volume of
nitrogen. It was left for Gay-Lussac in 1809 to show that
the ratios i : \ : J represent the composition of the gas by
volume more accurately than the figures given by Davy. In
doing this he had the advantage of being able to use the
very inflammable metal " potassium " which Davy had pre-
pared in 1807 by the action of an electric current on potash.
Gay-Lussac found that :

" On burning the new combustible substance from potash
in 100 parts by volume of nitrous gas, there remained over
exactly 50 parts of nitrogen, the weight of which, deducted
from that of the nitrous gas (determined with great care by
M. Berard at Arcueil), yields as result that this gas is com-
posed of equal parts by volume of nitrogen and oxygen "
(A.C.R. IV. 14).

Composition of nitrous fumes (nitrogen peroxide).—

The composition of the brown fumes of nitrogen peroxide
could not be determined by the methods employed in the
cases of nitrous oxide and nitric oxide, since it was
impossible to collect the gas over water or over mercury.
By admitting nitric oxide, oxygen, and water successively
into an exhausted globe and weighing at each stage, Davy
was able to calculate that the 100 parts of the gas contained
70 parts by weight of oxygen, and 30 parts by weight of
nitrogen ( Works, III. 15). But the experiment was
rendered complex by leakage of air into the globe
through the stopcocks, and by other sources of error for
which corrections had to be applied ; it will therefore be
convenient to describe in its place the more exact experiments
on the admixture of nitric oxide and oxygen which were
made by Gay-Lussac in 1816. Gay-Lussac measured the
contraction which took place when oxygen, confined over
mercury in a graduated tube, was admitted into a bulb con-

o 2

196 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

taining a known volume of nitric oxide, and conversely.
He describes his experiments as follows :

" The apparatus consists of a small bulb united to a
graduated tube by means of a capillary tube 6 centimetres
in length, divided into two equal parts by a tap . . . . "

" The manner of procedure is as follows : Knowing that
the little bulb holds 170 parts of the graduated tube, I
introduced into it, after having made a vacuum, 160 parts of
nitrous gas ; I put then into the tube a definite volume of
oxygen, for example, 200 parts, and having opened the tap,
I find that 158 parts have entered the bulb. This has then
received 160 + 158 = 318 of gas: but its capacity being
only 170, there must have disappeared a volume equal to
318-170 or 148. Working thus and taking the mean of
several results, I found that for 100 parts of nitrous gas the
contraction is 94.

" One cannot deduce from this experiment the quantity of
oxygen that is combined with the nitrous gas ; but if one
makes another experiment in which the oxygen disappears,
because the nitrous gas is in excess, and if one observes the
contraction, one will know the relationship in which the
oxygen and the nitrous gas are combined in each experiment,
assuming that the same product is formed, and consequently
that the contractions are the same in each case. For
example, I found that using 100 parts of oxygen and an
excess of nitrous gas, the contraction was 192. But since
in che first experiment a contraction of 94 corresponded
with 100 parts of nitrous gas, in the second experiment the
contraction of 192 should correspond with 204 of the same
gas. According to this, 100 parts of oxygen are combined
with 204 of nitrous gas, or, in round numbers, 200 ; and I
fix the contraction at 200 instead of 192, that is to say, I
make it equal to the volume of the nitrous gas or double
that of the oxygen." (" On the Combinations of Azote
with Oxygen," Ann. de Chimie, 1816, 1, 401 — 402).

Gay-Lussac's figures showed that :
100 parts of oxygen combining with
204 parts of nitrous gas contract to the extent of
192 parts and give rise to 304 - 192 =
112 parts of the brown nitrous fumes.

x NITRE, NITRIC ACID, AND NITROGEN 197

Gay-Lussac believed that these should be taken as round
numbers and concluded that 100 parts of oxygen combine
with 2w parts of nitrous gas to form 100 parts of the brown
fumes.

It is now known that the relative volume of the product
varies very greatly with the temperature ; at temperatures
above 150° C. 200 volumes are produced, but at atmospheric
temperatures the quantity may be only a little above 100
volumes. The gas resulting from the admixture is now
called NITROGEN PEROXIDE. Since it is formed by the com-
bination of one volume of oxygen with two volumes of nitric
oxide (containing i volume oxygen and i volume nitrogen)
it follows that nitrogen peroxide is a compound of nitrogen
with oxygen in the ratio of one volume of nitrogen to two
volumes of oxygen.

C. NITRIC AND NITROUS ACIDS.

Two varieties of nitric acid.— Although Cavendish had
discovered in 1784 the nature of the characteristic con-
stituents of nitric acid, it was not until 1816 that the
quantitative composition of the acid was finally determined.
This delay was due, mainly, to the fact that the name was
applied to acids which varied widely in composition and
properties.

During the earlier alchemistic period nitric acid was pre-
pared by throwing a mixture of green vitriol and nitre into
a retort or on a charcoal fire and condensing the brown
fumes in a receiver containing cold water. In this way a
red, fuming acid was produced.

On the other hand, the acid which Glauber prepared by
distilling saltpetre with oil of vitriol in glass vessels at a low
temperature was almost colourless. But it was not possible
to distinguish the two varieties sharply from one another,
and the whole range of acids, from the least to the most
highly coloured, were described under a single name as

198 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

"the nitrous acid." The name NITRIC ACID, now used to
distinguish the colourless acid prepared by Glauber's
method, was not introduced until 1787.

Nitric and nitrous acids. — It was recognised both by
Priestley and by Scheele that pure nitric acid is colourless,
but becomes red when reduced or deoxidised by contact
with inflammable substances. Thus Scheele in his Treatise
on air and fire (1777) asserts that :

" The colours of the acid of nitre are accidental. When
a few ounces of fuming acid of nitre are distilled by a very
gentle heat, the yellow separates itself from it and goes into
the receiver, and the residuum in the retort becomes white
and colourless like water. This acid has all the chief
properties of acid of nitre, except that the yellow colour is
wanting. This I call the pure acid of nitre ; as soon,
however, as it comes into contact with an inflammable
substance, it becomes more or less red. This red acid
is more volatile than the pure, hence heat alone can separate
them from one another ; and, for exactly the same reason
the volatile spirit must go over first in the distillation of
Glauber's spirit of nitre. When this has gone over, the
colourless acid follows" (A.C.R. VIII. 20-21).

The gradual reduction of the acid is described (A.C.R.
VIII. 21) as producing:

(1) Red nitrous vapours (nitrogen peroxide).

(2) A weak, volatile PHLOGISTICATED ACID OF NITRE
(nitrous acid).

(3) Colourless nitrous air (nitric oxide).

(4) Complete destruction of the acid (production of
nitrogen).

Scheele examined in detail the properties of the phlog-
isticated acid of nitre, prepared by the action of inflam-
mable substances upon the colourless acid. He made
the important discovery that a salt of the phlogisticated
acid was produced by heating common nitre or saltpetre
until it ceased to "boil." In addition to collecting the gas

x NITRE, NITRIC ACID, AND NITROGEN 199

(oxygen) set free during this action, Scheele noted
that :

" Nitre maintained in red-hot fusion in a glass retort for
half an hour, becomes moist in the open air, and deliquesces
after cooling, and still does not show any trace of alkali."

"This liquefied nitre permits its volatile acid to escape
immediately, when rubbed or mixed with the vegetable
acids"(A.C.R. VIII. 28-29).

A further distinction was discovered by Cavendish, who
found that nitre after heating in an earthenware retort
acquired the property of forming a precipitate when
mixed with nitrate of silver (A.C.R. III. 46), the silver salt
of the " phlogisticated " acid being almost insoluble in water
(see above, p. 191).

The name NITRIC ACID was given to the acid of common
nitre by the French chemists in 1787. The traditional
name of NITROUS ACID was transferred by them to the
"phlogisticated" or deoxidised acid of Scheele's ignited
salt.1 Thus they write :

" We have not hesitated to make the authority of the rule
prevail over that of custom, by naming for example, nitric
add that in which the azote is impregnated with all the
oxygen it is capable of containing,, and reserving the
appellation of nitrous acid for that much weaker acid where
the same base is mixed to a much less quantity of oxygen "
(Chemical Nomenclature, tr. 1788, p. 35).

The salts of nitric acid were called NITRATES, whilst those
derived from nitrous acid were called NITRITES.

Composition of nitric and nitrous acids. — The first
attempts to determine the composition of nitric acid were
made by Lavoisier in 1776 (Mem. Acad. Sci., 1776, 671;
Works, II. 129) eight years before Cavendish had discovered

1 Saltpetre which has been heated with metallic lead consists almost
entirely of potassium nitrite, the potassium salt of nitrous acid, but the
red fuming " acid of nitre " contains both nitric and nitrous acids.

2oa HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

in atmospheric " nitrogen " the principle from which the
acid is derived. Lavoisier's experiments were therefore
restricted to measuring the proportions in which nitrous air
(nitric oxide) and oxygen are united in the acid.

Lavoisier's analysis of nitric acid. — Lavoisier first
attempted to analyse nitric acid by acting upon it with
mercury in such a way that the mercury was recovered
unchanged, whilst the acid was resolved into a series of
gaseous constituents. For this purpose two ounces of acid
of density i '316 were heated with two ounces of mercury and
the gaseous products were collected as they were set free,

(1) during the dissolution of the mercury in the acid,

(2) during the decomposition of the nitrate to oxide
(red precipitate),

(3) during the decomposition of the oxide into mercury
and oxygen.

At the end of the experiments 226 cubic inches of nitrous
air and 238 cubic inches of oxygen had been collected, in
addition to a considerable quantity of water, whilst the
mercury was recovered intact. The experiment failed to
give correct results because nitrous fumes, produced at
various stages of the action, were absorbed and lost in the
water over which the gases were collected ; but the method
was so ingenious as to be well worthy of notice.

Lavoisier's synthesis of nitric acid. — Lavoisier also
attempted in 1776 to synthesise nitric acid by combining
together known quantities of nitrous air and oxygen in
presence of water. He found that 7 J measures of nitrous air,

mixed with 4 measures of oxygen over water,

-

" were reduced to about J of a measure, that is to say to ^¥ of
their original volume."

" The water in the tube was acid at the end of this
operation, or rather was nothing but a weak [nitric] acid ;
on saturating it with an alkali, I obtained from it by evapora-
tion a true saltpetre " ( Works, II. 135).

x NITRE, NITRIC ACID, AND NITROGEN 201

The result of this experiment, in which nitrous air
combined with only half its volume of oxygen, did not
agree with those obtained by the analytical method, in
which the volumes were approximately equal. The
discrepancy led Lavoisier to suggest that the acid prepared
from nitre, which he had used for analysis, differed essen-
tially from the acid synthesised from nitrous air, and that
it was considerably richer in oxygen.

Later investigators found that the proportions in which
nitric oxide and oxygen combined were not definite, but
varied widely with the conditions under which the experi-
ment was carried out ; the problem which Lavoisier had
attacked was indeed of altogether exceptional difficulty, and
forty years elapsed before the correct solution was given by
one of the brilliant group of Frenchmen who, in the early
years of the nineteenth century, resumed at Arcueil, near
Paris, the work which was so tragically interrupted by the
untimely death of Lavoisier.

Combination of nitric oxide with oxygen to form nitric
and nitrous acids. — When nitric oxide is added in suitable
proportions to common air or to oxygen confined over
water, the whole of the oxygen is removed and the water
becomes inpregnated with nitric or nitrous acid. The
method is fairly trustworthy as a means of absorbing or
estimating oxygen, but great variations were found in the
quantity of nitric oxide required to remove a given volume
of oxygen, and consequently also in the proportions of
nitrous air and oxygen in the resulting acid. Thus :

Lavoisier found : oxygen : nitric oxide = TOO : 187
Dalton found : oxygen : nitric oxide— 100 : 130 to 360
Davy found : oxygen : nitric oxide = 100 : 133 to 300

Such variations were not difficult to understand in view
of Cavendish's observation that when water impregnated
with the mixed gases was distilled in a glass retort :

" The first runnings were very acid, and smelt pungent,
being nitrous acid much phlogisticated ; what came next
had no sensible taste or smell ; but the last runnings were
very acid, and consisted of nitrous acid not phlogisticated "
(A.C.R. III. n).

Similar mixtures of nitric and nitrous acids were invariably
obtained, until Gay-Lussac in 1816 succeeded in obtaining
first one acid and then the other as the sole product of the
admixture of the gases, and gave the correct limiting
proportions as follows :

Oxygen : nitric oxide •= 100 : 133. Product, nitric add.
Oxygen : nitric oxide =* 100 : 400. Product, nitrous acid.

Gay-Lussac in his paper "On the Combinations of Azote
with Oxygen " writes :

" I come now to the combinations of nitrous gas with
oxygen : they seem to vary with the slightest change in
the conditions-; but I shall prove that there are three
distinct combinations which, by their mixing, can explain all
those which are not in definite proportions."

" When nitrous gas and oxygen are mixed, the absorption
varies according to the diameter of the tube, the rapidity of
mixing, and the order in which the gases are introduced
into the tube. Wishing to work over mercury, with water
to absorb the acid, but fearing that the mercury would be
attacked, I added potash to the water, and thus obtained
constant absorptions, independent of the conditions I have
just mentioned. I made a great many experiments, and
concluded that 100 parts of oxygen absorb 400 of nitrous
gas : provided that the solutions of potash are strong, the
absorptions are almost all comprised between 495 and 500,
and are rarely below 490. This combination of oxygen
and nitrous gas, which had not been distinguished, so far as
I know, and which I will call provisionally pernitrous add,1
cannot be obtained free ; as soon as one saturates the
potash with an acid, nitrous gas is liberated, and ordinary

]Now called nitrous acid.

x NITRE, NITRIC ACID, AND NITROGEN 203

nitrous acid l is produced, which remains dissolved in the
water. Reducing the nitrous gas to its elements, one finds
that the proportion of 100 of oxygen to 400 of nitrous gas
reduces to :

Nitrogen • 100,

Oxygen 150 "

(Annales de C/iimie, 1816, 1, 399 — 400).

" It remains now to consider nitric acid, and to determine
by what proportion of oxygen and nitrous gas it should be
represented ... I worked at first, like Dalton, in tubes of
5 millimetres diameter, using an excess of oxygen, and
obtained almost exactly the results that he gives, that is to
say, an absorption of 134 to 136 parts of nitrous gas for 100 of
oxygen ; but I found also that tubes of twice this diameter
can be used, provided that one does not shake the water,
and that one waits some minutes. The absorption obtained
by Dalton being only 130, I will adopt the figure 133,
which is also adopted by Davy, and it only remains to
determine to what acid it belongs. The red sulphate of
manganese,2 which I have already recommended as a re-
agent to test when a body is saturated with oxygen, will
fulfil our purpose admirably ; for it is at once decolorised
by nitrous acid, but not at all by nitric acid."

"I began by producing an absorption of 180 parts of
nitrous gas and 100 of oxygen, and found that the acid
formed .... decolorised the red sulphate of manganese
immediately. This salt was also decolorised when the
absorption of nitrous gas was 160, 150, and even 138 ; but
it no longer took place in the experiment in which only 134
parts of nitrous gas were absorbed. It is thus proved that
[the product] is ordinary nitric acid, and that it is formed
by the proportion of 100 of oxygen gas to 133 of nitrous gas,
which reduces to

Nitrogen 100,

Oxygen 250"

(Annaks de Chimie, 1816, 1, 403 — 404).

1 Yellow nitric acid saturated with nitric oxide.

2 A red oxidising agent, probably a permanganate.

204 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

Composition of the oxides and acids of nitrogen.— Gay-

Lussac had now determined the composition of the two
acids derived from saltpetre as well as of the three gaseous
oxides which could be prepared from it. He summarised
his results in a table which, on substituting modern names
for the different compounds, reads as follows :

Nitrogen. Oxygen.
Nitrous oxide ... ... ... 100 50

Nitric oxide ... ... ... 100 100

Nitrous acid ... ... ... 100 150

Nitrogen peroxide... ... ... 100 200

Nitric acid ... ... ... 100 250

This table, the sequence of which was anticipated in a
remarkable way by Scheele in 1777, affords an exception-
ally good illustration of the laws of chemical combination,
the proportions of oxygen in the five compounds being in
the ratios 1:2:3:4:5.

SUMMARY AND SUPPLEMENT
A. NITROGEN.

Cavendish, in 1781, found nitric acid in the water obtained
by exploding hydrogen with an excess of oxygen and concluded
that it had been produced by oxidation of atmospheric azote.
In 1784, by sparking over potash a mixture of three parts
of air with five parts of oxygen, he succeeded in absorbing
almost the whole of the gas and obtained 1*4 grains of nitre.
By sparking with an excess of oxygen, which was afterwards
absorbed by liver of sulphur, he oxidised the whole of the
azote except about 1^ ; the azote therefore consists almost
entirely of a gas from which nitre can be prepared and to
which Chaptal in 1790 gave the name nitrogen ; the residue was
shown by Rayleigh and Ramsay in 1895 to contain an inactive
gas which they called argon.

X NITRE, NITRIC ACID, AND NITROGEN 205

The oxidation of nitrogen in presence of potash may be
represented by the equations :

2N2 + 5O2 + 4KOH -> 4KNO3 + 2H2O

(Potassium
nitrate.)

2N2 + 3O2 + 4KOH -» 4KNO2 + 2H2O

(Potassium
nitrite.)

The fact that a part of the nitre produced in this way was
"phlogisticated," i.e. in the "reduced" form as nitrite, was
noticed by Cavendish.

B. OXIDES OF NITROGEN.

Davy, in 1800, determined the composition by weight cf
the three gases which Priestley prepared from nitric acid,
as follows :

Oxygen : Nitrogen.

"Diminished nitrous air (Nitrous

oxide, N2O) ......... 37 : 63

"Nitrous air" (Nitric oxide, NO) 56*5 : 43-5

Brown Fumes (Nitrogen peroxide,

NO2 or N2O4) 70 : 30

In the case of the two colourless oxides, he burned charcoal
in the gas and measured the quantities of nitrogen and
of carbonic anhydride which were produced. These are as
follows :

Nitrous oxide 2N2O-f C— >2N2 + CO2 (containing O2, \vol.}
i vol. i vol. \ vol.

Nitric oxide 2NO -1- C-»N2 + CO2 (containing O2, J voL\
i vol. \ vol. \ vol.

whence we have as the composition by volume :

Nitrous oxide. Nitrogen : oxygen=-\ : J
Nitric oxide. Nitrogen : oxygen = \ : \.

The composition of the brown gas was determined, but not
very accurately, by mixing known quantities of nitric oxide and
oxygen.

206 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

Gay-Lussac, in 1807, determined more accurately the compo-
sition of nitric oxide (which does not readily part with its oxygen)
by burning metallic potassium in it :

i vol. \ vol.

In 1816, he determined the composition of nitrogen peroxide
by measuring the contraction produced on adding oxygen to a
known volume of nitric oxide, and conversely. He found that
the same contraction (192 volumes) was produced by adding
nitric oxide to 100 volumes of oxygen, or oxygen to 204 volumes
of nitric oxide. Oxygen and nitric oxide therefore combined in
the ratio of 100 volumes of oxygen to 204 volumes of nitric oxide,
or very nearly one volume of oxygen to two of nitric oxide. As
two volumes of nitric oxide contained one volume of nitrogen
and one volume of oxygen, the composition of the product was
as follows :

Nitrogen peroxide. Nitric oxide : oxygen = 2 : i
Nitrogen : oxygen=\ : 2.

The combination of the two gases may be represented by the
equation :

2NO + O2 -> N2O4 or 2NO2.

2 VOls. I VOl. I VOL 2 VOls.

At low temperatures the product consists mainly of the
colourless oxide N2O4, and the volume is reduced, as Gay-
Lussac found, almost to one-third. Above 150° it consists
entirely of the brown oxide NO2, and occupies two-thirds of the
original volume.

R. W. Gray, in 1905 (Trans. Chem. Soc., 1905, 87, 1601-1620),
heated nickel in nitric oxide, and by using modern methods
was able to weigh both the nitrogen and the oxygen present in
a known weight of nitric oxide. The flask A (Fig. 40)
was provided with a glass stopper, B, leading through
a stopcock to a capillary ground-glass joint, D. A small
platinum boat, H, containing finely-divided nickel was sur-
rounded by a coil of platinum wire, which could be heated to
any desired temperature by an electric current, supplied through
the stout platinum electrodes, EE. The bulb M, containing

NITRE, NITRIC ACID, AND NITROGEN

207

charcoal, and provided with a stopcock, could be connected with
the bulb A, through the capillary ground-glass joints at K and D.
Air could be pumped out or nitric oxide admitted through the
stopcock P.

The bulb A was weighed (i) empty, and (2) filled with nitric
oxide. After burning the nickel in the boat H, the nitrogen
left in the bulb A was transferred to the bulb M by opening the

FIG. 40.— GRAY'S APPARATUS FOR DETERMINING THE COMPOSITION OF
NITRIC OXIDE.

two stopcocks (after pumping out the air between K and D) and
then cooling the charcoal in M by liquid air. The whole of the
nitrogen was thus condensed on the charcoal and a vacuum was
produced in the rest of the apparatus. A third weighing of the
bulb A gave the weight of oxygen which had been taken by the
nickel from the nitric oxide ; the increase in the weight of
the bulb M gave the weight of the nitrogen that had been set
free, thus :

2NO + 2Ni -> N2 + 2NiO.
In a typical experiment,
0.62I03 gran, nitric oxide gave {^ g- oxygen^

whence I4'oi3 grams of nitrogen combine with 16 grams of
oxygen to form 30*013 grams nitric oxide.

C. NITRIC AND NITROUS ACIDS.

The existence of two " acids of nitre," one strong and the
other weak, was recognised by Priestley, Scheele, Cavendish,

208 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

and Lavoisier. The strong acid from which saltpetre is
derived was distinguished in 1787 as nitric acid and its salts
as nitrates ; the weak acid, present in ignited saltpetre, was
distinguished as nitroics acid and its salts as nitrites ; the name
nitrous acid was, however, also applied to nitrogen peroxide
and to the red fuming variety of nitric acid. Typical formulae
are as follows :

/Nitric acid, HNO3) f Saltpetre (potassium nitrate), KNO3\
\Nitrous acid, HNO2/\ Ignited „( „ nitrite), KNO2J

/Silver nitrate, AgNO3 (soluble in water) |
\Silver nitrite, AgNO2 (sparingly soluble)]

Lavoisier, in 1776, attempted to analyse strong nitric acid
with the help of mercury, as in the case of sulphuric acid (p. 168).
He also synthesised a weak acid by mixing nitric oxide and
oxygen over water. Cavendish, in 1781, showed that the latter
process gives both the ordinary (nitric) and the phlogisticated
(nitrous) acid.

Gay-Lussac, in 1816, finally determined the composition of
the two acids by mixing nitric oxide and oxygen : (a) over potash,
producing a pure nitrite ; (d) in a narrow tube over water, pro-
ducing pure nitric acid.

The action is shown by the following equations :

(a)

t,vols, i vol. (Potassium

nitrite.)

4 vols. 3 vols. (Nitric acid.)

Since nitric oxide contains half its volume of nitrogen and
half its volume of oxygen, the four volumes of nitric oxide
shown in each of the above equations contain two volumes of
nitrogen and two volumes of oxygen ; the anhydrides of the
two acids therefore have the following composition :

Nitrous acid Nitric oxide : oxygen = $ i

Nitrogen : oxygen = 2 3

Nitric acid Nitric oxide : oxygen — ^ 3

Nitrogen : oxygen = 2 5.

NITRE, NITRIC ACID, AND NITROGEN

209

The composition by volume of the whole series of oxides
investigated by Gay-Lussac is as follows :

Nitrous oxide
Nitric oxide
Nitrous anhydride
Nitrogen peroxide
Nitric anhydride

Nitrogen

Oxygen.

N2O 100

50

NO 100

100

N4O6 or N2O3 100

150

N2O4 or NO2 100

200

N2O5 100

250

CHAPTER XI

MURIATIC ACID AND CHLORINE

A. THE DISCOVERY OF CHLORINE

Scheele's investigation of pyrolusite (1771-1774). —
During the years from 1771 to 1774 Scheele was occupied,
at the request of Bergman, in the investigation of a Swedish
mineral to which the name of " Brunsten " or " Manganese" l
was given. This mineral is now generally known as
PYROLUSITE, whilst the name MANGANESE has been trans-
ferred to the metal which it contains. This mineral was a
black powder which resembled the metallic calces in appear-
ance, but differed from them in being insoluble in dilute
oil of vitriol and in nitric acid ; it dissolved, however, in
presence of various inflammable substances, such as sugar
and gum, and had the curious quality of dissolving in the
weak acids (sulphurous and nitrous acids) which are pro-
duced by reducing or " phlogisticating " oil of vitriol and
nitric acid. The mineral could also be rendered soluble by
baking it with oil of vitriol, when a gas was given off identi-
cal in its properties with the "fire air," or oxygen, which
Scheele had prepared by heating nitre.

Pyrolusite was thus shown to have the properties of a calx
combined with oxygen. Such compounds of a calx with

1 "On Manganese, Manganesium, or Magnesia Vitrariorum,"
Scheele's Essays, tr. 1786, repr. 1901, pp. 52-104.

xi MURIATIC ACID AND CHLORINE an

oxygen are now generally described as PEROXIDES ; thus
pyrolusite is sometimes called PEROXIDE OF MANGANESE, but
is more often distinguished by its colour as BLACK OXIDE OF
MANGANESE. The composition of pyrolusite as a compound
of a calx or earth with oxygen is very similar to that of chalk
as a compound of lime with fixed air. But whilst the fixed
air contained in chalk is set free even by the weakest acids,
the oxygen of pyrolusite is only driven off by heating
strongly with oil of vitriol. Another method of separating
oxygen from pyrolusite was described in 1785 by Berthollet
(A.C.R. XIII. 1 6), who heated the mineral (compare the
burning of chalk to lime) and thus extracted from it " a large
quantity of vital air " ; in this process the pyrolusite " lost an
eighth of its weight."

Scheele's discovery of chlorine (1774). — The behaviour of
" manganese " towards spirit of salt differed from its behav-
iour with other acids, in that the mineral dissolved in the
cold acid without the addition of any reducing agent to
take away the oxygen. In the cold solution, as Scheele
pointed out, "the manganese has here attached itself so
loosely to the acid that water can precipitate it, and this
precipitate behaves like ordinary manganese." But when
the mixture was warmed, a pungent gas was liberated, and a
colourless solution was obtained, similar to those produced
by other acids with the help of a reducing agent. Scheele
describes his discovery of the new gas as follows :

" One ounce of pure spirit of salt was poured on half an
ounce of finely-ground manganese. After this mixture had
stood one hour in the cold, the acid had assumed a dark
brown colour. Part of this solution was poured into a
bottle, which was left open in a warm place. The solution
gave off a smell like warm aqua regis, and after a quarter of
an hour it was clear and colourless as water, and the smell
was gone."

p 2

212 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

"The remainder of the brown mixture was set to digest,
in order to see whether the marine 1 acid would saturate
itself with manganese. As soon as the mixture became warm,
its smell of aqua regis became considerably augmented, and
an effervescence also arose, which continued until the
following day, when the acid was found to be saturated. On
the residue which it had been unable to dissolve, there was
again poured one ounce of spirit of salt, whereupon all the
above-mentioned phenomena occurred, and the manganese
became completely dissolved, except a little siliceous
earth."

" In order clearly to apprehend this novelty I took a
retort containing a mixture of manganese and spirit of salt.
In front of the neck I bound a bladder emptied of air, and
set the retort in hot sand. The bladder became distended
by the effervescence in the retort. When the acid no longer
effervesced, which was an indication of its saturation, I
removed the bladder, and found that this air had coloured
it yellow, as if by aqua fortis, but did not contain any trace
of fixed air ; it had, however, a quite characteristically
suffocating smell, which was most oppressive to the lungs.
It resembled the smell of warm aqua regis. The solution
in the retort was clear, inclining to yellow, which last-men-
tioned colour was caused by its containing iron" (A.C.R.
XIII. 5-7).

The gas which Scheele had thus prepared, he regarded as
"marine acid, deprived of phlogiston," and called it " de-
phlogisticated marine acid," but when the theory of phlogiston
was overthrown, Davy gave to it the name CHLORINE (Greek
X^wpos greenish-yellow) on account of its characteristic colour.
Although this name was not introduced until 1810, it is used,
on account of its simplicity and convenience, throughout
the chapter in place of the cumbrous and misleading names
given to it during the intervening period.

The properties of chlorine. — In addition to describing
its suffocating smell and yellow colour, Scheele states that

1 Marine acid — muriatic acid, the acid derived from brine.

xi MURIATIC ACID AND CHLORINE 213

the gas " unites with water in very small quantity ; and gives
to water a slightly acid taste ; but as soon as it comes in
contact with a combustible matter it becomes again a proper
marine acid." As usually prepared it contained a certain
quantity of acid vapour, but this could be got rid of by
collecting it in wet bottles ; the acid was then dissolved by
the water, whilst the greater part of the chlorine remained
as a gas, into which various substances could be plunged.
The following properties were noticed :

" The corks in the bottles became yellow, as from aqua
fortis."

" Blue litmus paper became almost white ; all vegetable
flowers — red, blue, and yellow — became white in a short time;
the same thing also occurred with green plants. In the
meantime the water in the bottle became changed to a weak
and pure marine acid. The former colours of these flowers,
as well as those of the green plants, could not be restored
either by alkalis or by acids."

" Iron filings were put into the same bottle and they
dissolved. This solution was evaporated to dryness and
distilled with an addition of oil of vitriol, when a pure
marine acid, which did not dissolve gold, again passed
over."

" All metals were attacked, and with gold it is noteworthy
that its solution in this dephlogisticated marine acid forms
with volatile alkali a fulminating gold'." 1

" Insects immediately died in the vapours ; and fire was
immediately extinguished by them " (A.C.R. XIII. 8-9).

Berthollet (1785) regards chlorine as an oxygen com-
pound.— Chlorine was next investigated by the French
chemist Berthollet (Mem. Acad. Sa., 1785, 276-295 ; A.C.R.
XIII. 11-31), who made a special study of the properties
of CHLORINE-WATER. This was prepared by passing the
gas through a series of bottles ; the first bottle, empty and
cold, served to condense and remove the acid vapours ;

1 See Chapter XII.

2i4 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

three others almost filled with water were used to dissolve
the gas. By surrounding the bottles with ice, Berthollet
obtained yellow crystals which he considered to be the
solidified gas (A.C.R. XIII. 13-14) ; these have since been
shown to be a compound of the gas with water.

Berthollet found that solutions of chlorine in water were
decomposed by exposure to light, oxygen being liberated
and muriatic acid reproduced. Scheele had regarded
chlorine as muriatic acid which had " lost one of its
constituents," namely, phlogiston ; he would therefore have
interpreted this experiment as a decomposition in which
water was robbed of its phlogiston by the chlorine, whilst
the oxygen which it contained was set free. This view may
be expressed by the equation :

Chlorine + water -> muriatic acid + oxygen.

(Hydrogen +oxygen.) (Chlorine-)- hydrogen.)

Berthollet preferred, however, to regard the change as a
simple decomposition of chlorine into muriatic acid and
oxygen :

Chlorine -> muriatic acid + oxygen

(Muriatic acid+oxygen.)

The fundamental weakness of Berthollet's theory lay in
the fact that chlorine by itself could not be decomposed into
muriatic acid and oxygen even by the most drastic treat-
ment. This was in direct contradiction to his own view
that in chlorine

" the vital air .... adheres so feebly to the marine acid
that the action of light suffices to disengage it promptly"
(A.C.R. XIII. 20). ^

Berthollet shows that chlorine is not an acid.— Another
difficulty in regarding chlorine as a compound of muriatic
acid with oxygen arose when Berthollet himself discovered
that chlorine was not an acid. He points out that " Scheele

xi MURIATIC ACID AND CHLORINE 215

and Bergman were unable to recognise this essential quality,
as in the process which they employed the water of the
vessels in which they received the gas " was contaminated
with " a portion of marine acid which always passed over in
the distillation." In Berthollet's apparatus this acid was
" retained in the first bottle," which was left empty and
surrounded with ice or cold water; if by chance any acid
passed forward it was absorbed in the first of the three bottles
of chlorine-water (A.C.R. XIII. 15).

Chlorine-water purified in this way was found to have the
following properties :

" It has a harsh taste which does not resemble that of the
acids."

" It destroyed vegetable colours .... without any red
tint becoming apparent."

" It does not cause an effervescence with solution of fixed
alkali, even when the latter is saturated with fixed air"
(A.C.R. XIII. 14).

The fact that chlorine was not an acid was in direct
contradiction to Lavoisier's conception of oxygen as the
acid-producer. Lavoisier had concluded that muriatic acid,
like carbonic, nitric, sulphuric, and phosphoric acids, must
be a compound of oxygen ; if chlorine contained more
oxygen than muriatic acid it should have been a stronger
acid ; actually it was not an acid at all.

This difficulty was realised by the French chemists who
drew up the new system of chemical nomenclature in 1787 ;
in proposing to describe chlorine as an " oxygenated muri-
atic acid " they say of muriatic acid that " it is an acid of a
particular nature, because it imbibes an excess of oxygen,
and because in this state its acidity seems rather to decrease
than to augment " (Chemical Nomenclature ', p. 33).

Gay-Lussac and Thenard (1809) prove that muriatic acid
gas contains hydrogen. — That muriatic acid gas (i) con-

216 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

tains hydrogen and (2) is the sole product of the combina-
tion of hydrogen with chlorine was proved by the French
chemists Gay-Lussac and Thenard in 1809. The presence of
hydrogen was shown by the fact that this gas was liberated
freely when the muriatic gas was acted upon by metals.1
Thus they write :

" We have also examined the action of the metal of potash
on muriatic gas. At the ordinary temperature this action
is very slow ; but as soon as the metal is fused there is com-
bustion with disengagement of light, and there result muriate
of potash and hydrogen gas."

" The quantity of hydrogen collected in this experiment is
precisely the same as that which the metal would give in
contact with water."

"We passed a current of muriatic gas over well-cleaned
iron turnings at a dull red heat. Much hydrogen gas was
disengaged, without sensible admixture of muriatic gas ;
much muriate of iron was at the same time obtained ; the
residual turnings were not oxidised" (A.C.R. XIII. 34 and

38).

Hydrogen, combined with oxygen in the form of water, was
also collected " by passing muriatic acid gas at a moderate
heat over litharge which had been fused and then reduced
to coarse powder."

Muriatic acid gas a compound of hydrogen and
chlorine. — Gay-Lussac and Thenard observed further that
equal volumes of hydrogen and chlorine combined together
to form muriatic gas (a) by keeping the mixture for several
days, (b) by gently heating it, (c) by exposure to light.
They found

" that a mixture, in equal parts, of [chlorine] gas and hydro-
gen gas changes, in the course of several days, into ordinary
muriatic gas, and that no water is deposited."

" If a mixture of equal parts of these two gases is made
and a small piece of iron heated in mercury to 150° intro-

1 These were now generally recognised as elementary substances.

xi MURIATIC ACID AND CHLORINE 217

duced, there is a violent inflammation, and formation of
muriatic acid."

" We made two mixtures, each consisting of about J litre
of [chlorine] gas with the same volume of hydrogen gas,
which we knew acted only slowly on one another ; one of
them we placed in complete darkness, and exposed the
other to the light of the sun, which was that day very
feeble. At the end of several days the first mixture was
still coloured green, and appeared to have undergone no
change ; the second, on the. contrary, had been completely
decolorised in less than a quarter of an hour, and was almost
entirely decomposed."

" Being no longer able, after these experiments, to doubt
as to the influence of light on the combination of the two
gases, and judging from the rapidity with which it had
operated, that if the light had been much more vivid it
would have operated much more quickly, we made new
mixtures .... and placed them in complete darkness,
awaiting some moments of bright light. Two days after
having made the mixtures, we were able to expose them to
the sun. Scarcely had they been exposed when they
suddenly inflamed with a very loud detonation, and the
jars were reduced to splinters, and projected to a great
distance. Fortunately we had provided against such
occurrences, and had taken precautions to secure ourselves
against accidents" (A.C.R. XIII. 38, 41, 43).

These experiments proved conclusively that muriatic gas
was a compound of hydrogen and chlorine, and was, in fact,
the only substance produced by the combination of these
gases. The composition of the gas must therefore be
represented by the equation :

Chlorine + hydrogen -> muriatic gas,
instead of

Chlorine - oxygen -> muriatic gas.

This view is expressed in the modem system of nomen-
clature, in which muriatic acid gas is described as HYDROGEN

218 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

CHLORIDE and its solution in water as HYDROCHLORIC ACID/
the muriates derived from the acid being described as

CHLORIDES.1

B. CHLORINE AN ELEMENT.

Gay-Lussac and Thenard regard chlorine as the oxide
of an imaginary radical. — The French chemists were so
obsessed with the idea that all acids must contain oxygen,
that they refused to abandon this view, even when chlorine
had been proved to be a simpler substance than muriatic
acid. Thus, instead of admitting that chlorine might be an
element, they regarded it as the oxide of an unknown
radical, analogous with nitrogen, carbon, sulphur, or phos-
phorus. Muriatic acid, formed by the union of chlorine
with hydrogen, was regarded as a ternary compound of this
radical with hydrogen and oxygen, or perhaps with water.
According to this new view :

Chlorine = X + oxygen.
Muriatic acid = chlorine + hydrogen

= X -f oxygen + hydrogen.

It will be seen that this theory merely substituted for
chlorine the symbol (X + oxygen), where X was an imaginary
radical which had never been isolated. To prove the
correctness of their theory, it was necessary for Gay-
Lussac and Thenard to extract oxygen from chlorine, or
water from muriatic acid, without making use of substances
in which oxygen was already present. The experiment of
passing muriatic gas over litharge was obviously invalid (as
Davy pointed out in 1810), since the formation of water

1 The name "hydrochloric acid "was proposed by Gay-Lussac in
1814 (Ann. d. Chimie, 1814, 91, 9). The term " chloride "'was intro-
duced by Davy in 1816 {Works, V. 516), as a substitute for the
word " chlorure " suggested by Gay-Lussac.

xi MURIATIC ACID AND CHLORINE 219

could be accounted for by the combination of the hydrogen
of the acid with the oxygen of the litharge, quite independ-
ently of the existence of any oxygen in chlorine.

The difficulty of decomposing chlorine. Chlorine as an
agent for purifying charcoal. — Gay-Lussac and Thenard
made many attempts to separate oxygen from chlorine and
thus to set free its other hypothetical constituent. But all
their efforts were unsuccessful, since neither the metals, nor
phosphorus, nor any of the well-known absorbents of oxygen
were capable of separating oxygen from the gas.

" Finally, as a last method, we tried to decompose
[chlorine] by charcoal ignited at the extreme heat of the
forge. To avoid the presence of the smallest quantity of
water, we made the gas pass slowly through a large glass
tube a metre and a half in length, filled with muriate of
lime. This tube communicated with a porcelain tube in
which the charcoal was exposed to a red heat. The first
portions of the .... gas were completely converted into
ordinary muriatic gas. This effect diminished gradually in
spite of a very great elevation of temperature, and soon the
gas passed without alteration, mixed only, towards the end
of the experiment, with one thirty-third of an inflammable
gas, which we believe to be carbonic oxide gas. This result
clearly showed us that [chlorine] is not decomposed by char-
coal, and that the muriatic gas which we had obtained at
the commencement of the operation was due to the hydro-
gen of the charcoal. ... In fact, on taking ordinary char-
coal without igniting it, muriatic gas was disengaged during
a lengthened period, even at a temperature only slightly
elevated. . . . According as the charcoal lost its hydrogen,
however, the quantity of muriatic acid went on diminishing,
and finally nothing was obtained but [chlorine]" (A.C.R.
XIII. 39-40).

In this passage Gay-Lussac and Thenard disclose incident-
ally a very efficient method of purifying charcoal. The
charcoal which Stas used in determining the composition of
carbonic anhydride (p. 149) was purified by heating it

220 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

in a current of chlorine, and the same method is still in
general use.

A striking contrast was observed on comparing the effect
of heat alone on dry and on moist chlorine. The dry gas
resisted all attempts to decompose it, but the introduction
of moisture was followed by the immediate production of
oxygen and muriatic acid. In this experiment chlorine
11 alone or mixed with the vapour of boiling water " was
introduced " into a porcelain tube exposed to a red heat.
In the first case the gas did not suffer any alteration ; but
as soon as the vapour of water was introduced, oxygen gas
and muriatic acid were obtained. It is not necessary for
this decomposition that the temperature should be very
high, as it still takes place below a red heat." When
hydrogen was used in place of water the formation of
muriatic acid took place even more readily " at a tempera-
ture only a little higher than that of boiling water" (A.C.R.
XIII. 41).

At the conclusion of their experiments Gay-Lussac and
Thenard were obliged to admit that since chlorine " is not
decomposed by charcoal ... it might be supposed . . .
that this gas is a simple body," and agreed that " the pheno-
mena which it presents can be explained well enough on this
hypothesis" (A.C.R. XIII. 48). But they were not
prepared to abandon the view that it contained oxygen and
preferred to regard it still as a compound. It was therefore
left for Davy to adopt and to advocate the view that
chlorine was a simple or elementary substance, and finally
to win general recognition for a conception that gave the
death-blow to Lavoisier's oxygen theory of acids.
P Davy (1810) regards chlorine as an element. — At the
time when Gay-Lussac and Thenard were making their
experiments on chlorine, Davy was engaged in similar
researches on this gas. During the early part of 1810, the
failure of all attempts to decompose chlorine, even by the

xi MURIATIC ACID AND CHLORINE 221

action of charcoal heated to whiteness by a Voltaic battery,
led him :

"to doubt of the existence of oxygen in that substance,
which has been supposed to contain it above all others in a
loose and active state ; and to make a more rigorous invest-
igation than had been hitherto attempted for its detection "
(A.C.R. IX. 23).

He repeated the experiments of Gay-Lussac and Thenard
on the combination of hydrogen and chlorine, and con-
firmed their observation that the product was muriatic acid
gas free from any other substance except perhaps a mere
trace of water.

But he could not discover any evidence of the presence
of oxygen in muriatic acid gas or in chlorine, and at the
close of the year he expressed his belief :

"that the body improperly called in the modern nomen-
clature of chemistry, oxymuriatic add gas has not as yet been
decompounded; but that it is a peculiar substance,
elementary as far as our knowledge extends, and analogous
in many of its properties to oxygen gas."

" To call a body, which is not known to contain oxygen,
and which cannot contain muriatic acid, oxymuriatic acid,
is contrary to the principles of that nomenclature in which
it is adopted."

" After consulting some of the most eminent chemical
philosophers in this country, it has been judged most proper
to suggest a name founded upon one of its obvious and
characteristic properties — its colour-- and to call it CHLORINE
or Chloric gas."

"Should it hereafter be discovered to be a compound,
and even to contain oxygen, this name can imply no error,
and cannot necessarily require a change" (A.C.R. IX.
40, 59)-

Elements and compounds.— The controversy as to the
elementary nature of chlorine was of value in fixing clearly

222 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

the conception of an ELEMENT. The four elements of
Aristotle :

earth, air, fire, and water,
and the three principles of the alchemists :

mercury, sulphur, and salt,

served rather to describe the qualities than the actual com-
position of the bodies in which their presence was implied.
The first conception of an element as a substance which
cannot be decomposed into simpler substances was given
by Boyle. But no clearer statements of this view can be
found in the whole of the literature of chemistry than those
made by Davy in his controversy with the French chemists, and
afterwards with Berzelius. Two of these may be quoted :

" Some authors continue to write and speak with scept-
icism on the subject, and demand stronger evidence of
chlorine being undecompounded. These evidences it is
impossible to give. It has resisted all attempts at decom-
position. In this respect it agrees with gold, and silver, and
hydrogen, and oxygen. Persons may doubt whether these
are elementary bodies ; but it is not philosophical to doubt
whether they have not been resolved into other forms of
matter " (A.C.R. IX. 73).

" It is possible that oxymuriatic gas may be compound,
and that this body and oxygen may contain some common
principle ; but at present we have no more right to say that
oxymuriatic gas contains oxygen than to say that tin contains
hydrogen ; . . . and till a body is decompounded, it should
be considered as simple" (A.C.R. IX. footnote, p. 61).

Davy (1810) determines the composition by volume of
hydrogen chloride. — Until the year 1810 muriatic acid gas
was regarded either as a simple substance (Berthollet) or as a
compound of the unknown muriatic radical with one-third or
one-fourth of its weight of water (Davy ; Gay Lussac ; A.C.R.
XIII. 49). x In that year, however, Davy, having proved it

1 Berzelius (Chemical Proportions, 1819, 125) assumed that 442-65
parts of chlorine contained 142*65 parts of the muriatic radical and 300
parts of oxygen.

xi MURIATIC ACID AND CHLORINE 223

to be a binary compound of two simple gases, hydrogen and
chlorine, succeeded also in establishing its composition by
volume. Davy states that Cruikshank had shown th^
chlorine and hydrogen, when mixed in equal proportions,
gave a product almost entirely condensible by water ; Gay-
Lussac and Thenard showed that the product was muriatic
acid, but did not make any measurements of volume.
These could not be made in the ordinary way either over
water (which dissolved the acid gas) or over mercury (which
was acted on by chlorine). In Davy's experiments the two
gases " were mixed in equal volumes over water, and intro-
duced into an exhausted vessel and fired by the electric
spark." As the gases were moist, "there was always a
deposition of a slight vapour, and a condensation of from
one-tenth to one-twentieth of the volume ; but the gas
remaining was muriatic acid gas." The contraction was
reduced when the gases were dried, " by introducing them
into vessels containing muriate of lime, and by suffering
them to combine at common temperatures." Although
Davy was never " able to avoid a slight condensation," he
found that " in proportion as the gases were free from
oxygen or water, this condensation diminished." It was
therefore reasonable to suppose that the pure gases would
unite in equal volumes without contraction (A.C.R. IX. 26).

Davy also mixed together equal volumes of chlorine and
sulphuretted hydrogen, both dried : " in this instance the
contraction was not ^ ; sulphur .... was formed on the
side of the vessel, but no vapour was deposited. The
residual gas contained about i§ of muriatic acid gas, and
the remainder was inflammable " (A.C.R. IX. 26)

These experiments proved that hydrogen and chlorine
combine together in equal volumes and without any marked
contraction, as shown by the equation :

Hydrogen + chlorine -> hydrogen chloride,
i vol. i vol. 2 vols.

224 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

In the case of sulphuretted hydrogen, which was known
to contain its own volume of hydrogen, the volume
changes were similar :

Sulphuretted + chlorine — > hydrogen chloride + sulphur,
hydrogen
i vol. i vol. 2 vols. solid

Davy showed further that hydrogen chloride liberates half
its volume of hydrogen when the chlorine which it contains
is absorbed by a metal. Thus in :

" the decomposition of muriatic acid gas, by heated tin
and zinc, hydrogen equal to about half its volume was
disengaged, and metallic muriates the same as those pro-
duced by the combustion of tin and zinc in [chlorine]
resulted" (A.C.R. IX. 28).

" With potassium, in experiments made over very dry
mercury, the quantity of hydrogen is always from 9 to 11,
the volume of the muriatic acid gas used being 20 "
(A.C.R. IX. 27).

" When mercury is made to act upon i in volume of
muriatic acid gas, by Voltaic electricity, all the acid
disappears, calomel is formed, and about 0-5 of hydrogen
evolved" (A.C.R. IX. 27).

This decomposition of the gas by sparking in presence of
mercury was discovered by Priestley (Experiments on Air,
1775, II. 239) and studied in detail by Henry in 1800
(Phil. Trans. 1800, 191). It is of special interest as an
example of a reversible change. Usually the action
proceeds as shown by the upper arrow in the equation :

Hydrogen + chlorine ^± hydrogen chloride.

But chlorine acts upon mercury so strongly that a very
slight decomposition of the hydrogen chloride, as shown by
the lower arrow, is sufficient to corrode the mercury and
ultimately to remove the whole of the chlorine from the gas,
leaving a residue of pure hydrogen. The compound formed

xi MURIATIC ACID AND CHLORINE 225

by the action of chlorine on an excess of mercury is an
insoluble white powder known as CALOMEL. It contains
more mercury and less chlorine than the CORROSIVE
SUBLIMATE which is formed when the chlorine is in excess.

Compounds of chlorine with the metals. — Davy's view that
chlorine was an element involved also a new conception as
to the nature of the muriates, or chlorides. Most of the
salts that had been investigated contained a fixed, earthy
BASE, usually a metallic oxide, in combination with a volatile
ACID, usually the oxide of a non-metal such as carbon,
nitrogen, sulphur or phosphorus. These salts were, there-
fore, BINARY COMPOUNDS of the tWO Oxides Or TERNARY

COMPOUNDS of metal, non-metal and oxygen, thus :

= /METAL

CHALK = -( \OXYGEN

~ \CARBON

C /-« £\ p -p Tf TJ

COPPER OXIDE
v SULPHURIC ANHYDRIDE = (SULpHUR

The muriates on the other hand, as Scheele had observed,
were formed by the direct combination of chlorine with the
metals.

In order to maintain the analogy with the sulphates,
phosphates, etc., the French chemists assumed a further
sub-division of the chlorine, thus :

MURIATE

LEAD = LEAD

OF LEAD I CHLORINE = I OXYGEN

\ UNKNOWN MURIATIC RADICAL

Davy showed that this view was not justified by any
experimental evidence. He suggested, as a more reason-
able hypothesis, that chlorine " may possibly belong to the
same class of bodies as oxygen " (A.C.R. IX. 33), and that

Q

226 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

the compounds of chlorine with lead, silver, mercury,
potassium and sodium might " be considered as a class of
bodies related more to the oxides " than to the ordinary
salts (A.C.R. IX. 34)." Gay-Lussac recognised a still
closer analogy between chlorine and sulphur, and between
the chlorides and sulphides. According to these views a
true analogy is supplied by comparing the muriates or
chlorides with binary compounds, such as :

LITHARGE (OXIDE OF LEAD) = {oXYGEN

GALENA (SULPHIDE OF LEAD) = {SULPHUR

and not with ternary compounds, such as :

("LITHARGE = (LEAD

SULPHATE OF LEAD = J VOXYGEN

[SULPHURIC ANHYDRIDE = |SULPHUR

Displacement of sulphur and oxygen by chlorine. — The

analogy between chlorine, sulphur, and oxygen is shown by
the readiness with which these elements can be interchanged
in their compounds with the metals. Thus Scheele, in his
first description of chlorine, stated that by contact with the
gas " cinnabar became white on the surface, and when the
piece was washed in water a pure sublimate solution was
obtained, but the sulphur was not altered" (A.C.R. XIII.
9). This action may be represented by the equation :
Cinnabar + chlorine -> corrosive sublimate + sulphur

(Mercury-|- (Mercury+chlorine)

sulphur)

A similar process is now used on a large scale for con-
verting sulphide ores of zinc, lead, and silver into chlorides.
The conversion of oxides into chlorides was described
by Davy, who thus obtained a direct proof of the presence
of oxygen in the earths. Davy found that :

" When baryta, strontia, or lime, is heated in [chlorine] to

xi MURIATIC ACID AND CHLORINE 227

redness, a body precisely the same as a dry muriate is
formed, and oxygen is expelled from the earth. I have
never been able to effect so complete a decomposition of
these earths by [chlorine], as to ascertain the quantity of
oxygen produced from a given quantity of earth. But in
three experiments made with great care I found that one of
oxygen was evolved for every two in volume of [chlorine]
gas absorbed" (A.C.R. IX. 47).

The displacement of oxygen by chlorine is effected even
more easily by the action of muriatic acid gas or hydrogen
chloride. This gas being a HYDRIDE, i.e. a compound of
hydrogen with another element, interacts with metallic
oxides to form muriates or chlorides and water, thus :

LITHARGE + MURIATIC ACID GAS -> MURIATE OF LEAD + WATER
(Lead + ( Hydrogen + (Lead + chlor- ( Hydrogen +

oxygen) chlorine) ine) oxygen)

Under similar conditions an acidic OXIDE, such as fixed
air or sulphuric anhydride, unites with the metallic oxide to
form a ternary salt and nothing else.

Berthollet (1788) prepares chlorate of potash.—
Although the muriates, like the sulphides, were found to be
binary compounds containing no oxygen, oxidised muriates
(ternary compounds analogous with the sulphates) were
discovered at an early date. The most important of these,
now known as CHLORATE OF POTASH or POTASSIUM CHLORATE,
was first prepared in a pure state by Berthollet in 1788, by
passing chlorine gas into hot aqueous potash. (Journ. de
Physique, 1788, 33, 217-224). After crystallising out a
great deal of muriate of potash (potassium chloride) the new
salt was obtained in hexagonal plates with a lustre like
mica. These did not taste like muriate of potash, but
produced in the mouth " a feeling of freshness closely
resembling that of nitre." The salt detonated with char-
coal ; after the detonation nothing was left but ordinary
muriate of potash • the salt was therefore an " oxygenated

Q 2

228 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

muriate of potash." On heating it, oxygen was liberated
more easily and in much larger quantity than from nitre ;
100 grains of salt gave 75 cubic inches or nearly 37 grains
of oxygen ; the salt could therefore be used very conveni-
ently for the preparation of oxygen and has been employed
for that purpose ever since. On account of the readiness
with which its oxygen was set free and the brilliancy of its
detonation with charcoal, Berthollet predicted that "the
powder which I propose to prepare with this salt, will have
remarkable properties " ; the experiment was carried out
shortly afterwards but produced a disastrous and fatal
explosion. The acid from which the oxidised muriate is
derived was prepared by Gay-Lussac in 1814 (Ann. de
Chimie, 1814, 91, 107-110). He described the acid (ibid,
p. 9) as CHLORIC ACID, and its salts CHLORATES.

Berthollet (1788) prepares bleaching solutions, contain-
ing alkaline hypochlorites. — By passing chlorine into a
cold aqueous solution of a mild or caustic alkali, Berthollet
obtained a liquid which possessed all the bleaching qualities
of chlorine-water, .and decomposed slowly (like chlorine-
water exposed to sunlight) with liberation of oxygen gas.
When the liquid was heated, these properties disappeared
and nothing was left but the chloride and chlorate. The
bleaching properties are due to the presence of a salt
(derived from an unstable HYPOCHLOROUS ACID), which
contains less oxygen than the chlorate and is distinguished
as a HYPOCHLORITE. Hypochlorite solutions were used by
Berthollet as a convenient source of chlorine in the
experiments in which he laid the foundations of the
bleaching industry (Ann. de Chimie, 1789, 2, 151-190).

Tennant (1799) prepares bleaching powder. — A some-
what similar compound, prepared by the direct union of
chlorine with lime, and widely known as BLEACHING
POWDER, was introduced by the English chemist Tennant in
1799 as a substitute for hypochlorite of potash or soda.

xi MURIATIC ACID AND CHLORINE 229

SUMMARY AND SUPPLEMENT
A. THE DISCOVERY OF CHLORINE

Scheele, in 1774, described the properties of the mineral
"manganese," now known as pyrolusite or black oxide of
manganese, MnO2. It is insoluble in dilute nitric or sulphuric
acid, but liberates oxygen and is converted into a sulphate by
baking with oil of vitriol.

MnO2+H2SO4 -

(Manganese (Manganous

dioxide.) sulphate.)

It also liberates oxygen when heated (Berthollet, 1785), giving
rise to an oxide analogous with smithy scale, Fe3O4.

It dissolves in cold muriatic acid to a dark solution, which
contains an unstable manganic salt and from which the black
oxide can be recovered by dilution ; this reversible action may
perhaps be represented by the equation : —

but no chloride of the formula MnCl4 has yet been isolated.
When the liquid is heated, a pungent, green gas escapes, to
which Davy, in 1810, gave the name chlorine.

(Manganous (Chlor-
chloride.) ine)

The gas bleaches coloured flowers, extinguishes flame, and
unites directly with nearly all metals, thus : —

Lead Pb + CL->PbCl2.

(Lead
chloride.)

Iron 2Fe + 3Cl2->2FeCl3.

(Ferric
chloride.)

Gold 2Au + 3Cl2->2AuCl3.

(Auric
chloride) ^)

Berthollet, in 1785, collected chlorine-water free from muri-
atic acid and showed that the gas is not an acid. From the

230 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

solutions he froze out chlorine hydrate, C12 + 8H2O. He found
that chlorine water was decomposed by light, liberating oxygen
and reproducing muriatic acid ; he therefore regarded chlorine
as a compound of oxygen with muriatic acid and called it
"oxymuriatic gas." The action is really a decomposition of
water by chlorine :

the action is reversible, as shown by the double arrows, since
muriatic acid exposed to air and light becomes impregnated
with chlorine.

Gay-Lussac and Thenard, in 1809, proved that muriatic gas
was a compound of hydrogen with chlorine, a fact that is
expressed by the modern name hydrogen chloride, HC1, for the
gas and hydrochloric acid for the solution. They separated the
hydrogen from it by the action of metals, thus :

Potassium 2 K + 2 H Cl-> 2 KC1 + H2

(Potassium
chloride.)

Iron Fe + 2HCl->FeCl2 + H2

(Ferrous
chloride.)

It should be noted that the ferrous chloride, FeCl2, produced in
this way contains less chlorine than the ferric chloride, FeCl3,
produced by the direct action of chlorine upon iron.

Hydrogen, in the form of water, was also separated from
muriatic gas and litharge :

2HC1 + PbO -> PbCl2 + H2O

(Hydrogen (Litharge.) (Lead (Water.)
chloride.) chloride.)

They united hydrogen and chlorine to form hydrogen chloride :
(a) slowly by keeping in diffused daylight ; (b) rapidly by heat-
ing ; (c) with a violent explosion by exposure to bright sunlight.
Although they could not extract oxygen from chlorine even by
red-hot charcoal, they contended that chlorine gas was the oxide
of an unknown radicle and that muriatic gas was a ternary
compound of this radical with oxygen and hydrogen, thus :

Chlorine = X + oxygen

Muriatic gas = X + oxygen + hydrogen.

xi MURIATIC ACID AND CHLORINE 231

B. CHLORINE AN ELEMENT.

Davy, in 1810, being quite unable to separate oxygen from
chlorine, suggested that the gas was an element and gave to it
the name chlorine. He determined the volumetric composition of
hydrogen chloride, which is formed without change of volume
when chlorine acts upon hydrogen or upon sulphuretted
hydrogen.

H2 + C12 -> 2HC1

I -VOl. I Vol.

H2S + C12 -> 2HC1 + S

i vol. i vol. 2 vols. solid.

He showed that half its volume of hydrogen is liberated when
the hydrogen chloride is acted on by metals, thus :

-. Tin Sn + 2HC1 -> SnCl2 + H2.

2 vols. (Stannous i vol.

chloride.)

Zinc Zn + 2HC1 -> ZnCl2 + H2.

2 vols. (Zinc chloride.) i vol.

Potassium 2K + 2HC1 -> 2KC1 + H2.

2 vols. (Potassium i vol.

chloride.)

Mercury 2Hg+ 2HC1 -> Hg2Cl2 + H2.

(By sparking.) 2 vols. (Calomel.) i vol.

It should be noted that the stannous chloride, SnCl2,
prepared in this way contains less chlorine than the stannic
chloride, SnCl4, prepared by the action of chlorine on tin and
that calomel or mercurous chloride, Hg2Cl2, contains a smaller
proportion of chlorine than the corrosive sublimate or mercuric
chloride, HgCl2, prepared by the action of chlorine on mercury
or its sulphide. In the case of zinc and potassium the products
are identical with those obtained by the action of chlorine on
the metal.

All these chlorides are binary compounds of metal and
chlorine analogous with the sulphides and oxides, rather than
with ternary salts such as the sulphates and carbonates. This

232 HISTORICAL INTRODUCTION TO CHEMISTRY CH. xi

analogy is illustrated by the mutual displacement of chlorine,
sulphur and oxygen, for instance :

HgS + C12 -> HgCl2 + S.

(Cinnabar.) (Corrosive

sublimate.)

CaO + C12 -> CaCl2 + |O2.

(Lime.) (Muriate of

lime.)

.

Berthollet, by the action of chlorine on potash, prepared
ternary salts, containing metal, chlorine and oxygen :

2KOH + C12 -> KC1 + KC1O + H2O.

(Potassium
hypochlorite.)

6KOH + 3C12 -> 5KC1 + KC1O3 4- 3H2O.

(Potassium
chlorate.)

The potassium hypochlorite prepared from cold potash liberates
chlorine when acted on by acids, and was used as a convenient
source of chlorine for bleaching ; but it was soon displaced by
the bleaching powder, CaOCl2, which Tennant prepared by the
direct action of chlorine on slaked lime. The potassium
chlorate, which Berthollet prepared by the action of chlorine on
hot potash, is a crystalline salt, is readily decomposed by
heat and forms a convenient source for the preparation of small
quantities of oxygen :

CHAPTER XII

THE HALOGENS.

A. FLUORINE.

Fluorine. — In 1813 and 1814 Davy published two papers
on " Fluor Spar " and the " Fluoric Compounds " ( Works,
V. 408—424, 425 — 436). In these he recognised the
existence of an element, resembling chlorine, which he
proposed to call FLUORINE. The origin of these fluoric
compounds is described in his " Elements of Chemical
Philosophy" (1812), as follows :

" There is a substance found abundantly in nature called
FLUOR SPAR ; it is usually either blue, green, yellow, or
white, transparent, and crystallised in cubes. It is a common
product of the mines in Derbyshire.

" When this substance, in fine powder, is mixed with oil
of vitriol and distilled in retorts of silver or lead, connected
with receivers of the same metal artificially cooled, an
intensely active fluid is produced. It has the appearance
of sulphuric acid, but is much more volatile, and sends off
white fumes when exposed to air. It must be examined
with great caution, for when applied to the skin it instantly
disorganises it, and produces very painful wounds. When
potassium is introduced into it, it acts with intense energy
upon it, and produces hydrogen gas and a neutral salt : when
lime is made to act upon it, there is a violent heat produced,
water is given off, and the same substance as fluor spar is
produced. When it is dropped into water a hissing noise is

233

234 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

produced with much heat, and an acid fluid not disagreeable
to the taste is formed if the water be in sufficient quantity.
It instantly corrodes and dissolves glass " (Works, IV.
347—348).

Hydrofluoric acid. — The acid vapour prepared by the
action of oil of vitriol on fluor spar had been used since 1670
at Nuremberg for etching glass. The liquid acid was dis-
covered in 1771 by Scheele (Essays, pp. 1-51), who called it
" fluor acid " ; but it was first obtained in its pure form in 1809

FIG. 41. — TWINNED CUBES OF FLUOR SPAR.
British Museum (Natural History).

by Gay-Lussac and Thenard, who distilled the acid in metal
vessels in order to avoid contamination by glass. Davy at
first regarded the liquid acid as a compound of water with an
anhydrous oxy-acid, strictly analogous with sulphuric acid.
But, after recognising chlorine as an element, he "was
forcibly struck by the analogy" between the chlorine
compounds and those prepared from fluor spar. After
many fruitless attempts to isolate oxygen from them, Davy
concluded :

xii THE HALOGENS 235

" That there exists in the fluoric compounds a peculiar
substance, possessed of strong attractions for metallic
bodies and hydrogen, . . . and which, in consequence of
its strong affinities and high decomposing agencies, it will
be very difficult to examine in a pure form, and, for the
sake of avoiding circumlocution, it may be denominated
FLUORINE, a name suggested to me by M. Ampere "
(Works, V. 423).

Davy concluded further that :

" Fluor spar and other analogous substances, . . . must
be regarded as binary compounds of metals, and fluorine"
( Works, V. 424), and

" That the pure liquid fluoric acid consists of hydrogen
united to a substance, which, from its strong powers of
combination, has not as yet been procured in a separate
form, but which is detached from hydrogen by metals"
(Works, V. 425).

This acid, composed of hydrogen and fluorine, is now
known as HYDROFLUORIC ACID.1 Davy made many attempts
to decompose the acid by the electric current, but did not
succeed in isolating fluorine. The element was, however,
prepared by this method by Moissan in 1886.

B. IODINE.

Iodine discovered by Courtois (1811). — The view that
chlorine was an element was greatly strengthened by the
isolation, in 1811 (Hist. Acad. Sri., 1813, p. cxxiv), of a
substance which resembled chlorine in its properties, and
like chlorine could not be decomposed into simpler con-
stituents.

1 This name was first suggested by Davy (Works, IV. 350) in 1812
to express the view that the liquid was a hydrated oxy-acid, but was
abandoned by him in the following year in favour of the older name
" fluoric acid." It was introduced again by Gay-Lussac, in 1814, to
express the fact that the acid contained, not water as Davy had suggested,
but hydrogen.

236 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

Davy, writing in 1814, says :

"This substance was accidentally discovered about two
years ago by M. Courtois, a manufacturer of saltpetre at
Paris. In his process for procuring soda from the ashes of
sea-weeds, he found the metallic vessels much corroded ; and
in searching for the cause of this effect, he made the dis-
covery. The substance is procured from the ashes after the
extraction of the carbonate of soda, with great facility, and
merely by the action of sulphuric acid : — when the acid is
concentrated, so as to produce much heat, the substance
appears as a vapour of a beautiful violet colour, which
condenses in crystals having the colour and the lustre of
plumbago " (" On a New Substance which becomes a
violet-coloured Gas by Heat," Davy's Works , V. 437).

The properties of iodine. — The new substance which
Courtois had discovered was described by Desormes and
Clement, at a meeting of the Paris Academy of Sciences on
November 29, 1813. Their paper was not published in
the Memoirs of the Institute, but a summary was given in
the Annales de Chimie under Courtois's name. The
properties of iodine are there described as follows :

" The new substance, which has been called iodine on
account of the beautiful violet colour of its vapour, has the
appearance of a metal. Its specific gravity is about
four times that of water. It is very volatile ; its odour is
analogous to that of [chlorine] ; it imparts a red-brown stain
to paper and to the hands, but this disappears after a short
time ; it is neither acid nor alkaline ; on putting it into a
retort and heating, it vaporises at a very gentle heat, about
75° C. It boils under water and produces a magnificent
violet vapour ; when sublimed in considerable quantity,
large, brilliant plates are produced, but these are not
massive ; it is little soluble in water, more in alcohol and
much in ether.

A red heat does not change the nature of iodine ; it
traverses a red-hot porcelain tube without alteration.

It is the same with iodine vapour in oxygen ; it is not
changed at all by a red heat. The violet vapour escapes

xii THE HALOGENS 237

from the action of the oxygen as if it were alone, and the
whole of the iodine is found again in the vessels in which
it is condensed" (Ann. de Chimie, 1813, 88, 305-306).

Iodine examined by Gay-Lussac and by Davy (1813-
1814). Iodine an element. — At the request of Clement,
iodine was examined more completely by Gay-Lussac l and
by Davy,1 between whom a strong and healthy rivalry had
arisen.

The two chemists agreed in regarding iodine as an
element. Thus Davy writes :

"The new substance, I find, is not decomposed when
voltaic sparks are taken in it in its gaseous state from
ignited points of charcoal : at first there are white fumes,
probably from the action of moisture or hydrogen in char-
coal, on the substance ; but these fumes soon cease, and
when the tube in which the experiment is made is cooled,
the substance appears unaltered.

" From all the facts that have been stated there is every
reason to consider this new substance as an undecompounded
body. In its specific gravity, lustre, . . . and colour, it
resembles the metals ; but in all its chemical agencies it is
more analogous to oxygen and chlorine " (Davy's Works,
V. 452—453).

The new element had been described in France as " ione "
or " iode " (Greek, 1008179, violet) Davy proposed to
describe it in English as IODINE as this " name will be
more analogous to chlorine and fluorine " (Davy's Works,
V. 454—455).

1 Davy's first paper was sent from Paris on December 10, 1813.
His " Further Experiments and Observations on Iodine " ( Works, V.
457-477) were dated from Florence on March 23, 1814, whilst a third
paper, "On a Solid Compound of Iodine and Oxygen" (Works, V.
492-502), was dated from Rome, February 10, 1815. Gay-Lussac read
two memoirs to the Paris Academy of Sciences on December 6 and
December 20, 1813 ; his completed " Researches on Iodine" were read
on August I, 1814 (Annales de Chimie, 1814, 91, 5-96; Ostwald's
Klassiker, No. IV).

238 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

Gay-Lussac expressed his opinion as follows : —

" Iodine is not inflammable ; it cannot even be combined
directly with oxygen. I consider it as a simple substance,
and I place it .... between sulphur and chlorine, because
it has stronger affinities than the former, and weaker than the
latter" (Ann. de Chimie, 1814, 91, 8; Klassiker, IV. 5).

The iodides. — Iodine combines directly with a large
number of metals, forming compounds which Davy de-
scribes as IODIDES. Thus Clement, describing the dis-
covery of Courtois, writes : —

" With metallic mercury it forms in the cold, by shaking,
a beautiful red powder resembling vermilion.

" Iodine readily attacks, in the cold, iron, zinc, tin, and
antimony. . . . These compounds are soluble in water.
Those with lead and silver are not soluble ; the former is of
a beautiful yellow colour" (Ann. de Chimie^ 1813, 88,
308—309).

Davy found that it formed a salt-like substance by com-
bination with the metal potassium : —

" I heated some potassium in a little glass tube, and
passed some of the substance in vapour over it ; at the
moment the vapour came in contact with the potassium,
there was an inflammation and the potassium burnt slowly
with a pale blue light. The substance formed by the action
of potassium was white, fusible at a red heat, and soluble
in water. It had a peculiar acrid taste " (Davy's Works,
V. 441).

Iodine combined with chlorine to form a yellow solid,
volatile by heat. Although it did not combine directly with
oxygen, a solid oxide was prepared by Davy in 1815 : —

" The compound of oxygen and iodine appears as a white
semi-transparent solid ; it has no smell, but a strong
astringent sour taste. . . . When strongly heated it decom-
poses, undergoing fusion at the moment, and is entirely
converted into gaseous matter and iodine, leaving no

xii THE HALOGENS 239

residuum whatever. . . . The gas formed is found to be
pure oxygen, and the solid sublimate produced is pure
iodine" (Davy's Works, V. 494).

Clement also records (Ann. de Chimie, 1813, 88, 509)
the formation by the action of ammonia on iodine of an
explosive black powder, which Gay-Lussac (Ann. de Chimie,
1814, 91, 29 ; Klassiker, IV. 15) recognised as an iodide
of nitrogen.

Hydriodic acid. — De'sormes and Clement found that
iodine (like chlorine) is capable of uniting directly with
hydrogen to form an acid : —

" Iodine is not acted on by passing it over red hot
charcoal ; but hydrogen effects a complete change in the
appearance of this substance.

" If a mixture of hydrogen with dry or moist iodine
vapour is passed through the red-hot tube, the violet colour
disappears, scarcely any trace of it can be seen, and a
colourless gas is collected, of which one part is promptly
absorbed by water and the other part is pure hydrogen.

" The water in which the soluble gas was collected was
very acid ; it acquired a red colour and became perceptibly
warm" (Ann. de Chimie, 1813, 88, 306 — 307).

This experiment was repeated by Gay-Lussac, who made
the further discovery, that

"Hydriodic acid gas is partially decomposed at a red
heat; the decomposition is complete if it is mixed with
oxygen ; water is formed, and iodine reappears. On the
other hand, I have found that on passing water and iodine
vapour through a porcelain tube at a red heat, there was no
decomposition. There is here a great difference between
iodine and chlorine, for the latter removes hydrogen from
oxygen ; but there is also a resemblance between sulphur
and iodine, since oxygen removes the hydrogen from both "
(Ann. de Chimie, 1814, 91, 18).

The partial decomposition of hydriodic acid by heat is
an excellent example of a balanced action —

hydriodic acid ^± hydrogen + iodine.

24o HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

In presence of oxygen the action proceeds to completion —
oxygen + hydriodic acid -> water + iodine.

This complete decomposition of the acid is directly contrary
to the main course of the action which Berthollet had
discovered in chlorine water exposed to sunlight —

oxygen + hydrochloric acid <— water + chlorine.
The properties of hydriodic acid. — Gay-Lussac found that
hydriodic acid was prepared more readily by combining iodine
with phosphorus and decomposing the iodide with water.
In this action water is decomposed ; the oxygen unites with
phosphorus to form phosphorous acid, whilst the hydrogen
unites with iodine to form hydriodic acid. Gay-Lussac
made a careful study of the acid and records its properties
as follows : —

" Hydriodic gas decomposes as soon as it comes into
contact with mercury ; the surface of the metal becomes
covered with a yellow-green substance, which is the iodide
of mercury, and if the contact is sufficiently prolonged, or if
it is shaken, the hydriodic acid is promptly decomposed ;
the iodine combines completely with the metal, and there
remains a volume of hydrogen gas which is exactly
the half of that of the hydriodic gas. I passed the gas over
zinc and potassium, and the product was always hydrogen
and an iodide. Thus, according to this analysis . . . there
can remain no doubt as to the nature of hydriodic gas.

" This gas is colourless ; its odour resembles that of hydro-
chloric gas ; its taste is very acid ; it contains half its volume
of hydrogen, and saturates its own volume of ammonia.
Chlorine removes its hydrogen instantly ; a beautiful violet
vapour is produced, and hydrochloric gas is formed " (Ann.
de Chimie^ 1814, 91, 14 — 15).

C. BROMINE.

Balard (1826) discovers bromine. — In a memoir "On a
peculiar substance contained in sea-water" (Ann. Chim.
Phys., 1826, 32, 337), Balard, a young French chemist,

xii THE HALOGENS 241

described the discovery of a new member of the group of
halogens :

"I have observed several times that, on treating with
chlorine-water the extract of the iodine-containing ashes of
seaweed, after having added a solution of starch, there
appears not only a blue zone due to iodine, but above this
a zone of bright yellow."

" This orange-yellow colour was shown also when I
treated the mother-liquor of sea-salt in the same way. . . .
The appearance of this hue was accompanied by a strong,
peculiar smell."

" The saline mother-liquor, treated thus with chlorine,
loses its colour and its characteristic smell by exposure
to air during one or two days " (loc. cit. p. 337).

" The yellow saline liquor, submitted to distillation,
liberated, during the first moments of boiling, thick, red
vapours which condensed on cooling to a liquid, which
retained most of the properties of the coloured liquor ; but
these properties were much more pronounced."

" In order to obtain this substance in a pure state, . . .
I passed the red vapours over calcium chloride. They
condensed in a small receiver, in volatile, deep-red drops,
filling the little vessel in which they were contained with
vapours comparable in colour with nitrous vapour " (loc. cit.
PP- 33

The aqueous liquid was decolorised by the action of
alkalis, and by reducing agents such as zinc and sulphuric
acid, sulphurous acid, hydrogen sulphide, and ammonia,
but the colour could be restored by the addition of more
chlorine. The properties of the substance suggested that it
might be a new chloride of iodine, but it gave none of the
usual tests for this element, and could not be decomposed,
even by the electric arc. It appeared, indeed, to be an
element, closely resembling chlorine and iodine, forming
a series of analogous compounds, but clearly to be dis-
tinguished from these elements both in its physical properties
and in its chemical actions.

242 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

On account of its odour the new element was described
(loc. at. p. 341) as BROMINE (Greek, /Spw/xos, a stink).

Liebig fails to recognise bromine as an element. — The
German chemist Liebig (1803-1873), in a candid piece of
autobiography, wrote in 1838 : —

" I know a chemist who, while at Kreuznach, many years
ago, undertook an investigation of the mother-liquor from
the salt-works. .He found iodine in it ; he observed,
moreover, that the iodide of starch turned of a fiery yellow
by standing overnight. The phenomenon struck him ; he
procured a large quantity of the mother-liquor, saturated it
with chlorine, and obtained by distillation a considerable
amount of a liquor colouring starch yellow, and possessing
the external properties of chloride of iodine, but differing in
many of its reactions from the latter compound. He
explained, however, every discrepancy most satisfactorily to
himself; he contrived for himself a theory on it."

" Several months later he received the splendid paper
of M. Balard, and, on the very same day, he was in a con-
dition to publish a series of experiments on the behaviour
of bromine with iron, platinum, and carbon, for Balard's
bromine stood in his laboratory, labelled liquid chloride of
iodine. Since that time, he makes no more theories unless
they are supported and confirmed by unequivocal experi-
ments ; and I can positively assert that he has not fared
badly by so doing" (Ann. Chem. Pharm., 1838, 25, 29-30;
quoted by Hofmann, Trans. C/iem. ^"^.,1875, 28, 1098.)

D. CYANOGEN.

Scheele (1782) prepares prussic acid from prussian blue.
— Prussian blue, or Berlin blue, was first described in 1710,
as a product obtained by mixing lixivium sanguinis x with
a salt of iron. Scheele, in 1782, distilled both Prussian
blue itself and the lixivium from which it was made with
vitriolic acid in a glass retort. The distillate was a watery
liquor with a "peculiar smell and taste" (Essays^ p. 237).

1 An extract of calcined blood containing potassium ferrocyanide

xii THE HALOGENS 243

When re-distilled from powdered chalk to free it from traces
of oil of vitriol, the liquid appeared to be " neither acid nor
alkaline," since it neither reddened paper dyed with litmus,
nor restored the blue colour to a paper which had been
reddened by acids. It combined with ammonia and with
potash to form salts, but these were " decomposed by all
acids," even by carbonic acid (Essays, p. 280). It formed
salts with lime and magnesia, but was not able to liberate
carbon dioxide from chalk.

On account of its origin and its power of forming salts,
the French chemists, in 1787, described the liquid as PRUSSIC
ACID, and its salts as PRUSSIATES (Chemical Nomenclature,
tr. 1788, pp. 58 and 73).

Berthollet (1787) determines the composition of prussic
acid. — Scheele found that the vapour of prussic acid was
inflammable, and after combustion gave a precipitate with
lime-water (Essays, p. 285). Ammonia could also be
separated in many ways from the prussic compounds.
Bergman therefore thought " that prussic acid is composed
of carbonic acid, of volatile alkali, and of phlogiston."

Berthollet, in 1787 (Mem. Acad. Set., 1787, 148-162;
compare Ann. de Chimie, 1789, 1, 30-39), made a careful
study of the acid, and especially of the action of chlorine
upon it, and expressed his views as to its composition as
follows :—

" I conclude from this, that hydrogen and azote exist in
prussic acid, that they are combined with charcoal, and that,
when oxygen is added, all the principles necessary for the
formation of carbonate of ammonia are present together ;
but in order that they may take the form of ammonia and of
carbonic acid, the concurrence is needed of alkali or of
lime which tend to combine with the carbonic acid. . . ."

" It would seem to me then that there remains no doubt
as to the composition of prussic acid. . . It is a compound
of azote, of hydrogen, and of pure charcoal or carbon"
(Joe. at., p. 159 ; compare Ann. de Chimie, 1789, 1, 37 and 38).

R 2

244 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP

Gay-Lussac (1815) prepares and analyses anhydrous
prussic acid. — Gay-Lussac, in 1811, (Ann. de Chimie, 1811,
77, 128) prepared anhydrous prussic acid by distilling its
mercury salt with concentrated hydrochloric acid. In his
memoir" On Prussic Acid" (Ann. de Chimie, 1815, 95,
136-231), he describes it as follows : —

" It is a colourless, odorous liquid, with a taste at first
fresh, then burning . . . and a true poison. Its density
at 7° is 07058; at 18°, 0*6969: it boils at 26*5°, and
solidifies about 15° below the temperature of melting ice;
it then crystallises regularly and sometimes assumes the
fibrous form of nitrate of ammonia. The cold which it
produces in vaporising, even in air at 20°, is sufficient to
freeze it. This phenomenon is easily produced by putting
a drop at the end of a slip of paper or a tube of glass.
Although I rectified the acid several times over powdered
marble, it always kept the property of reddening blue litmus
paper slightly : the red colour disappeared as the acid
evaporated" (Ann. de Chimie, 1815, 95, 145-146).

The anhydrous acid was vaporised into oxygen and
exploded in a Volta's eudiometer. It gave an equal volume
of carbonic anhydride and half its volume of nitrogen ; one
and a quarter volumes of oxygen were used, of which one
volume was required to produce one volume of carbonic
anhydride, whilst the one-quarter volume was used in burn-
ing one-half a volume of hydrogen.

Gay-Lussac concluded that the acid "contained one
volume of carbon-vapour, one-half volume of azote and
one-half volume of hydrogen," but no oxygen. The correct-
ness of this analysis was proved (i) by the density of the
gas, (2) by passing it over iron heated in a porcelain tube,
when carbon was deposited and a gaseous mixture of equal
volumes of hydrogen and nitrogen was set free (Ann. de
Chimie, 1815, 95, 147-152).

Gay-Lussac regards prussic acid as a hydracid. — After
completing his analyses, Gay-Lussac writes : —

xii THE HALOGENS 245

"This acid, compared with the animal substances, is
distinguished by the great quantity of azote which it contains,
by less hydrogen, and above all by the absence of oxygen.
Its acidifying properties cannot depend on hydrogen, which
has an alkaline influence, but on carbon and nitrogen. It
must be considered as a true hydracid, in which carbon and
nitrogen replace the chlorine in hydrochloric acid, the
iodine in hydriodic acid and the sulphur in hydrosulphuric
acid"1 (Ami. de Chimie^ 1815, 95, 155).

Gay-Lussac regards prussic acid as a derivative of
the " compound radical " cyanogen. — Gay-Lussac found
that prussic acid, when acted on by potassium, liberated
half its volume of hydrogen and gave a salt, just as
hydrochloric and hydriodic acids did in similar circum-
stances. He therefore suggested that it contained a
COMPOUND RADICAL, CYANOGEN, which acted the part of the
simple radicals, chlorine and iodine :

"There is, then, a very great analogy between prussic
acid and hydrochloric and hydriodic acids : like them, it
contains half its volume of hydrogen, and like them it has
a radical which combines with potassium, and forms a com-
pound quite analogous with the chloride and iodide of
potassium : only this radical is compound, where chlorine
and iodine are simple."

" It seemed to me to be necessary to create a new name
to designate the radical of prussic acid. That of cyanogen 2
having seemed most suitable to the chemists of this capital,
I have adopted it and shall use it in this memoir. Ordinary
prussic acid will then receive the name of hydrocyanic acid
and the prussiates that of hydrocyanates. The compounds
of cyanogen with simple substances, when it plays the same
role as chlorine in the chlorides, will be designated by the
name of cyanides " (Ann. de Chimie, 1815, 95, 161, and
162-163).

Gay-Lussac's view of the nature of the cyanogen-compounds
was vindicated by the discovery of CYANOGEN GAS, which he

1 i.e., sulphuretted hydrogen. 2 Greek KVOVOS, blue.

246 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

obtained by heating mercuric cyanide (loc. cit. pp. 172-199).
As one of the first examples of a " compound radical," it
became of dominant importance in the development of
structural chemistry during the ensuing forty years.

E. HYDRACIDS AND HALOGENS.

Gay-Lussac (1814) abandons the oxygen-theory of
acids. — The discovery of hydriodic acid, confirming Davy's
contention that muriatic acid was a simple hydride, con-
taining no oxygen, made it necessary to abandon the
limitations imposed by Lavoisier's oxygen-theory of acids.
Gay-Lussac was impressed by the fact that sulphur, chlorine,
and iodine could form acids by combining either with
oxygen or with hydrogen. He suggested that these com-
pounds should be distinguished as follows :

Oxygen-Acids Hydrogen- Acids

SULPHURIC ACID HYDROSULPHURIC ACID1

CHLORIC ACID HYDROCHLORIC ACID

IODIC ACID HYDRIODIC ACID.

" Like [sulphur and chlorine, iodine] forms two acids, the
one by combining with oxygen, and the other by combining
with hydrogen. . . . The compounds of chlorine, of
iodine and of sulphur with hydrogen, which show the
properties of the acids formed by oxygen, must be included
in the same class under the same name of acid ; but to
distinguish them, I propose to add to the specific name of
the acid, the generic word hydro \ so that the acid compounds
of hydrogen with chlorine, iodine and sulphur, would bear
the name of hydrochloric acid, hydriodic acid, and hydro-
sulphuric acid ; and the acid compounds of oxygen with the
same substances, would have, according to the recognised
principles of nomenclature, the names of chloric acid, iodic
acid, etc." (Ann. de Chimie, 1814, 91, 8-9).

1 i.e. , sulphuretted hydrogen.

xii THE HALOGENS 247

Gay-Lussac on Hydracids. — In a section on Acidity and
Alkalinity, appended to his " Researches on Iodine," Gay-
Lussac urged, as a general proposition, that the name " acid "
must be extended to include a series of acidic hydrides,
the HYDRACIDS, in addition to the well-known OXYACIDS.

" For a long time I have regarded an acid, in the most
general sense of the word, as merely a substance, containing
oxygen or not, which neutralises alkalinity, and an alkali
likewise is only a body which neutralises acidity.

" Whatever may be the definition of acids at which one
arrives, they must be formed into several groups, because
they do not derive their acid character from the same-
substance. We have :

" i. The acids properly so-called, in which oxygen may be
considered as the acidifying principle, and which contain
only two elements ; such as ... sulphuric, sulphurous,
. . . carbonic acids.1

" 2. Acids formed by hydrogen and another substance :
this group includes the hydrochloric, hydriodic, and hydro-
sulphuric 2 acids. It is probable that in these acids, chlorine
iodine and sulphur are the acidifying principles ; but as
hydrogen is present in all of them, I thought it was more
convenient to derive their generic name from it. These
different acids can be designated by the name of
hydracids" (Ann. de Chimie^ 1814, 91, 145, 147, 148).

Berzelius (1825) on halogens and haloid-salts.— Gay-
Lussac' s dual classification of the acids was accepted by Ber-
zelius. It carried with it the recognition of two types of salts,
since the union of two oxides produced a ternary salt, whilst
the interaction of an acidic hydride with a basic oxide gave rise
to a binary salt and water. He therefore proposed in 1825 3
to describe the binary compounds as HALOID SALTS, and
their non-metallic components as HALOGENS.

1 i.e., sulphuric, sulphurous, and carbonic anhydrides.

2 i.e., sulphuretted hydrogen.

3 The Swedish paper was translated into French in 1826.

248 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

" What we call a salt must be defined according to a
certain electro-chemical relation and without any regard to
the number of the constituent elements. Thus we say that
the compound of chlorine with sodium is a salt, because
these two substances completely neutralise the electro-
chemical properties of one another. . . .

u According to this view, electro-negative substances may
be divided into three classes. First : those which form
salts by neutralising the electro-positive metals : I call these
halogens (salt-producers) ; they are chlorine, iodine, and
fluorine. . . ."

" Salts may be divided into two classes : those which
result from the combination of a halogen with an electro-
positive metal, and those which are formed by the union
of an acid with a base : I will call the first haloid-salts,
and the second amphi-salts" ("Memoir on Sulpho-salts,"
Ann. Chim. Phys., 1826, 32, 62 — 63 and 65).

In the second class of salts Berzelius included not only oxy-
salts formed from two oxides, \mi"sulpho-salts formed from
two sulphides, etc. The discovery of bromine in the
following year completed the series of four halogen-elements
(fluorine, chlorine, bromine and iodine), and added a new
group of haloid-salts, but did not lead to any further
theoretical developments.

The hydrogen-theory of acids. — The uniformity of com-
position, which was implied in Lavoisier's oxygen-theory of
acids, was destroyed by the discovery of the hydracids. It
was restored later by a theory which regarded all acids as
hydrogen compounds.

The " hydrogen theory " of acids was foreshadowed, in
1816, by Davy (Works, V. 514 — 515), who regarded it as
"a simple statement of facts to say that liquid nitric acid is
a compound of ... hydrogen . . . azote and . . . oxygen,"
and suggested that " hydrogen in [an] acid, may be con-
sidered as acting the part " played by a metal in the salts
derived from it.

xii THE HALOGENS 249

A very clear statement was made by Auguste Laurent
(1807 — 1853), who regarded all acids as salts of hydrogen,
strictly analogous with the salts of silver or zinc, but
differing from them in taste and volatility.

After considering the general resemblance between the
compounds of zinc and those of an element X, which turns
out to be hydrogen, he writes :—

" If hydrogen had been a comparatively fixed body, and
its oxide fixed and solid, no one would have thought of
excluding it from the metals. . . .

" Dumas has even gone so far as to conceive, that if we
could but condense hydrogen, it would present a metallic
aspect.1 . . .

" For my part I do not see any difference between acids
(meaning thereby the salts of hydrogen) and ordinary salts ;
neither do I recognise any difference between these two
kinds of bodies, and the group of oxides in general "
(Chemical Method, Cavendish Society Reprints, 1854, 41).

In Laurent's opinion, there is a profound difference
between the acids as salts of hydrogen and the oxides
from which many of them are derived by the addition of
water.

" Having thus associated certain compounds, which
chemists separate, I proceed now to separate certain other
compounds which they associate. The anhydrous and the
hydrated acids are always confounded and employed,
theoretically at least, the one for the other. Nevertheless,
these bodies do not present any analogy ; under no circum-
stances whatever, do they comport themselves in the same
manner : their functions are entirely different. Take for
example the anhydrous sulphuric acid and the sulphate of
hydrogen. The first does not combine with metals ; the
second does combine with them, at the same time liberating
hydrogen. The first unites directly with the metallic oxides ;

1 Liquid hydrogen, which Dumas described as " hydrogenium " to
indicate its metallic character, is not a metal, but resembles the
paraffins, to which series it really belongs.

250 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

the second gives rise to a double decomposition. The first
does not decompose chlorides, and even combines with
some of them ; whereas the second decomposes them. . . .

" The anhydrous and the hydrated acids, then, have not
any analogy with one another, and ought to be placed in
two perfectly distinct classes. . . .

" The anhydrous and the hydrated sulphuric acid have
this in common : they both contain sulphur and oxygen ;
but with regard to analogy of properties, they do not
manifest any whatever" (Chemical Method, 44 and 45).

Acids and acid anhydrides. — At the close of his
" Chemical Method" (pp. 376 — 377) Laurent describes the
acidic oxides as ANHYDRIDES, derived from the acids
directly or indirectly by the removal of water. Laurent's
views on the nature of acids were soon accepted ; thus, in
1864, the following footnote was added to the third edition

of Miller's " Elements of Chemistry" (Part II. p. 4) :

.

" The term acid has been employed by chemical writers
up to a late period, to designate indifferently either the
anhydrous bodies formed by the union of oxygen with the
non-metallic elements . . . (now more commonly called
anhydrides . . . ) or the hydrated compounds produced by
the action of water upon the anhydrides, such as ...
oil of vitriol or sulphuric acid. To avoid this confusion,
produced by the application of the same term to two
substances essentially distinct, it will be convenient to follow
the practice of many later authors, and to limit the term
acid to those hydrated bodies which are really salts of
hydrogen."

The system of nomenclature introduced by Laurent has
persisted to the present day. The oxides of sulphur are
described as SULPHUROUS ANHYDRIDE and SULPHURIC
ANHYDRIDE, whilst the names SULPHUROUS ACID and SUL-
PHURIC ACID are restricted to the hydrated acids formed
by combining the oxides with water to form hydrogen sulphite
and hydrogen sulphate respectively. So too the acidic oxide

xii THE HALOGENS 251

of carbon is called CARBONIC ANHYDRIDE instead of carbonic
acid, whilst the name CARBONIC ACID is applied to an
unstable hydrate, hydrogen carbonate, which is produced
when the gas is dissolved in water.

SUMMARY AND SUPPLEMENT
A. FLUORINE

Davyin 1813-14 recognised the presence in fluorspar (calcium
fluoride, CaF2) and in hydrofluoric acid (hydrogen fluoride,
HF ; described by Scheele in 1771 as "fluor acid") of an
element fluorine, closely resembling chlorine. The analogy
between hydrofluoric acid and hydrochloric acid appears in its
physical properties, in its method of preparation :

CaS04 + 2HF.

(Fluor- (Hydrofluoric

spar.) acid.)

2NaCl + H2SO4 -> Na2SO4 + 2HCl.

(Rock-salt.) (Hydrochloric

acid.)

and in the action of metals upon it :

+H2,
2HCT+2K->2KCl+Ht5

but mainly in the fact that no oxygen can be extracted from
any of the fluoric compounds. The acid acts readily on the
silica present in glass :

4HF + SiO2 ^± SiF4 + 2H2O,

(Silica.) (Silicon

tetrafluoride.)

giving rise to a gas, silicon tetrafluoride, SiF4, which is decom-
posed again by water, somewhat in the manner indicated by the
lower arrow in the above equation ; the actual product of the
decomposition by water is, however, a hydrosilicofluoric acid :

(SiF4-f2HF = H2SiF6)

252 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.
formed according to the equation :

(Hydrosilico-
fluoric acid.)

Davy attempted to isolate fluorine by passing an electric
current through hydrofluoric acid. This method was used
successfully by Moissan in 1886 (Ann. Chim. Phys. 1887,
[vi], 12, 473). An electric current was passed through a

FIG. 42. — MOISSAN'S APPARATUS FOR PREPARING FLUORINE.

mixture of hydrogen fluoride and potassium fluoride, con-
tained in a platinum U-tube (Fig. 42), and cooled to
-23° C. by the evaporation of methyl chloride. The elec-
trodes, //, of platinum iridium alloy, were insulated from the
U-tube by stoppers, FF, of fluorspar, the joints being made

xii THE HALOGENS 253

air-tight by leaden washers, p, screwed down by brass caps, E.
Hydrogen was liberated at the negative and fluorine at the
positive electrode. It is a light-yellow gas, condensing to a
liquid at - 187° C., and crystallising to a pale yellow solid at
- 233° C. It is a very active gas, unites explosively with hydrogen,
and at once decomposes water, forming hydrofluoric acid,
ignites charcoal, combines vigorously with phosphorus, liberates
chlorine from potassium chloride, and unites directly with
nearly all the metals, but has no marked action on dry glass.

B. IODINE

Iodine, discovered by Courtois in 1811 in the ashes of
seaweed, resembles chlorine in its properties, and in its resist-
ance to all efforts to effect its decomposition. It was examined
by Clement, by Gay-Lussac, and by Davy, four memoirs on
iodine being published within a fortnight at the close of 1813.
The beautiful violet vapour is not decomposed by passing
through a red-hot tube, alone or with oxygen, nor by the action
of white-hot charcoal. But it unites directly with the metals to
form fusible iodides. It also unites with sulphur and phos-
phorus ; with chlorine it forms a solid yellow chloride Id. It
unites directly at a red heat with hydrogen to form gaseous
hydrogen iodide or hydriodic acid, but this action is reversible
as the acid is decomposed partially by the same treatment :

(Hydrogen
iodide.)

Hydriodic acid is prepared more easily by the action of water
on iodide of phosphorus :

2PI3 + 6H2O -> 2H3PO3-t-6HI,

(Phosphorous (Hydriodic
acid.) acid.)

or by the action of sulphuretted hydrogen on iodine suspended
in water :

SH,+ I2-> S + 2HI.

The gas is intensely soluble in water, giving rise to a fuming

254 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

acid. It is decomposed by metals, liberating half its volume of
hydrogen, e.g. :

2Hg + 2HI->Hg2I2+H2.

(Mercurous
iodide.)

With ammonia, iodine interacts to form an explosive black
compound, nitrogen iodide, NH3,NI3. With potash it gives a
mixture si potassium iodide and potassium iodate (compare the
preparation of potassium chlorate) :

3l2 + 6KOH -> 5KI + KIO3 + 3H2O

(Potass- (Potass-

ium ium
iodide.) iodate.)

When this mixture is acidified iodine is reproduced,

(Hydriodic (lodic
acid.) acid.)

but iodic acid, HIO3, can be prepared by the action of sulphuric
acid on barium iodate,

Ba(IO3)2+H2SO4 -> BaSO4 + 2HIO3
or by dissolving iodine in nitric acid :

3l2+ioHNO3 -» 6HIO3+ioNO + 2H2O ;

when dried at 200° iodic acid loses water and is converted into
iodic anhydride, I2O6. A stable periodic acid, HIO4,2H2O, and
a very unstable hypoiodous acid, HIO, are also known.

C. BROMINE

Bromine, discovered by Balard in 1826, is a volatile red-
brown liquid, intermediate in its properties between chlorine
and iodine.

D. CYANOGEN

Scheele (1782), by the action of oil of vitriol on Prussian blue,
obtained a volatile acid, weaker than carbonic acid, to which the
French chemists in 1787 gave the name prussic acid. He

XIT THE HALOGENS 255

showed that it was inflammable and gave carbon dioxide when
burnt.

Berthollet (1787), by acting on prussic acid successively with
chlorine and an alkali, oxidised prussic acid to carbonic acid and
ammonia, and concluded that it was a compound of hydrogen,
carbon, and nitrogen. The action is expressed by the following
equations :

HCN + C12 -> CNC1 + HC1

(Prussic (Cyanogen

acid.). chloride.)

CNCl + KOH-f H2O->CO2 + NH3 + KC1.

Gay-Lussac (1811), by distilling the mercury salt of the acid
with hydrochloric acid, obtained anhydrous prussic acid as a
liquid boiling at 26-5° C. and freezing at - 15° C. In 1815, he
exploded the vapour with oxygen and found that it gave i vol.
carbonic anhydride, and \ vol. nitrogen, and that an additional
j vol. of oxygen was used to burn \ vol. of hydrogen to water.
The actual figures for 100 vols. of vapour were :

Contraction ... ... ... = 78*5 vols.

Contraction by potash ... ... = ici'ovols.

Nitrogen .. ... ... ... = 46*0 vols.

Oxygen used ... ... ... about 125 vols.

These numbers agree approximately with those required by
the equation :

4HCN + 5O2 -> 4CO2 + 2N2 + 2H2O Contraction

4 vols. 5 •vols. 4 vols. 2 vols. liquid — 3 vols.

Gay-Lussac was struck by the analogy of prussic acid with
hydrochloric and hydriodic acids and suggested that it contained
a "compound radical" cyanogen which acted the part of the
simple radicals chlorine and iodine. Prussic acid was therefore
called hydrocyanic acid and its salts became known as cyanides.
Typical formulas are :

Hydrocyanic acid (compare HC1), HCN

Potassium cyanide (sol. like KC1), KCN

Silver cyanide (insol. like AgCl), AgCN

Mercuric cyanide (compare HgCl2), HgC2N2.

These formulas are often written, HCy, KCy, AgCy, HgCy2,
to bring out their close analogy with HC1, KC1, AgCl, HgCl2,

256 HISTORICAL INTRODUCTION TO CHEMISTRY CH. xn

the symbol Cy being used as an abbreviation for the cyan-
ogen radical CN. Gay-Lussac prepared cyanogen gas by
heating mercuric cyanide :

E. HYDRACIDS AND HALOGENS

Iodine was recognised as an element both by Davy and
by Gay-Lussac ; its acid hydride was therefore a binary
compound, containing no oxygen. Gay-Lussac (1814) pro-
posed to divide the acids into two groups, namely : —

(1) the oxy-acids (SO2, SO3, CO2, etc.), and

(2) the hydro-acids, including in the latter group hydrochloric
acid (HC1), hydriodic acid (HI), and " hydro sulphuric acid"
(sulphuretted hydrogen, H2S) ; hydrofluoric acid (HF) and
hydrocyanic acid (prussic acid, HCN) were soon added to the
second group of acids.

Berzelius (1825) recognised the existence of two correspond-
ing series of salts, namely : —

(1) amphi- salts, which were ternary compounds, formed by
the union of an acidic and a basic oxide, sulphide or selenide, and

(2) haloid-salts (e.g. common salt, NaCt), which were
binary compounds of a metal with a halogen, fluorine, chlorine,1
iodine, etc.

Laurent (1854), following Davy (1816), regarded all acids
as salts of hydrogen ; sulphuric acid was then hydrogen
sulphate, H2SO4, whilst the oxide (SO3) which had been
regarded previously as sulphuric acid was now called sulphuric
anhydride ; so, too, carbonic acid was H2CO3 (compare Na2CO3)
whilst the oxide (CO2) was called carbonic anhydride.

1 Bromine was not discovered until the following year.

CHAPTER XIII

THE DECOMPOSITION OF THE ALKALIS

A. THE COMPOSITION OF AMMONIA

Alkalis regarded as elements. — At the time when
Priestley succeeded in collecting " acid air " and " alkaline
air" over mercury as permanently-elastic gases, there was
no reason for supposing that either of these gases was other
than a simple substance. The fixed alkalis were also still
undecomposed and were generally thought to be simple, or
elementary, substances.1 Thus Higgins, in a treatise, " On
Bleaching," published in 1799 (p. 18), writes :

" Pure pot-ash, according to our present knowledge of
chemistry, is a simple elementary substance."

The first alkali to be decomposed was the volatile alkali
which Priestley called "Alkaline air," and Bergman
" Ammonia." The decomposition was effected almost
simultaneously by Priestley himself and by Scheele,
Bergman's illustrious pupil.2

Priestley (1775) decomposes " alkaline air." — Priestley,
in 1773, found that "alkaline air" was inflammable. In

1 Lavoisier, however, suspected that they might be compound (see
below, p. 280).

2 " It was well observed to me by a near relation of Bergman that
the greatest of Bergmarfs discoveries was the discovery of Scheele "
(Thomas Beddoes, 1786).

258 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

1775 ne discovered further that it was decomposed by
electric sparks, increasing largely in volume, and was thus
converted into an inflammable gas which was insoluble in
water.

" I dipped a lighted candle into a tall cylindrical vessel,
filled with alkaline air, when it went out three or four times
successively ; but at each time the flame was considerably
enlarged, by the addition of another flame, of a pale yellow
colour ; and at the last time this light flame descended
from the top of the vessel to the bottom. At another time,
upon presenting a lighted candle to the mouth of the same
vessel, filled with the same kind of air, the yellowish flame
ascended two inches higher than the flame of the candle.
The electric spark taken in alkaline air is red, as it is in
common inflammable air " (Priestley, Experiments on Air,

I. 175).

" I took the electric explosion in a small quantity of
alkaline air, in the same manner as in the two preceding
experiments, and observed that every stroke added consid-
erably to the quantity of air ; and when water was admitted
to it, just as much remained unabsorbed as had been added
by the explosions. I then took about an hundred explosions
of the same jar, in a larger quantity of alkaline air ; after
which, so much of it remained unabsorbed by water, that I
could examine it with the greatest certainty. It neither
affected common air, nor was affected by nitrous air, and
was as strongly inflammable as any air that I had ever
produced " (Experiments on Air, II. 239-240).

Scheele (1774) oxidises ammonia by means of "man-
ganese."— The inflammable nature of ammonia was also
recognised by Scheele in his essay " On Manganese "(17 74).
It has been shown in a preceding chapter (p. 210) that this
mineral dissolves in dilute acids only in presence of reducing
or deoxidising agents, substances "rich in phlogiston," as
Scheele described them. Amongst these substances was.
the volatile alkali ammonia, the salts of which were able to
dissolve the mineral when acids alone failed to act upon it.

xin THE DECOMPOSITION OF THE ALKALIS 259

The ammonia was, therefore, a " phlogisticated " or
combustible substance.

Scheele also collected the non-inflammable constituent of
ammonia, and showed that it was an inert gas resembling
" phlogisticated air " or azote.

" If finely powdered manganese be exposed with pure
[nitric] acid, and some volatile alkali, to digestion
for several weeks, you will observe a great number of air
bubbles rise to the surface. This kind of air, collected in a
bladder tied to the neck of the flask, is not aerial acid [i.e.
carbonic anhydride], but of a quite different nature.
During the digestion, the volatile alkali is entirely decom
posed ; for if the solution be mixed and distilled with a
sufficient quantity of quicklime, not the least smell of
volatile alkali appears in the receiver. In this process, the
phlogiston, one of the constituent parts of volatile alkali,
has combined with the manganese, and thus rendered the
acid of nitre capable of acting upon the manganese. The
elastic fluid, collected in the bladder, has been either
separated from the volatile alkali, and is then its other
constituent part, or it is a product arising from its decom-
position. That the nitrous acid has no share here is
clearly proved by the following process :

" I repeated the very same distillation with manganese and
sal-ammoniac. ... It happened here as I expected; I
obtained the same kind of air in the bladder as in the
preceding experiment " (Scheele's Essays, p. 83).

Scheele (1777) oxidises ammonia by the calx of gold.

— Scheele prepared the same inert gas by detonating
FULMINATING GOLD. This substance is a compound of the
volatile alkali with calx of gold, prepared by adding the
volatile alkali to a solution of gold in aqua regia. It had
been known from the time of Basil Valentine to possess the
property of detonating violently when heated. Bergman
determined its composition by precipitating the calx of gold
by means of a fixed alkali, and then converting it into " aurum
fulminans " by the action of ammonia, or of one of its salts

s 2

260 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

(" On the Fulminating Calx of Gold," Bergman's Essays, II.,
134-165 ; pp. 155, 156). He also drove off the volatile
alkali by gentle heat, when the gold lost its power of
detonating (loc. cit. p. 145), and remained in the form of an
inert calx, the " aurum mutum " of the alchemists. When the
fulminate is detonated, the oxygen of the calx combines
with the inflammable constituent of the ammonia, the inert,
gaseous constituent is suddenly set free, and a residue of
metallic gold remains.

Scheele in the course of his Experiments on Air and Fire
confirmed Bergman's observations and actually extracted a
small quantity of ammonia from the fulminate. Following
Bergman he detonated half a drachm of fulminating gold
over water and obtained six ounce-measures of a gas, which
had the following properties :

was not miscible with water.

did not precipitate lime-water.

"(3) It extinguished immediately the flame of a burning
candle ; and was consequently air, in every respect,
similar to that generated by the destruction of volatile
alkali " (Experiments on Air and Fire, tr. J. R. Forster,
1780, pp. 142-143).

The liberation of gas by the action of chlorine on
ammonia is noted in his essay "On Manganese" (Essays,
p. 71).

Berthollet (1785) proves that ammonia is a compound of
hydrogen and nitrogen. — To the French chemist Berthollet
belongs the credit of proving definitely in 1785 that the
inflammable constituent of ammonia is hydrogen, and the
inert constituent nitrogen. He also made an accurate
quantitative analysis of the gas.

The nature of the inflammable constituent was determined
by neutralising the ammonia with nitric acid and heating the
crystalline nitrate thus produced. Nitrous oxide was set
free, and a considerable volume of water was condensed.

"(i) It wa
"(2) It di<

xin THE DECOMPOSITION OF THE ALKALIS 261

The formation of water was attributed to the oxidation of
the inflammable constituent of the ammonia by the oxygen
of the acid. Berthollet concluded from this experiment
that " the inflammable gas of water is a constituent part of
the volatile alkali" (Mem. Acad. Set., Paris, 1785, 99, 319).
The inert constituent was first examined as a product of
the action of chlorine on ammonia.

" The gas obtained from a mixture of volatile alkali and
[chlorine] showed all the negative properties which
characterise phlogisticated air or atmospheric [azote].1 ... It
follows that the volatile alkali is a compound of inflammable
air and [azote], and this result will be confirmed by the
experiments which will follow" (ibid. p. 320).

A more careful study of the gas from fulminating gold
proved that this gas was actually nitrogen.

" I made use of the gas which is set free in the detonation
of fulminating gold, in order to assure myself that it is
really [azote] that is contained in the volatile alkali; to
obtain it pure I filled a small glass retort with boiled
water, in which I placed the fulminating gold ; I distilled
this water at the pneumatic trough ; when the retort was
dry, most of the fulminating gold detonated, and although
there were seven grains of it, the retort was not broken,
because the mouth was free; by plunging the retort into
water I collected the gas which had been set free; I
mixed it with the proportions of vital air indicated by
Cavendish; I submitted this mixture to the action of
electricity, and the diminution which it experienced was
similar to that which one obtains when dealing with
atmospheric [azote]" (ibid. pp. 321 — 322).

"One can now explain the properties of fulminating
gold : the gold is there present as a calx, that is to say,
united with vital air ; it is also combined with volatile
alkali, as Bergman has so well shown ; the inflammable
gas of the volatile alkali forms water with the vital air, its

1 " Mofette" is the word used in the original paper.

262 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

[azote] is suddenly liberated and the metal is revived"
(ibid. p. 321).

Although the actual conversion into nitre of the nitrogen
of the volatile alkali was not realised in this experiment,
Berthollet was able to quote two cases in which the converse
change had been effected, since ammonia had actually been
prepared by the action of zinc on a solution of nitre, and by
the action of a fixed alkali on the products of interaction of
tin and nitric acid. "It is necessary to note that in each
case nitric acid was present which must have provided the
[azote]" (ibid. p. 326).

Berthollet's analysis of ammonia.— In a paper published
in the Journal de Physique for April, 1786, vol. 28,
p. 273, Berthollet compared the weight of ammonia driven
off by gentle heat from fulminating gold, with the weight
of water produced when the fulminate was detonated ;
he " concluded that the inflammable gas formed almost
one-sixth by weight ... of the volatile alkali," the pro-
portion now accepted as correct being 3/17 instead of
3/18. In 1785 J he carried out a volumetric analysis which
showed that two volumes of ammonia gas were increased
to four volumes by sparking, and that these four volumes
consisted of nearly three parts of hydrogen with one part of
nitrogen.

" We made some alkaline gas from a mixture of one part of
sal-ammoniac and three to four parts of freshly-burnt lime,
so that it should be free from water . . . : we introduced
it into a glass tube provided with an exciter; the space
which it occupied was ... 1*7 cubic inches. It was
electrified until an increase in volume was no longer
perceptible ; after that, a little water was introduced into
the tube, and no absorption was perceived, although it was
shaken, so that the alkaline gas was completely decomposed ;
the increase of its volume was i '6 cubic inches ; then the

1 The paper was not published till 1788.

xni THE DECOMPOSITION OF THE ALKALIS 263

gas was introduced into a Volta eudiometer, mixing each
measure successively with vital air, and detonating after each
addition, in the following manner :

" (i) Two measures of vital air.

One measure of decomposed alkaline gas.

(2) One measure of vital air.
One measure of gas.

(3) Two measures of vital air.
One measure of gas.

(4) One measure of gas.

Total: five measures of vital air and four measures
of gas.

"These nine measures were reduced to 4*6 measures, so
that 4 '4 measures were destroyed. Care was taken, in this
experiment, to put an excess of vital air in order to be sure
of destroying all the combustible part ; but one knows that
combustion causes 145 measures of inflammable gas to
disappear for 74 of vital air. It follows then that the 4*4
measures destroyed represent 2*9 measures of inflammable
gas and 1*5 measures of vital air. The four measures of
electrified gas which were tested, would then represent 2*9
measures of inflammable gas and i'i of [azote]" (Mem.
Acad. Set., Paris, 1785, 99, p. 324).

St. Claire Deville in 1865 showed that the decomposition
of ammonia into nitrogen and hydrogen is not quite complete
(see Chapter XX). The decomposition is, in fact, another
example of a balanced action, as may be indicated by the
reversible equation,

Ammonia ^± nitrogen 4- hydrogen.

The proportion of ammonia which persists is so small
that Berthollet failed to detect any contraction of volume
on adding water to the gas.1

The composition of sal-ammoniac.— When Priestley had
prepared " acid air " and " alkaline air," he tried the experi-

1 The evaporation of water into the dry gas would produce an
increase of volume which would compensate for the decrease of volume
due to the absorption of a little undecomposed ammonia.

264 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

ment of mixing them, expecting that common air might be
produced. Instead, he obtained a crystalline deposit of
sal-ammoniac. This salt is, therefore, a compound of the
two gases with one another. In 1809, Gay-Lussac showed
(A.C.R. IV. 10) that sal-ammoniac is formed by the union
of ammonia and of hydrogen chloride in equal volumes.

Both gases, although regarded as simple substances when
they were first prepared, were afterwards shown to be com-
pounds containing hydrogen. Sal-ammoniac is, therefore,
a ternary compound of nitrogen, hydrogen and chlorine : '

[AMMONIA = {NITROGEN

SAL-AMMONIAC = -[ J HYDROGEN

[MURIATIC ACID GAS = {CHLORINE

It is analogous with the oxygen-salts such as :

SULPHATE 0F/COPPER OXIDE

COPPER = j SULPHURIC ANHYDRIDE =

but differs from them in being formed by the union of two
hydrides instead of two oxides. Such salts were described
as hydrogen- or HYDRO-SALTS, in contrast with the better-
known oxygen- or OXY-SALTS (see, for instance, Turner's
"Chemistry," 1847, I. 566).

Davy (1807) suspects the presence of oxygen in
ammonia. — Davy's success in isolating metals from the fixed
alkalis (see below, pp. 273-280) led him, in 1807, to
suspect that the volatile alkali might also contain a metallic
constituent in combination with oxygen and hydrogen.

"As the two fixed alkalies contain a small quantity of
oxygen united to peculiar bases, may not the volatile alkali
likewise contain it? was a query which soon occurred to
me in the course of inquiry ; and in perusing the accounts
of the various experiments made on the subject, some of
which I had carefully repeated, I saw no reason to consider
the circumstance as impossible" (Davy's Works, V. 92).

xin THE DECOMPOSITION OF THE ALKALIS 265

He made many attempts to prepare oxygen from nitrogen or
water from ammonia, but concluded at last that these could
not be produced from the pure dry gases. The presence of
oxygen in nitrogen was, however, accepted by Berzelius up
to about 1820, both chlorine and nitrogen being tabulated
as oxides in his essay on "Chemical Proportions "in 1819.
Davy (1812) also suspected the presence of hydrogen in
sulphur and phosphorus and, by analogy, in inflammable
substances and metals generally (Works, IV. 358-359).
These tentative hypotheses are of interest as showing the
great difficulty of establishing the elementary character of
any given substance.

Seebeck (1808) prepares an amalgam from ammonia.—
Davy's views as to the composition of ammonia led him to
attach great importance to the discovery of a metallic
amalgam prepared from it by the action of an electric
current. Berzelius and Hisinger had found (Ann. de Chimie,
1804, 51, 167) that concentrated aqueous ammonia was
decomposed into nitrogen and hydrogen by passing an electric
current through it, between platinum wires. Seebeck at
Jena and Berzelius and Pontin (Gilbert's Annahn, 1810, 6,
247-280) at Stockholm, discovered early in 1808 that an
amalgam was produced, instead of hydrogen, when mercury
was substituted for platinum. The amalgam was examined
immediately by Davy, who included a section " On the
production of an Amalgam from Ammonia " ( Works, V.
122) in a paper on the Decomposition of the Earths, read
before the Royal Society on June 30, 1808. Davy describes
its preparation and properties as follows :

" When a globule of pure mercury is negatively electrified
by a Voltaic apparatus of 100 pairs of plates, it being in
contact with solution of ammonia in a cavity made in a
piece of muriate of ammonia, or any ammoniacal salt,
moistened in such a manner, and so placed on a disc of
platina, that the circuit is completed ; the globule increases

266 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

rapidly in volume, the quicksilver loses its fluidity, and at
length becomes of the consistency of soft butter, and
arborescent crystallisations shoot from it, which are quite
solid. The amalgam so formed has perfectly metallic
properties. It effervesces copiously when thrown into water,
hydrogen gas is given off, and a solution of ammonia is
found in the water. When exposed to the air it gradually
loses its consistence ; it emits a strong odour of ammonia,
and reddens paper tinged with turmeric held above it ; and
at last is found merely quicksilver.

" I found a still more easy mode of making the amalgam
by employing mercury combined with a minute quantity of
potassium, sodium or barium.1 When a compound of this
kind is placed in contact with a solution of ammonia, or
any moistened ammoniacal salt, it enlarges to eight or ten
times its bulk, and becomes a soft solid, and may be pre-
served for a much longer time than the amalgam formed by
electrical powers ; it changes very slowly even under water "
(Works, IV. 353-354; the quotation is from Davy's
"Elements of Chemical Philosophy," published in 1812).

Composition of the amalgam from ammonia. — Davy
thought at first that the amalgam might have been formed
by deoxidation of ammonia and "attempted to procure a
peculiar metallic substance from it by distillation out of the
contact of air, but without success ; . . . on the application of
heat, hydrogen and ammonia were always evolved, and the
mercury recovered its former state. ... In the most
accurate experiments the proportions of ammonia and
hydrogen were two to one by volume " ( Works, IV, 355).

As these products were formed, even from carefully-dried
amalgam, he was obliged to adopt the alternative view of
Gay-Lussac and Thenard that " the amalgam consists of
mercury united to azote and hydrogen, the hydrogen being
in larger proportion than in ammonia" (Works, IV. 354).
Since two volumes of ammonia contained one volume of
nitrogen and three volumes of hydrogen, the compound
1 See below, pp. 275, 278, 281.

xiii THE DECOMPOSITION OF THE ALKALIS 267

present in the amalgam (which contained two volumes of
ammonia in combination with one volume of hydrogen) was
composed of one volume of nitrogen united to four volumes
of hydrogen.

Davy (1808) suggests the name "Ammonium."— In
suggesting the name AMMONIUM for the substance present
in the amalgam Davy writes :

"The more the properties of the amalgam obtained
from ammonia are considered, the more extraordinary do
they appear.

" Mercury by combination with about TWOO^ Part °f lts
weight of new matter, is rendered a solid, yet has its specific
gravity diminished from 13*5 to less than 3, and it retains
all its metallic characters; its colour, lustre, opacity, and
conducting power remain unimpaired.

"It is scarcely possible to conceive that a substance
which forms with mercury so perfect an amalgam, should
not be metallic in its own nature ; and on this idea to assist
the discussion concerning it, it may be conveniently termed
ammonium" (Works, V. 130-131).

Ampere (1816) and Berzelius (1823) on the ammonium
radical. — The discovery of ammonium amalgam had the
effect of emphasising the close analogy of the compounds of
ammonia with those of the alkali metals. Ampere in 1816
pointed out that the salts of the fixed and volatile alkalies
would be strictly analogous if the " ammonium " of Seebeck's
amalgam were regarded as the metallic base of the salts of
ammonia.

" This difficulty would disappear by admitting that . . .
the compound of one volume of nitrogen and four volumes
of hydrogen, which is united to mercury in the amalgam
discovered by Seebeck, and to chlorine in the hydrochloride
of ammonia, behaves as a metal in all the compounds which
it forms ; in such a way that a volume of ammonia gas
combined with half a volume of water-vapour would be
considered as a sort of oxide ; with half a volume of

268 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

sulphuretted hydrogen as a sulphide . . . ; finally with a
volume of hydrochloric or hydriodic gas, as a chloride or
iodide of this compound metal" ("On a Natural Classifica-
tion of Simple Substances," Ann, Chim. Phys., 1816, 2,
1 6, footnote).

According to Ampere's view the composition of sal-
ammoniac should be represented as follows :

SAL-

AMMONIAC = -

AMMONIA

NITROGEN I

HYDROGEN

MURIAT.CGAS • CHLORINE = CHLQRINE

This view was supported by the analogy of cyanogen (p. 245),
a poisonous gas containing nitrogen and carbon, which
"although a compound substance, shows all the properties
of the simple substances [chlorine, iodine, sulphur, etc.]
capable of acidifying hydrogen."

Berzelius, convinced at last that ammonia contained no
oxygen, adopted Ampere's view, and in 1823 {Jahresbcrickt,
1823, 2, 57) proposed to represent the composite metal
ammonium by a symbol Am, analogous with those which he
had already proposed for the metallic elements (p. 295).

Absorption and liberation of water in salt-formation, —
For many years confusion prevailed in reference to the
composition of salts, because no attention was paid to the
appearance or disappearance of water ;• this water cost nothing,
and, even when not in use as a solvent, could only be
excluded by very careful drying.

It was recognised that salts can be formed by the direct
combination of two oxides, two sulphides or two hydrides, e.g.

Two oxides : litharge + ^dride ~^ lead sulPhatev
Two hydrides \ ammonia + muriatic gas — >• sal-ammoniac.

xiii THE DECOMPOSITION OF THE ALKALIS 269

But when a basic oxide combines with an acid hydride,
water is formed in addition to the salt, e.g.

litharge + muriatic gas -- > lead chloride + water.

On the other hand, when a basic hydride combines with
an acid oxide ivater is absorbed1 in the formation of the salt.
Thus ammonia does not unite with sulphuric anhydride to
form a salt, but

ammonia -f water + — > ammonium sulphate.

The water is needed to convert the ammonia into am-
monium oxide, which can then unite with the acid oxide to
form a salt : alternatively, it might be used to convert the
acid oxide (sulphuric anhydride) into a hydro-acid (sulphuric
acid) which can then unite directly with ammonia to form
a salt, thus :

rr, . , ammonium , sulphuric ammonium

Twooxute +ar/ydride- ' su!phate.

sulphuric ammonium

Twohydndes: ammonia + • su,phate

It is evident from these considerations that the ammonium
salts of oxy-acids are quaternary compounds, containing
both hydrogen and oxygen, in addition to nitrogen and (e.g.)
sulphur.

The case of ammonium nitrate is specially interesting, as
the acid oxide and basic hydride are both compounds of
nitrogen. When the salt is heated, the hydrogen of the
"ammonium" unites with oxygen from the acid to form
water, and the nitrogen escapes in the form of nitrous oxide.
In the case of ammonium nitrite the decomposition is
almost explosive, the salt being decomposed completely
into water and nitrogen.

1 Sal-volatile, which Priestley prepared from ammonia and carbonic
anhydride, -is an exception ; the product is not ammonium carbonate,
but ammonium carbamate, which forms a carbonate only when dissolved
in water.

270 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

Daniell (1840), regards salts as binary compounds of two
radicals. — Gay-Lussac, in 1815, had recognised in
cyanogen a compound-radical which could act the part of a
non-metallic element, such as chlorine, iodine or sulphur.
Ampere, in 1816, recognised in ammonium a compound
radical which could act the part of a metal. The way
was thus opened up for a final unification of all the different
types of salts. The binary salts formed by the union of
metal and halogen now became the type instead of the
exception. In the ammonium salts, the metal was replaced
by a compound-radical : in the cyanides, the halogen was
replaced by a compound-radical. It only remained to include
the oxygen salts in the same category by regarding the sul-
phates as compounds of a metal, with a compound
SULPHATE-RADICAL. This radical was composed of sulphuric
anhydride united with the oxygen of the base, thus :•

LEAD OXIDE = {LEAD = LEAD'

LEAD SULPHATE = SULpHURIC J* 1 SULPHATE.

ANHYDRIDE ^SULPHUR J

just as the ammonium of sal-ammoniac (p. 268) was com-
posed of ammonia united with the hydrogen of the acid.

This view of the nature of salts was generally adopted
when Daniell, in 1840, showed that the first action of an
electric current on salt solutions was to decompose the salt
into metal and non-metal, if the radicals were simple, or if
the radicals were compound, into "metallic" and "non-
metallic "compound-radicals.

B. DECOMPOSITION OF THE ALKALIES AND EARTHS.

Volta (1790) produces an electric current by means of
the voltaic pile.— The Italian physicist Volta discovered in
1790 that insulated discs of zinc and copper, when brought
into contact for a moment and then separated, acquired

xin THE DECOMPOSITION OF THE ALKALIS

271

respectively a positive and a negative charge of electricity.
This method of producing electric charges was much
improved in the VOLTAIC PILE,1 which he described in the
Philosophical Transactions for 1800 (pp. 403-431). This
consisted of a series of pairs of zinc (or tin) and silver (or
copper) discs, separated by cards, which were soaked in
brine to render them conducting (Fig. 43). The voltaic
pile had the advantage that the electric charges were
replenished as quickly as they were drawn off, so that a
continuous ELECTRIC CURRENT could be obtained. The

FIG. 43.— THE VOLTAIC PILE.

It is composed of 32 discs of copper or silver, A A (e.g. silver coins) and

32 discs of zinc, z z, separated by cards soaked in brine and terminating in

two basins of water, b b.

current was not sufficiently intense to produce sparks,
except with the help of a Leyden jar, and would only
produce a severe electric shock when the hands were
moistened, or, better, dipped into bowls of water, bb \
but even a small pile was capable of producing a far greater
quantity of current than the largest of the electrical
machines then in use. The power of the pile could be
increased almost indefinitely by augmenting the number of

1 A voltaic pile presented by Volta to Davy is still preserved in the
Library of the Royal Institution.

272 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP

pairs of discs, whilst the efficiency of each unit was much
improved by replacing the wet cards by vessels of wood,
earthenware or glass, arranged side by side and filled with
brine, or with dilute alkali (Fig. 44) ; each CELL of the
BATTERY contained a zinc and a copper plate separated by a
layer of solution, the zinc and copper plates of adjacent cells
being soldered together.

Nicholson and Carlisle (1801) decompose water by the
electric current. — Nicholson and Carlisle discovered in 1801
that two platinum wires, attached to the terminal plates of a
voltaic pile and dipped in a glass of water, became
covered with bubbles of gas. Closer examination showed
that the gas collected from the positively-charged wire

FIG. 44.— THE VOLTAIC BATTERY OR "CROWN OF CUPS."

Plates of zinc, z z, and copper or silver, A A, are soldered together at a a and
immersed in cups of brine.

attached to the zinc plate of the pile, was oxygen, and that
hydrogen was liberated from the negatively-charged wire
attached to the silver plate of the pile. At the end of
13 hours, 72 grain-measures of oxygen and 142 grain-
measures of hydrogen were collected, i.e. " nearly the
proportions in bulk of what are stated to be the component
parts of water" (Nicholson's Journal, 1801, IV. 186). The
electric current had thus been able to overcome the affinity
of oxygen and hydrogen for one another, and had caused
both gases to be liberated in the free state. The discovery
of the decomposing power of the electric current was of
great importance because compounds had been decom-
posed previously only by heat, or by the action of some

xiii THE DECOMPOSITION OF THE ALKALIS 273

substance which combined very strongly with one con-
stituent and thus set the other free.

Berzelius and Hisinger (1803) decompose salts by the
electric current. — Berzelius and Hisinger in 1803 made
the further discovery that dissolved salts were also
decomposed by the current, the acid of the salt collecting
with the oxygen round the positively-charged wire and the
base with the hydrogen round the negatively-charged
wire (Ann. de Chimie, 1804, 51, 172).

Davy constructs an electric battery. — On leaving the
Pneumatic Institute at Bristol in 1801 to take up the Pro-
fessorship of Chemistry at the Royal Institution, Davy at
once took up the work of investigating the decomposing
action of the electric current. He built up for this purpose
a large battery " containing 24 plates of copper and
zinc 12 inches square, 100 plates of 6 inches, and
150 of 4 inches square, charged with solutions of alum
and [nitric] acid." After describing, in his Bakerian lecture
of 1806, "a number of decompositions and chemical
changes produced by electricity," he concluded "that the
new methods of investigation promised to lead to a more
intimate knowledge than had hitherto been obtained, con-
cerning the true elements of bodies " which had not hither-
to been decomposed (A.C.R. VI. 5).

Davy (1807) prepares a metal from potash. — During the
autumn of the following year Davy used the experience
gained in these researches in his brilliant and successful
attempt to decompose the caustic alkalies. Although Black
had demonstrated fifty years earlier the relation of these
substances to the mild alkalies, nothing whatever was known
of their nature. They had usually been regarded as
elements. Davy himself, knowing that ammonia was a com-
pound of nitrogen and hydrogen, expected that the caustic
alkalies would prove to be compounds of nitrogen with other
combustible substances, perhaps sulphur or phosphorus.

T

274 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

As the result of little more than a month's work, Davy
found himself in a position to "demonstrate the decom-
position and composition" of the caustic alkalies.

On applying the current to saturated solutions of potash
and soda, Davy found that " the water of the solutions alone
was affected, and hydrogen and oxygen disengaged with
the production of much heat and violent effervescence "
(A.C.R. VI. 7).

" The presence of water appearing thus to prevent any
decomposition," Davy passed the current through " potash
in igneous fusion," and found that " a most intense light was
exhibited at the negative wire, and a column of flame, which
seemed to be owing to the development of combustible
matter, arose from the point of contact."

All attempts to isolate the combustible matter under these
conditions failed, because the heat required to melt the
potash was more than sufficient to ignite the inflammable
product. Bu-t when Davy employed "electricity as the
common agent for fusion and decomposition," metallic
globules appeared at the negatively-charged surface ; whilst
oxygen was produced at the positive pole.

"A small piece of pure potash, which had been exposed
for a few seconds to the atmosphere, so as to give conduct-
ing power to the surface, was placed upon an insulated disc
of platina, connected with the negative side of the battery
. . . and a platina wire, communicating with the positive
side, was brought in contact with the upper surface of the
alkali. The whole apparatus was in the open atmosphere."

" Under these circumstances a vivid action was soon
observed to take place. The potash began to fuse at both
its points of electrization There was a violent effervescence
at the upper surface ; at the lower, or negative surface, there
was no liberation of elastic fluid ; but small globules having
a high metallic lustre, and being precisely similar in visible
characters to quicksilver, appeared, some of which burnt
with explosion and bright flame, as soon as they were formed,

xin THE DECOMPOSITION OF THE ALKALIS 275

and others remained, and were merely tarnished, and finally
covered by a white film which formed on their surfaces "
(A.C.R. VI. 8).

To the new metal thus produced, the " basis of potash,"
Davy gave the name POTASSIUM.
Davy describes the physical properties of potassium.—

The new metal at ordinary temperatures is

" a soft and malleable solid, which has the lustre of polished
silver," but " at about the freezing point of water it becomes
harder and brittle, and when broken in fragments, exhibits
a crystallised texture, which in the microscope seems
composed of beautiful facets of perfect whiteness and a
high metallic splendour."

" To be converted into vapour, it requires a temperature
approaching that of red heat ; and when the experiment is
conducted under proper circumstances, it is found unaltered
after distillation."

" It is a perfect conductor of electricity," and " an
excellent conductor of heat."

" Resembling the metals in all these sensible properties
it is however remarkably different from any of them in
specific gravity ; I found that it rose to the surface of naphtha
distilled from petroleum, and of which the specific gravity
was '86 1 and it did not sink in double distilled naphtha,1 the
specific gravity of which was about '770, that of water being
considered as i " (A.C.R. VI. 14).

Gay-Lussac and Thenard (^Physico- Chemical Researches,
1811, I. in) who prepared the metal in larger quantities by
allowing molten potash to flow on to iron filings heated to
whiteness in a gun-barrel jacketed with clay, found that the
density of potassium at 15° C. relatively to water at the same
temperature was 0*865. They also gave the melting point
as 58°, the value now accepted being 62-5°.

The chemical properties of potassium, — (a) Oxidation —
Davy had considerable difficulty in preserving the new

1 This statement made twice by Davy, is remarkable in that it makes
the density of sodium less than 077.

T 2

276 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

substance, since the globules " acted more or less upon
almost every body to which they were exposed." It could
be preserved in " recently distilled naphtha " ; but when
exposed to the atmosphere its metallic lustre was immediately
destroyed and a white crust formed upon it.

"This crust I soon found to be pure potash, which
immediately deliquesced, and new quantities were formed,
which in their turn attracted moisture from the atmosphere
till the whole globule disappeared, and assumed the form of
a saturated solution of potash."

When exposed to air or oxygen confined by mercury,

" an absorption of oxygen took place ; a crust of alkali
instantly formed upon the globule ; but from the want of
moisture for its solution, the process stopped, the interior
being defended from the action of gas."

The new metal was readily combustible ; when strongly
heated in oxygen " a rapid combustion with a brilliant white
flame was produced, and the metallic globules were found
converted into a white and solid mass," which Davy was not
able to distinguish from ordinary potash (A.C.R. VI. u).

(b) Combination with Chlorine, Sulphur, Phosphorus and
the Metals. — When introduced into chlorine gas, potassium
" burns spontaneously with a bright red light, and a white
salt, proving to be muriate of potash is formed " (A.C.R.
VI. 1 6). Muriate of potash is, therefore, a binary com-
pound of a metal with chlorine, of the same type as those
prepared from lead, silver, and iron ; this fact is expressed
in the name POTASSIUM CHLORIDE which is now given to
the salt.

Potassium also resembles the other metals in that it
combines readily with sulphur and phosphorus.

It possesses the property of forming alloys with other
metals. With mercury it forms an amalgam, " a substance
exactly like mercury in colour " ; " the compound is fluid at
the temperature of its formation ; but when cool it appears as

xin THE DECOMPOSITION OF THE ALKALIS 277

a solid metal, similar in colour to silver " even when con-
taining only one part of potassium to seventy parts of
mercury. When these amalgams are exposed to air, "they
rapidly absorb oxygen ; potash which deliquesces is formed ;
and in a few minutes the mercury is found pure and
unaltered." " When a globule of the amalgam is thrown into
water, it rapidly decomposes it with a hissing noise ; potash
is formed, pure hydrogen disengaged, and the mercury
remains free " (A.C.R. VI. 20)..

(c) Decomposition of Water. — The strong affinity of
potassium for oxygen was shown in a very striking manner
by its violent action upon water even when quite cold :

" When it is thrown upon water, or when it is brought
into contact with a drop of water at common temperatures, it
decomposes it with great violence, an instantaneous explo-
sion is produced with brilliant flame, and a solution of pure
potash is the result."

" When water is made to act upon the basis of potash
out of the contact of air and preserved by means of a glass
tube under naphtha, the decomposition is violent ; and there
is much heat and noise, but no luminous appearance, and
the gas evolved when examined in the mercurial or water
pneumatic apparatus is found to be pure hydrogen."

" So strong is the attraction of the basis of potash for
oxygen, and so great the energy of its action upon water,
that it discovers and decomposes the small quantities of
water contained in alchol and ether, even when they are
carefully purified."

" The basis of potash when thrown into solutions of the
mineral acids, inflames and burns on the surface" (A.C.R.
VI. 16, 17, 18).

(d) Reduction of Oxides. — Finally potassium was found to
be an active reducing agent, removing the oxygen from
metallic oxides and liberating the metal.

" When a small quantity of the oxide of iron was heated
with it, to a temperature approaching its point of distillation,

278 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

there was a vivid action ; alkali and grey metallic particles
which dissolve with effervescence in muriatic acid, appeared.
The oxides of lead and the oxides of tin were revived still
more rapidly ; and when the basis of potash was in excess,
an alloy was formed with the revived metal " (A.C.R.
VI. 22).

The decomposition of caustic soda. — Davy immediately
examined caustic soda to see if he could decompose it as he
had done potash. He found that :

" Soda, when acted upon in the same manner as potash
exhibited an analogous result ; but the decomposition
demanded greater intensity of action in the batteries, or
the alkali was required to be in much thinner and smaller
pieces" (A.C.R. VI. 9).

The " basis of soda " was a white, metallic substance very
similar to potassium.

To the metal thus prepared from soda Davy gave the
name SODIUM.

Physical properties of sodium. — The properties of sodium
were described as follows :

" The basis of soda .... is a solid at common tempera-
tures. It is white, opaque, and when examined under a
film of naphtha, has the lustre and general appearance of
silver. It is exceedingly malleable, and is much softer than
any of the common metallic substances."

" It conducts electricity and heat in a similar manner to
the basis of potash ; and small globules of it inflame by the
voltaic electrical spark, and burn with bright explosions."

" Its specific gravity is less than that of water. It swims
in oil of sassafras of 1-096, water being i, and sinks
in naphtha of specific gravity 0-86 1. This circumstance
enabled me to ascertain the point with precision. I mixed
together oil of sassafras and naphtha, which combine very
perfectly, observing the proportions till I had composed a
fluid, in which it remained at rest above or below ; and this
fluid consisted of nearly twelve parts naphtha, and five of oil
of sassafras, which gives a specific gravity to that of water,

xin THE DECOMPOSITION OF THE ALKALIS 279

nearly as nine to ten, or more accurately as 0-9348 to i "
(A.C.R. VI. 23—24).

Gay-Lussac and Thenard found the density at 15° to be
0-972 relatively to water at the same temperature.

" The basis of soda has a much higher point of fusion
than the basis of potash." Davy gave the melting point as
120° to i8o°F., but the value now accepted is higher, viz.
98° C. He was not " able to ascertain at what degree of
heat it is volatile ; but it remains fixed in a state of ignition
at the point of fusion of plate glass" (A.C.R. VI. 24).

Chemical properties of sodium. — In its chemical properties
sodium resembled potassium, but was much less vigorous in
its action on air, oxygen and water.

" When the basis of soda is exposed to the atmosphere,
it immediately tarnishes, and by degress becomes covered
with a white crust, which deliquesces much more slowly
than the substance which forms on the basis of potash. It
proves, on minute examination, to be pure soda.

"The basis of soda combines with oxygen slowly, and
without luminous appearance at all common temperatures ;
and when heated, this combination becomes more rapid ;
but no light is emitted till it has acquired a temperature
nearly that of ignition.

" The flame that it produces in oxygen gas is white, and
it sends forth bright sparks, occasioning a very beautiful
effect ; in common air, it burns with light of the colour of
that produced during the combustion of charcoal, but much
brighter.

" Its operation upon water offers most satisfactory evidence
of its nature. When thrown upon this fluid, it produces a
violent effervescence, with a loud hissing noise ; it combines
with the oxygen of the water to form soda, which is dissolved,
and its hydrogen is disengaged. In this operation there is
no luminous appearance" (A.C.R. VI. 24 — 25).

" When introduced into [chlorine], it burnt vividly with
numerous scintillations of a bright red colour. Saline
matter was formed in this combustion, which, as might have

280 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

been expected, proved to be muriate of soda" (A.C.R.
VI. 25).

Common salt was thus shown to be a binary compound
of sodium and chlorine, a fact which is expressed by the
name SODIUM CHLORIDE.

Sodium, like potassium, combined with non-metals such
as sulphur and phosphorus, formed alloys with tin and other
metals, and a solid amalgam with mercury.

The bases of potash and soda are metals.—" Should the
bases of potash and soda be called metals ? " This question
Davy answered in the affirmative. u They agree with metals
in opacity, lustre, malleability, conducting powers as to heat
and electricity, and in their qualities of chemical combination.
Their low specific gravity does not appear a sufficient reason
for making them a new class ; for amongst the metals them-
selves there are remarkable differences in this respect,
platina being nearly four times as heavy as tellurium."
Davy therefore decided " to adopt the termination which
by common consent has been applied to other newly-
discovered metals," and to call the new substances
potassium and sodium. He added, "Whatever future
changes may take place in theory, there seems, however,
every reason to believe that the metallic bases of the
alkalies, and the common metals, will stand in the same
arrangement of substances ; and as yet we have no good
reasons for assuming the compound nature of this class of
bodies" (A.C.R. VI. 33—35).

Davy (1808) separates new metals from the alkaline
earths. — Although the nature of the caustic alkalies had
scarcely even been conjectured l before their decomposition
by Davy, the similarity between the alkaline earths, such as
lime and baryta, and the metallic oxides, had long been

1 Lavoisier, in 1789, regarded the alkalis as "evidently compound,
although we are still ignorant of the nature of the principles which
enter into their composition" (Works, I. 137).

xin THE DECOMPOSITION OF THE ALKALIS 281

recognised and had indeed led Lavoisier to suggest that
they were of the same nature ( Works, I. 122, 126, 137).

Fresh from his recent successes with potash and soda,
Davy, in 1808, attacked the alkaline earths, and after some
failures isolated their metals in the form of amalgams with
mercury.

" The earths were slightly moistened, and mixed with one-
third of red oxide of mercury, the mixture was placed on a
plate of platina, a cavity was made in the upper part of it to
receive a globule of mercury, of from fifty to sixty grains in
weight, the whole was covered by a film of naphtha, and the
plate was made positive, and the mercury negative, by a
proper communication with the battery."

"The amalgams obtained in this way, were distilled
in tubes of plate glass, or in some cases in tubes of common
glass. These tubes were bent in the middle, and the
extremities were enlarged, and rendered globular by blowing,
so as to serve the purposes of a retort and receiver."

" The tube after the amalgam had been introduced, was
filled with naphtha, which was afterwards expelled by boiling,
through a small orifice in the end corresponding to the
receiver, which was hermetically sealed when the tube
contained nothing but the vapour of naphtha, and the
amalgam."

" I found immediately that the mercury rose pure by
distillation from the amalgam, and it was very easy to
separate a part of it; but to obtain a complete decom-
position was very difficult" (A.C.R. VI. 46 — 47).

In this manner Davy obtained metals from baryta, from
lime, and from the earth now known as strontia j these
metals he named BARIUM, CALCIUM, and STRONTIUM. He
seems also to have obtained a specimen of the metal con-
tained in magnesia ; he first called this metal " magnium,"
but afterwards reverted to the name MAGNESIUM.

Properties of the metals of the earths. — Barium was
described by Davy as " a white metal of the colour of silver,"
solid at ordinary temperatures but " fluid at a heat below

282 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

redness." "When exposed to air, it rapidly tarnished,
and fell into a white powder, which was barytes. When
this process was conducted in a small portion of air, the
oxygen was found absorbed, and the nitrogen unaltered;
when a portion of it was introduced into water, it acted
upon it with great violence and sunk to the bottom, pro-
ducing in it barytes ; and hydrogen was generated" (A.C.R.
VI. 49)-

Strontium was described by Davy as closely resembling
barium.

In reference to calcium, Davy writes :

" The metal from lime, I have never been able to examine
exposed to air or under naphtha. In the case in which I
was able to distil the quicksilver from it to the greatest
extent, the tube unfortunately broke, whilst warm, and at
the moment that the air entered, the metal, which had the
colour and lustre of silver, instantly took fire, and burnt
with an intense white light into quicklime" (A.C.R.
VI. 50).

The metal is now manufactured on a considerable scale
by passing an electric current through the muriate, CALCIUM
CHLORIDE, to which fluor-spar (CALCIUM FLUORIDE) is added
to render it more easily fusible.

" The metal from magnesia seemed to act upon the glass,
even before the whole of the quicksilver was distilled from
it. In an experiment in which I stopped the process before
the mercury was entirely driven off, it appeared as a solid,
having the same whiteness and lustre as the other metals of
the earths. It sunk rapidly in water, though surrounded
by globules of gas, producing magnesia, and quickly
changed in air, becoming covered with a white crust, and
falling into a fine powder, which proved to be magnesia"
(A.C.R. VI. 50).

Composition of the earths. — Having isolated metals
from the four alkaline earths, Davy was able to prove by

xiii THE DECOMPOSITION OF THE ALKALIS 283

direct experiment that they were actually metallic oxides.
Although he was not able to determine their composition
by direct weighing, Davy satisfied himself:

" That when the metals of the earths were burned in a
small quantity of air they absorbed oxygen, gained weight
in the process, and were in the highly caustic or unslacked
state ; for they produced strong heat by the contact of water,
and did not effervesce during their solution in acids."

Lime was thus proved to be CALCIUM OXIDE. Slaked
lime, formed by combining lime with water, is sometimes
regarded as HYDRATE OF LIME, but is more usually described
as CALCIUM HYDROXIDE, i.e. a ternary compound of calcium,
hydrogen and oxygen. Similar names are given to the
other earths.

Composition of the alkalis. — Davy at first regarded the
alkalis as " highly combustible metallic bases united to
oxygen" (A.C.R. VI. 43.), but he soon discovered that the
oxides required to be combined with water in order to
convert them into caustic alkalis. He also confirmed the
observation of Gay-Lussac and Thenard that when the
metals were burned in oxygen, PEROXIDES were formed,
from which the excess of oxygen could be detached by
further heating, or by the action of water in converting the
peroxide into a common caustic.

These relationships were demonstrated very clearly in his
experiments on the elementary nature of chlorine. He
combined chlorine gas (i) with metallic potassium, (2) with
the oxide and peroxide, (3) with the hydrated oxide (identi-
cal with ordinary caustic potash) and found that :

(i) In the direct combination of potassium with chlorine
no water or oxygen was separated, the only product being a
neutral muriate or chloride formed by the direct combination
of about i -i cubic inch of chlorine with each grain of pot-
assium.

284 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

(2) A grain of potassium burning in oxygen was found
(after igniting the calx to decompose the peroxide) to pro-
duce an absorption amounting to J cubic inch. By ab-
sorbing i J cubic inches of chlorine the oxide was converted
into chloride ; exactly half a cubic inch of oxygen was
regenerated, but " there was no separation of moisture,"
provided that the chlorine had been properly dried.

(3) Similar phenomena were observed in the action of
chlorine on the peroxide, the whole of the absorbed oxygen
being regenerated without liberating any water. But when
chlorine was admitted to " a white sublimate of hydrate "
formed by the action of moisture on the peroxide "it
instantly became transparent from the evolution of water ;
and on heating the glass in contact with the sublimate, its
opacity was restored, and water driven off." (A.C.R. IX.

41—43)-

Caustic potash was, therefore, a binary compound of
oxide of potassium with water, corresponding with slaked
lime, or a ternary compound of potassium, oxygen and
hydrogen. It is therefore described as POTASSIUM HYDROXIDE.
Similar experiments show that caustic soda is correctly des-
cribed as SODIUM HYDROXIDE.

SUMMARY AND SUPPLEMENT.
A. THE COMPOSITION of AMMONIA.

Priestley, in 1/73, showed that ammonia was inflammable
and in 1775 decomposed it by sparking. This method was used
by Berthollet in 1785 to determine the composition of the gas.
He found that

ammonia gave hydrogen + nitrogen

2 vols. 3 vols. i vol.

The action was shown by St. Claire Deville, in 1865, to be
incomplete and reversible.

2NH3 ZT 3H2 + N2

(Am- (Hydro- (Nitro-
monia.) gen.) gen.)

xni THE DECOMPOSITION OF THE ALKALIS 285

since the gas after sparking gave a cloud of sal-ammoniac when
mixed with hydrogen chloride ; he was even able to recombine
a mixture of hydrogen and nitrogen by sparking it in presence
of hydrogen chloride.

3H2 4- N2 + 2HC1 -> 2NH4C1

(Sal-ammoniac.)

The presence of an inflammable constituent in ammonia had
already been recognised in 1774 by Scheele who found that the
mineral "manganese" could be dissolved by acids when combined
with ammonia, in the same way that it was dissolved in presence
of combustible substances. In this process the non-inflammable
constituent was liberated as an inactive gas :

3MnO2+2NH4NO3-> Mn(NO3)2 +2MnO + 4H2O + N2

(Manganese (Ammonium (Manganous (Manganous (Water.) (Nitro-

dioxide.) nitrate.) nitrate.) oxide.) gen.)

The same inert gas was liberated when ammonia was oxidised
by the calx of gold during the detonation tf fulminating gold, an
explosive compound obtained by precipitating gold chloride
with ammonia.

Au2O3,2NH3-> 2Au + 3H2O + N2

(Fulminating (Gold.) (Water.) (Nitro-

gold.) gen.)

The gas prepared from fulminating gold was used by Ber-
thollet, when he sparked the inactive constituent of ammonia with
oxygen, and showed that it was diminished in the same way as
atmospheric nitrogen. Berthollet also showed that the inflam-
mable constituent was really hydrogen ; he oxidised ammonia
with nitric acid, by heating crystalline ammonium nitrate, and
collected the water produced

NH4NO3 -> 2H2O + N2O ;

(Ammonium (Water.) (Nitrous

nitrate.) oxide.)

the other product was pure nitrous oxide.

The discovery of the composition of ammonia, together with
that of muriatic acid (hydrogen chloride) showed that hydro-
salts could be produced by combining two hydrides, e.g.

NH3 + HC1 -> NH4C1

(Am- (Hydrogen (Sal

monia.) chloride.) ammoniac.)

286 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

in just the same way that oxy-salts were formed by combining
two oxides, e.g.

PbO + SO3 -> PbSO4

(Lith- (Sulphuric (Lead
arge.) anhydride.) sulphate.)

In 1808, Seebeck at Jena, and Berzelius and Pontin at
Stockholm, discovered that an amalgam could be produced by
passing an electric current through ammonia in the presence of
mercury. The amalgam is decomposed by heat into mercury,
hydrogen (i vol.} and ammonia (2 vols.). As ammonia (2 vols}
contains hydrogen (3 vols.} and nitrogen (i vol.\ the pseudo-
metal in the amalgam contains nitrogen (i vol.} combined with
hydrogen (4 vols} ; Davy in 1808, proposed to call it ammonium.
Amp&re, in 1816, showed that the salts of ammonia could be
regarded as strictly analogous with the salts of the metals if am-
monium were regarded as a " compound metal," which united
directly with halogens to form salts such as sal-ammoniac or
ammonium chloride :

NH3 +HC1 = NH4C1 or (NH4) Cl

(Am- (Hydrogen (Sal- (Ammonium

monia.) chloride.) ammoniac.) chloride.)

A compound of ammonia with half its volume of water vapour
constituted an ammonium oxide, which could unite with oxy-
acids (anhydrides) to form oxy-salts, thus :

2NH3 + H2O -> (NH4)2O

(Ammonium oxide.)

(NH4)20 + 80s -» (NH4)2S04

(Ammonium (Sulphuric (Ammonium
oxide.) anhydride.) sulphate.)

The oxy-salts of ammonia contain both oxygen and hydrogen.
In some cases these may be driven off together in the form of
water, e.g. :

NH4 NO3 -> 2H2O + N2O

(Ammonium (Water) (Nitrous

Nitrate.) oxide.)

NH4 NO2 -> 2H2O + N2

(Ammonium (Water) (Nitrogen.)

nitrite.)

The decomposition of ammonium nitrate, the "nitrum
flammans" of the alchemists, had been known from an early
period and was used by Berthollet in 1785 to prove the presence

xin THE DECOMPOSITION OF THE ALKALIS 287

of hydrogen in ammonia ; the decomposition of ammonium
nitrite takes place so readily that this salt was separated from
its solutions in a crystalline form for the first time in 1909 (P. C.
Ray, Trans. Chem. Soc., 95, 345).

BerzelillS, in 1823, proposed to represent the compound
radical ammonium, NH4, by the symbol Am, thus :
ammonium chloride = NH4C1 or AmCl.
Two compounds having the composition of

ammonium oxide, 2NH3+H2O = (NH4)2O or Am2O and
ammonium hydroxide, NH3 + H2O = (NH4)OH or AmOH,

were actually isolated in 1909 (Annual Reports, 6, 14) ; it is
remarkable that both compounds, like ammonia itself, melt at
- 78° C.

Daniell, in 1840, proposed to regard all salts as binary com-
pounds of two radicals. In the halogen salts of the metals both
radicals are simple or elementary. In the ammonium salts the
<' metallic " radical is compound. In the cyanides, sulphates,
nitrates, etc. the " non-metal" is replaced by a compound radical,
cyanogen CN, sulphate SO4, nitrate NO3, etc. These radicals are
the first products of the action of an electric current on the
salts. '

B. DECOMPOSITION OF THE ALKALIS

Volta, in 1790, discovered that electric charges appeared on
insulated discs of zinc and copper if brought into contact and
then separated ; in 1800, he devised a Voltaic pile in which the
charges were constantly renewed and .a continuous electric
current was produced. The pile was soon developed into the
Voltaic battery and very large batteries were constructed in
England for Davy and in France for Gay-Lussac and
Thenard.

Nicholson and Carlisle, in 1801, showed that water was
decomposed by the electric current into oxygen and hydrogen

2H20 -> 2H2 + 02

Berzelius and Hisinger, in 1803, showed that dissolved salts
were decomposed into alkali and acid by the electric
current, e.g. :

Na2SO4 + 2H2O -> 2NaOH + H2SO4

(Sodium (Caustic (Sulphuric

sulphate.) soda.) acid.)

288 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

These two actions were regarded by Daniel! in 1840 as due to
a decomposition of the salt into its two radicals, e.g.

Na2SO4 -> 2Na + SO4 ;

the radicals then acted on the water, so that caustic soda and
hydrogen were produced together at the negatively-charged
electrode (Faraday's kathode] where the current left the liquid,
whilst sulphuric acid and oxygen were produced together at the
positively charged electrode (Faraday's anode] where the
current entered the liquid, thus :

Kathode: 2Na + 2H2O -> 2NaOH + H2
Anode : SO4 + H26 -> H2SO4 + |O2

Davy, in 1807, after having failed to separate a metal from
aqueous caustic potash, passed an electric current across a piece
of solid potash, slightly moistened on the surface. The potash
was melted by the heat of the current andgave a light inflam-
mable metal, which Davy called potassium. A similar metal,
sodium, was separated from caustic soda. These metals were
intensely active ; they combined directly with chlorine to form
salt-like substances, and had so great an attraction for oxygen
that they were able to decompose water at ordinary temperatures,
liberating hydrogen and reproducing the caustic alkali ; they
were used constantly during the succeeding years as the most
powerful agents for extracting oxygen from substances in which
its presence was suspected.

Thus boron, the characteristic non-metal of borax (sodium
borate) and of ^nV<2«W(prepared by Homberg in 1702 and com-
monly known as " sedative salt "), was isolated in 1808, by Davy,
and by Gay-Lussac and Thenard, by heating boric anhydride
with potassium. Silicon, which exists as an oxide (silica or silicic
anhydride} in quartz and sand, and in the mineral silicates, was
prepared by Gay-Lussac and Thenard in 1809 by the action of
potassium on the fluoride. Aluminium, the characteristic metal
of alumina, of alum, and of clay, was isolated by Wohler in 1827
by the action of potassium on the chloride.

Davy, in 1808, separated a new series of metals from the
alkaline earths — magnesia, lime, strontia and baryta. The
earths were mixed with oxide of mercury, covered with naphtha
and submitted to the action of the electric current. Amalgams
were obtained, from which the mercury was distilled off, more or

xin THE DECOMPOSITION OF THE ALKALIS 289

less completely, leaving behind a residue pf metallic magnesium,
calcium, strontium or barium. The metals oxidised rapidly
and were converted into unslaked or caustic earths, which were
thus shown to be metallic oxides :

Magnesia = magnesium oxide MgO

Lime = calcium oxide CaO

Strontia = strontium oxide SrO

Baryta = barium oxide BaO

The slaked oxides, formed by uniting the oxides with water, are
called hydroxides, e.g.

slaked lime = lime + water

= calcium hydroxide, CaO,OH2 or Ca(OH)2

Davy at first regarded the caustic alkalis as oxides of pot-
assium and sodium, but further experiments showed that they
were really hydroxides like slaked lime, since
potassium and chlorine gave potassium chloride only

2K + C12 -> 2KC1,

potassium oxide and chlorine gave potassium chloride and oxygen
K2O + C12 -> 2KC1 + £O2,

(i vol.) (i vol.)

potassium peroxide and chlorine gave potassium chloride and
oxygen

K2O4 + C12 -> 2KC1 + 2O.J,

caustic potash and chlorine gave potassium chloride, oxygen
and water

2KOH + C12 -> 2KC1 4- K)2 + H2O.

u

PART II

CHEMICAL THEORIES

CHAPTER XIV

THE ATOMIC THEORY

A. DALTON'S ATOMIC THEORY

Theories as to the composition of matter. — From the
time of the Greek philosophers there had existed two
opposing theories of the constitution of matter (a) the
continuous, according to which a void could not exist, and
(b) the atomic, according to which all material substance
consisted of particles separated by spaces.

The great English philosopher Isaac Newton gave his
powerful support to the latter opinion, with the result that it
became more and more prevalent, until at the close of the
eighteenth century it was almost universal to think of sub-
tances as composed of minute particles, too small to be
visible even by means of the microscope ; these tiny
particles were generally spoken of as " atoms," i.e. as
indivisible. Newton's views as to the nature of atoms are
set out in the following passage :

" It seems probable to me, that God in the beginning
formed matter in solid, massy, hard, impenetrable, moveable
particles ; of such sizes and figures, and with such other
properties, and in such proportions to space, as most
conduced for the end for which he formed them ; and that
these primitive particles being solids, are incomparably
harder than any porous bodies compounded of them ; even

29' U 2

292 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

so very hard, as never to wear or break in pieces : no
ordinary power being able to divide what God himself made
One, in the first creation " (Optics, Book III ; Works, 1782,
IV. 260. Not included in the first edition of the Optics
published in 1704).

Atoms and molecules. — The vague atomistic theory of
the Greek philosophers was of little real value until Dalton
showed that it could be used to explain the laws which
govern chemical combination. The new ideas introduced
into the Atomic Theory by Dalton may be summarised as
follows :

(1) The ultimate particles of a pure substance, whether
simple or compound, are perfectly alike in size and weight.

(2) The " simple atoms" of an elementary . substance are
indivisible, and ca?i neither be created nor destroyed.

(3) The "compound atoms" of a chemical compound are
formed by the union of two or more elementary atoms.

(4) Combination between atoms takes place in the simplest
integral ratios, e.g. i atom of A with i, 2, or 3 atoms of B.

In order to avoid the difficulty of describing a group of
particles as an indivisible atom, it has become customary to
reserve the name ATOM for the simple elementary atoms of
which elements and compounds are alike composed.
Groups of atoms, whether of the same kind or of different
kinds, are distinguished as MOLECULES, i.e. little masses.

Dalton (1808) describes his atomic theory.— Dalton's
views as to the nature and methods of combination of the
atoms are set out clearly in his " New System of Chemistry,"
Vol. I., published in 1808 :

" Whether the ultimate particles of a body, such as water,
are all alike, that is, of the same figure, weight, etc., is a
question of some importance. From what is known, we
have no reason to apprehend a diversity in these particulars :
if it does exist in water, it must equally exist in the elements
constituting water, namely, hydrogen and oxygen. Now it

xiv THE ATOMIC THEORY 293

is scarcely possible to conceive how the aggregates of
dissimilar particles should be so uniformly the same. If
some of the particles of water were heavier than others, if
a parcel of the liquid on any occasion were constituted
principally of these heavier particles, it must be supposed to
affect the specific gravity of the mass, a circumstance not
known. Similar observations may be made on other sub-
stances. Therefore we may conclude that the ultimate
particles of all homogeneous bodies are perfectly alike in weight,
figure, etc. In other words, every particle of water is like
every other particle of water ; every particle of hydrogen is
like every other particle of hydrogen, etc." (New System,
1808, I. 142 ; A.C.R. II. 28).

" Chemical analysis and synthesis go no farther than to
the separation of particles one from another, and to their
reunion. No new creation or destruction of matter is within
the reach of chemical agency. We might as well attempt to
introduce a new planet into the solar system, or to annihilate
one already in existence, as to create or destroy a particle of
hydrogen. All the changes we can produce, consist in
separating particles that are in a state of cohesion or
combination, and joining those that were previously at a
distance" (A.CR. II. 29).

" If there are two bodies, A and B, which are disposed
to combine, the following is the order in which the com-
binations may take place, beginning with the most simple :
namely,

i atom of A + i atom or B = i atom of C, binary

1 atom of A + 2 atoms of B = i atom of D, ternary

2 atoms of A + i atom of B = i atom of E, ternary

i atom of A + 3 atoms of B = i atom of F, quaternary

3 atoms of A + i atom of B = i atom of G, quaternary

etc., etc.

(New System, 1808, I. 212 ; A.C.R. II. 30.)

" From the application of these rules, to the chemical
facts already ascertained, we deduce the following con-
clusions :

(i) That water is a binary compound of hydrogen and

294 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

oxygen, and the relative weights of the two elementary
atoms are as i : 7, nearly.

(2) That ammonia is a binary compound of hydrogen
and azote, and the relative weights of the two atoms are
as i : 5, nearly.

(3) That nitrous gas is a binary compound of azote and
oxygen, the atoms of which weigh 5 and 7 respectively ;
that [nitrogen peroxide] l is a binary or ternary compound
according as it is derived, and consists of one atom of azote
and two of oxygen, together weighing 19 ; that nitrous oxide
is a compound similar to [nitrogen peroxide], and consists of
one atom of oxygen and two of azote, weighing 17. ..

(4) That carbonic oxide is a binary compound, consisting
of one atom of charcoal, and one of oxygen, together
weighing nearly 12 ; that carbonic acid is a ternary compound
(but sometimes binary) consisting of one atom of charcoal,
and two of oxygen, weighing 19; etc., etc. In all these
cases the weights are expressed in atoms of hydrogen, each
of which is denoted by unity" (A.C.R. II. 30-31).

Dalton's atomic symbols. — Much of the success of
Dalton's atomic theory was due to the system of symbols
which he introduced (A.C.R. 11.32), to represent the atoms
of elements, and the molecules of compounds. Typical
examples were :

Dalton's Berzelius's

Atoms. Symbol. Symbol.

Hydrogen .... Q H

Azote (J) N

Carbon .... @ C

Oxygen .... Q O

Sulphur . . • . 0 S

Dalton's Berzelius's

Molecules. Symbol. Symbol.

Water OO II + O

Ammonia .... ^^CD H + N

1 "Nitric acid" in the original.

xiv THE ATOMIC THEORY 295

Dalton's Berzelius's

Molecules. Symbol. Symbol.

Nitric oxide . . . ®O O + N

Nitrous oxide . . , ®O® O + 2N

Nitrogen peroxide . . O®O 2° + N

Carbonic oxide . . . O© C + O

Carbonic anhydride . . O C + 2°

Berzelius's atomic symbols (1819) — Dalton's system of
symbols was much improved by Berzelius, who, in his Essay
On the Theory of Chemical Proportions, published in
Paris in 1819, suggested that "Chemical symbols should be
letters of the alphabet, in order to be easily drawn and
printed without disfiguring the text " ; he therefore " chose
for this purpose the initial letter of the Latin name of each
element '?1 (loc. tit., p. in). Since many elements had the
same initial he proposed to use single letters to represent the
chief non-metallic elements, and to add a second letter to
distinguish the metals or other non-metals having the same
initial. As examples he gives :

S = Sulphur. C =Carbonicum (carbon.)

Si =Silicium (silicon). Co = Cobaltum (cobalt).

Sb = Stibium (antimony). Cu = Cuprum (copper).
Sn — Stannum (tin).

O =Oxygenium (oxygen).
Os = Osmium.

Compounds were distinguished by adding together the
symbols of the constituent elements, e.g. the two oxides of
copper were written Cu + O and Cu + 26. Berzelius
abbreviated his symbols by writing these two compounds
as Cu and Cu, the dots representing atoms of oxygen. In
the systems as used in the present day the two formulae

1 In a Swedish pamphlet published in 1814 Berzelius used initial
letters to represent the oxides of the chief mineral-forming elements.

296 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

would be written CuO and CuO2, but as the molecules of
the two compounds are now regarded as containing only a
single atom of oxygen the formulae are actually written as
Cu.2O and CuO.

Berzelius gave a list of fifty elements, with corresponding
symbols. The majority of the symbols are easily-recognised
abbreviations ; exceptions are found in the case of :

Gold Aurum Au

Silver Argentum Ag

Copper Cuprum Cu

Iron Ferrum Fe

Lead Plumbum Pb

Tin Stannum Sn

Mercury Hydrargyrum Hg

Antimony Stibium Sb

Sodium Natrium Na

Potassium Kalium K

The number of recognised elements has now grown to
over 80.

The laws of chemical combination explained by the
atomic theory. — One of the chief merits of Dalton's atomic
theory was that it afforded a simple and obvious explanation
of the three chief laws of chemical combination, namely :

(1) The law of fixed proportions.

(2) The law of multiple proportions.

(3) The law of reciprocal proportions.

B. FIXED PROPORTIONS.

The law of fixed proportions. — Throughout the alche-
mistic period very little attention was paid to the weights of
the substances obtained in chemical processes. The study
of chemical composition was confined to qualitative
experiments, designed to find out the nature and not the

xiv THE ATOMIC THEORY 297

proportions of the constituent elements of various sub-
stances. A similar disregard of quantitative measurements
prevailed during the vogue of the phlogiston theory, which
became untenable as soon as the balance became a recog-
nised tool of chemical science.

The credit of introducing quantitative conceptions into
chemistry belongs mainly to Black, whose work on
" Magnesia Alba " was a masterly demonstration of the
possibilities of the new method. Black's methods were
used with even greater effect by Lavoisier, the whole of
whose work was carried out upon a quantitative basis.

The work of Black and of Lavoisier was based upon the
tacit assumption of the LAW OF FIXED PROPORTIONS, that :

" Chemical compounds are formed by the combination of
their elements in fixed proportions by weight."

This law can be deduced at once from Dalton's theory
that " the ultimate particles of all homogeneous bodies are
perfectly alike in weight, figure, etc." Conversely, if the
law of fixed proportions were not true, Dalton's theory
would be a false hypothesis which could not stand the test
of experiment.

Proust (1799-1802) formulates the law of fixed propor-
tions.— The law of fixed proportions, which had been tacitly
assumed by Black and Lavoisier, became a definite doctrine,
to be tested and verified by experiment, in the hands of the
French chemist Joseph Louis Proust (1755-1826), who
carried out much analytical work bearing on this question
whilst Professor of Chemistry at Madrid. Proust was im-
pressed by the fact that the proportions in which the ele-
ments combine are " fixed by nature " and that the " power
of augmenting or diminishing " these proportions " is not
given to men " (Journ. de Physique, 1802, 55, 325). Proust
had already developed this view in 1799, when he found
that there was no difference of composition between natural

298 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

and artificial carbonate of copper and that " in art as in nature
copper never oxidises beyond 26%," whether the oxidation
was brought about by fire, or by dissolving in nitric acid or
by conversion into verdigris. Thus, in reference to the native
carbonate of copper, he writes :

" If 100 parts of this carbonate, dissolved in nitric acid
and thrown down by alkaline carbonates, give us 100 parts of
artificial carbonate ; if the base of these two compounds is
the black oxide, one must recognise an invisible hand which
holds the balance in the formation of compounds and
fashions their attributes at its will ; one must conclude that
nature does not act otherwise in the depths of the earth, than
on its surface, or in the hands of man " (Ann. de Chimie,
1799, 32, 30).

In the case of iron and sulphur, Proust found an apparent
exception to his rule ; iron pyrites was not identical with the
artificial sulphide of iron (p. 175), but was " surcharged
with an excess of sulphur," about 20% of which could be
driven off by heating the mineral in a retort (Nicholson's
Journal, 1802, 1, in; tr. homjourn. de Physique, 1801,
53, 89-97). But in the course of the year 1802 he pre-
pared an artificial sulphide, which appeared to have the
same composition as iron pyrites, and concluded that
iron and sulphur could combine in two fixed proportions.

" From the foregoing facts it follows, that iron can fix
sixty per cent, of sulphur by a considerably elevated tem-
perature. This proportion constitutes iron sulphurated to
the minimum."

" By a lower heat it can also attract a [further] quantity
which is equal to half of this weight ; and this result is iron
sulphurated to the maximum, or with ninety parts of sul-
phur. If this last combination be exposed to the temper-
ature which formed the first, it returns to that state ; that is
to say, it returns to the minimum of sulphuration, by giving
out all the sulphur it was capable of fixing above the pro-
portion of sixty parts for each quintal of iron " (Nicholson's

xiv THE ATOMIC THEORY 299

Journal, 1802, 1, 272; tr. from Journ. de Physique, 1802,
54, 89-95).

These facts are only partially correct. The artificial
sulphide contains 57 parts of sulphur combined with 100 of
iron : in iron pyrites there are 114 (not 90) parts of sulphur
combined with 100 parts of iron. When iron pyrites is
heated it loses one-third of its sulphur, as Proust supposed,
but the product is not identical with the artificial compound,
ferrous sulphide. *

Berthollet (1803) challenges the law of fixed propor-
tions.— In his Essay on Chemical Statics, published in 1803,
Berthollet definitely challenged Proust's opinion that in-
variable proportions and constant attributes characterise all
the true compounds of art or of nature, and that the chemist
is no more able to control these proportions and attributes
than he is able to control the affinities which the elements
possess for one another (Chemical Statics, tr. 1804, II.
315-316). Berthollet maintained that the elements
can combine in variable proportions, constancy of composi-
tion being secured only when some constituent crystallises
out, or distils out from the mixture of interacting substances.
In the special case of the oxides he proposed, in opposition
to Proust, to show

" that the proportions of oxygen in the oxides depend on
the same conditions as those which enter into the other
combinations ;

" that these proportions can vary, progressively, from the
term at which the combination becomes possible, to that at
which it acquires the highest degree ;

" and, that when this effect does not take place, it is only
because the conditions which I have pointed out become
an obstacle to this progressive action " (Chem. Statics, tr.
1804, II. 316).

He suggested that metals like zinc, which oxidise by
volatilising, take at once those proportions of oxygen which
may be considered as constant ; but that in the case of

1 The equation, 3FeS2 = Fe3S4 + S2, has already been given on p. 182.

300 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

metals such as tin and lead, which enter into a tranquil
fusion, oxidation advances progressively from a minimum
towards a maximum, producing a succession of colours and
of other properties. Thus lead forms an oxide which begins
by being grey ; afterwards it passes to different shades of
yellow, and finishes by becoming red. Iron also passes
through different shades and acquires different properties as
oxidation advances : similar effects may be observed in
several metals (loc. cit. p. 317).

In the case of sulphur, Berthollet expresses the view that

" The metals can combine with sulphur in very various pro-
portions, and the combinations which are thus formed
have different properties according to their proportions : I
am, in this case, again in opposition to the opinion of
Proust, who asserts, that by the invariable law of proportion,
sulphur and iron are fixed at —^ " (loc. cit. 372-373).

Proust (1804-1808) defends the law of fixed proportions ;
fixed proportions as a test of chemical combination.

— Proust replied to the criticisms of Berthollet in a series of
seven memoirs which appeared in the Journal de Physique
during- the last three months of 1804, It is not necessary
to follow the controversy in detail : it will be sufficient to
notice that Proust finished by using the law of fixed propor-
tions as a test to distinguish chemical compounds from mere
solutions of one compound or element in another. He was
able to appeal to the general feeling of chemical workers
to support the idea that there are in fact two different kinds
of combination, that the dissolution of sugar in water is
something different from the union of carbon, hydrogen and
oxygen to form sugar, and so forth. It was no small ser-
vice to point out that variable composition and fixed com-
position provide a test by which the two kinds of combina-
tion can be distinguished sharply from one another. These
views were expressed by Proust, in 1806, as follows :

"The attraction which makes sugar dissolve in water,

xiv THE ATOMIC THEORY 301

may or may not be the same as that which makes a definite
quantity of charcoal and of hydrogen dissolve in another
quantity of oxygen, to form the sugar of our plants ; but we
can see clearly, that the two sorts of attractions are so
different in their results, that it is impossible to confound
them."

" Thus the dissolution of nitre in water is for me quite
distinct from that of nitrogen in oxygen which produces
nitric acid, or from that of nitric acid in potash which
produces saltpetre."

" The dissolution of ammonia in water is in my eyes not
at all like that of hydrogen in nitrogen, which gives birth to
ammonia."

" The dissolution of silver sulphide in antimony sulphide,
which produces red silver ore,1 is not at all like that of
silver in sulphur which produces the sulphide of this metal."

" Finally, the dissolution of antimony sulphide in the
lower oxide of antimony is, to my feeling, not at all like that
of antimony in sulphur, and the reason of these distinctions
seems to me to be evident, for the dissolution of sugar, of
nitre, of ammonia in water, can be obtained in a range of
proportions of which the extremes are infinitely separated ;
but the dissolution, in one another, of the elements of nitric
acid, of saltpetre, or of ammonia has only been permitted to
us under the rigorous condition of a proportion of the two
or more constituents " (Journ. de Physique, 1806, 63, 369).

The distinction which Proust first made between a
DISSOLUTION and a COMBINATION (Journ. de Physique, 1806,
63, 370), and his description of the characteristics of a
CHEMICAL COMPOUND, have become regular features of the
chemical text-books of the hundred years which have
elapsed since his memoirs were published.

Stas (1865) tests the accuracy of the law of fixed
proportions. — The experiments of Proust were carried out
before the introduction into chemistry of the methods of
exact analysis which we owe to Berzelius ; they were there-
fore subject to errors varying from i per cent, to as much
1 Proust, Journ. de Physique, 1804, 59, 403.

302 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

as 20 per cent, in extreme cases. The exact analytical work
of Berzelius and others showed that constancy of composi-
tion could be relied on within very narrow limits ; but the
suggestion was made, in 1860, by the Swiss chemist Marignac
(Geneva Archives, 1860, 9, 103) that small variations of
composition might perhaps exist, even in the most stable
and definite of chemical compounds. This suggestion was
submitted to a most stringent test by Stas, who published a
memoir on the subject in 1865 ( Works, I. 445-481). Stas
made a very careful comparison of the quantities of common
salt required to convert different samples of silver into
silver chloride. The standard was a sample of silver which
had been distilled with the help of an oxy-hydrogen blowpipe.
The relative quantities of salt were as follows :

Method of Preparation. Salt used.

Distilled silver 100*000

Silver electrolysed from a cyanate solution ; metal I .^ \
USCd \99'997

f 99' 994

Silver nitrate reduced by milk-sugar ; metal fused . -j 99*995

1 99 '999

Silver nitrate reduced by sulphites ; metal fused 99'997
Silver chloride reduced by fusing with sodium car-
bonate and nitrate 99 995

Silver chloride reduced by heating with charcoal and
lime • . 99 '991

In these experiments ( Works, I. 466) the average deviation
from the mean value amounted only to 0*002 per cent.

A similar series of experiments (Works, I. 478) was made
on the precipitation of silver by means of different samples
of ammonium chloride, both at atmospheric temperatures
and at 100°. The results were as follows : —

Quantity used to
precipitate 100 parts

Method of Preparation. of silver.

Chloride prepared from cold solutions of hydro- ( 49 '600

chloric acid and ammonia. Ratio determined at -I 49 '599

ordinary temperature . . . . . . [ 49*598

THE ATOMIC THEORY 303

Quantity used to
precipitate 100 parts
Method of Preparation. of silver.

Chloride sublimed at ordinary pressure. Ratio I 49*597
determined at ordinary temperature . . .1 49*593

I 49*597

Chloride sublimed at ordinary pressure. Ratio I ? J?^
determined at 100° . . . . . . j ^Q..Q7

Chloride sublimed in a vacuum. Ratio determined / 49*598
at ordinary temperature \ 49*592

Mean .... 49 '997

In these experiments the average deviation from the mean
was only 0*004 Per cent.

The conclusion may be drawn that the law of fixed pro-
portions is exact up to the extreme limit of experimental
investigation.

Soddy and Hyman (Trans. Chem. Soc., 1914, 105,
1402 — 1408) have shown recently that different samples of
" lead " may unite with quantities of chlorine which differ
by one part in 225, a result that has been confirmed by
Richards and Lembert (/. Amer. Chem. Soc., 1914, 36,
1329 — 1344). The difference is attributed to the presence
in the metal of two elements, which are even more difficult
to separate than NEODYMIUM and PRASEODYMIUM, which
were regarded for a long time as a single element under the
name of DIDYMIUM. The exceptional difficulty of resolving
such mixtures cannot be regarded as affecting the validity of
the law of fixed proportions, as it applies exclusively to pure
chemical compounds.

C. MULTIPLE PROPORTIONS.

The law of multiple proportions. — The LAW OF
MULTIPLE PROPORTIONS states that :

" If several compounds be formed, the fixed proportions in
which two elements combine together are in simple integral
ratios to one another"

304 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

This law recognises that elements may combine together
in different proportions, but adds (i) that each of these
proportions is fixed -and (2) that the different fixed propor-
tions are not independent of one another, but are related in
the simplest possible way. Thus, if hydrogen and oxygen
unite together in the ratio i : 8 to form water, they may
also unite in the ratio 2:8, 3 : 8, 4 : 8, or i : 16, i : 24,
1:32 (or, generally, in the ratio n x i : m x 8, where n and
m are whole numbers) in forming other compounds ; but
they cannot combine together in intermediate ratios such as
1*107 : 8 or i : 7*823, which are wholly independent of the
ratio in which they are present in water. Berthollet, who
claimed "that a substance may combine in all proportions
with another on which it acts by a reciprocal affinity "
(Journ. de Physique, 1805, 60, 347), had no place for
the law of multiple proportions. Proust, who recognised
the existence in many cases of two or more fixed proportions,
and especially of a fixed maximum and minimum, might
easily have discovered the law of multiple proportions if his
analyses had been more exact. Thus, in the case of iron
and sulphur he gave the proportions as

sulphur to iron = 60 : 100, minimum
„ ,, „ =90 : 100, maximum

These two proportions are in the simple integral ratio 2:3;
but as the actual proportions are 57 : 100 and 114 : 100,
ratio 1:2, it is not surprising that his measurements as a
whole did not disclose any simple relationship between the
maximum and minimum proportions.

j The law of multiple proportions was therefore first dis-
covered by Dalton as a direct and obvious deduction from
fiis atomic theory. No formal statement of the law was
made, but several cases were quoted to which the law
can be applied, e.g. the three oxides of nitrogen in which

xiv THE ATOMIC THEORY • 305

the three proportions of oxygen to nitrogen l were in the
ratio 1:2:4, and the two oxides of carbon in which the
two proportions of oxygen to carbon2 were in the ratio 1:2.

The analytical data which Dalton used were no more
exact than those of Proust, and were, indeed, much less
exact than those on the strength of which he repudiated
Gay-Lussac's Law of Volumes (see below, p. 335). But,
guided by his theory, he was able to recognise the existence
of simple ratios, even in the imperfect data which were
at his disposal. The opinions of Berthollet, " that the
chemical agency is proportional to the mass, and that in all
chemical unions there exist insensible gradations in the
proportions of the constituent principles" (A.C.R. II. 27),
was quite incompatible with the atomic theory, according
to which one or more atoms of an element could combine
with i, 2, 3, etc., atoms of a second element, but not with
(any intermediate fractional number. Dalton therefore
jdenounced it as inconsistent "both with reason and
Observation " (A.C.R. II. 28), whilst Proust was able to
show that it had no sufficient experimental basis.

The law of multiple proportions supported by Thomson
(1808), by Wollaston (1808) and by Gay-Lussac (1809).

As soon as the law of multiple . proportions had been
discovered, ample evidence in support of it was forthcoming.
Thomas Thomson, whose History of Chemistry is of undi-
minished value after being in use for nearly a century,
directed attention to the existence of two series of OXALATES,
and especially of two potassium oxalates and two strontium
oxalates. The composition of these oxalates showed the
following simple relationship :

"Suppose TOO parts of potash; if the weight of acid

] Dalton gives the proportions as 7 : 10, 7 : 5, 14 : 5.
The actual proportions are . . 4 : 7, 8 : 7, 16 : 7.

Dalton gives the proportions as 7 : 5, 14 : 5.
The actual proportions are . .4:3, 8:3.

3o6 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

necessary to convert this quantity into oxalate be x, then
2X will convert it into superoxalate."

" It appears that there are two oxalates of strontian. . . .
It is remarkable that the first contains just double the pro-
portion of base contained in the second ;> (A.C.R. II. 41).

In a paper (A.C.R. II. 34-40) read before the Royal
Society on Jan. 28, 1808, a fortnight after Thomson's,
William Hyde Wollaston, the Secretary, showed that a third
oxalate of potash existed, the relative quantities of acid in
the three salts being in the ratio 1:2:4. He also showed
that in the two series of CARBONATES the quantities of gas
were in the ratio i : 2, since 4 grains of the bicarbonate,1
when converted into the carbonate by heating it, lost the
same amount of gas as 2 grains of the same salt when
wholly decomposed by acid ; these observations are com-
firmatory of those made 40 years before by Cavendish
(p. 69). Again, in the case of the SULPHATES, Wollaston
found that "super-sulphate1 of potash may be shown to con-
tain exactly twice as much acid as is necessary for the mere
saturation of the alkali present."

The gases which supplied Dalton with illustrations of the
law of multiple proportions were studied much more care-
fully by Gay-Lussac, who in his paper on the combining
volumes of gases (1809) quoted additional cases and pro-
vided more exact data, all tending to uphold the validity of
the law (see below, pp. 331-335).

The law of multiple proportions stated and proved by
Berzelius (1810). — One of the first chemists to feel
the fascination of the new problem of chemical pro-
portions was Berzelius, who published in 1810 a
"Research, to determine the fixed and simple Propor-
tions, in which the Constituents of inorganic Nature

1 Wollaston speaks of sub-carbonate (= carbonate) and carbonate
( = bicarbonate) ; sulphate and super-sulphate ( = bisulphate) ; oxalate,
binoxalate, and quadroxalate.

xiv THE ATOMIC THEORY 307

are combined with one another ; " this was translated
info German and published as a series of papers in Gilbert's
Annalen, 1811-1812, and has been reprinted as No. 35 of
Ostwald's Klassiker. Admitting at once the truthfulness of
Proust's view that substances could only combine together
in a series of fixed proportions, he had already found in his
own work some evidence that these fixed proportions were
related together in a very simple way. Further experiments
on the same range of compounds that had been studied by
Proust enabled him to assert that

" If two substances, A and B, combine with one another
in different ratios, this always takes place in the following
definite ratios: i A with iB (minimum); i A with ijB (or
perhaps more correctly 2 A with 36) ; lA with 26 ; lA with
46. In my experiments there is no example of lA with
36 " (Fixed Proportions, 1811-1812; Klassiker, XXXV. 5).

The following cases were studied by Berzelius : —

Lead and Oxygen.

Brown oxide ioo:i5'61 „ .. 0 AA

Yellow oxide 100 : 7-8 } Katw 2 °° : '

Sulphur and Oxygen.

Sulphuric acid J 100 : 146 -427 \ ,, . . .4Q7 .
Sulphurous acid ' 100 : 97-83 / Katl° 1 497 ' '

Copper and Oxygen.
Black oxide 100 : 25 \

Red oxide IOO:I2'5/

Iron and Sulphur.

100:117
Minimum 100 : 58

Maximum 100:117 1 z> , •

73} Ratl°

Iron and Oxygen.

Ferric oxide 100:44 '25! /> 4 • i AQZ ,

Ferrous oxide 100 : 29-6 ) Katw 1>495 : '

In each of these cases, the exact analyses of Berzelius
enabled him to recognise the validity of the law of multiple
proportions, which had remained disguised in the rougher
analyses of Proust.

1 i.e. sulphuric anhydride and sulphurous anhydride.

X 2

3o8 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

Accuracy of the law of multiple proportions. — The

measurements of Stas provide two opportunities of testing
the exactness of the law of multiple proportions.

(1) Dumas and Stas, in 1841 (see above, p. 148), showed
that 100 parts of carbonic anhydride contain

carbon 27 '27 per cent.,

oxygen 72 '73 per cent.

Stas, in 1849 (see above, p. 151), showed that 224*683
grams of carbonic oxide combined with 128*367 grams of
oxygen to form 353*050 grams of carbonic anhydride.
Therefore, TOO parts of carbonic anhydride contain

carbonic oxide 63 "64 per cent.

oxygen 36 '36 per cent.

Thus 27*27 grams of carbon unite with 72*73 grams of oxygen
to form 100 grams of carbonic anhydride, and with
63*64 — 27-27 = 36-37 grams of oxygen to form 63-64 grams

of carbonic oxide. The ratio ?2 '?3 = 1/9997. This ratio

36-37

differs from the integral ratio 2 : i by only one part in 7000
or 0*015 Per cent.

(2) Stas's analysis of silver sulphate (Works, I. 410)
showed that it contained 69*203 per cent, of silver, united
with 69*203 x 14*852 -i- 100= 10*278 per cent, of sulphur and
100-69-203 — 10*278 = 20-519 per cent, of oxygen. In
the sulphate, then, 100 parts of silver are united with
20-5 19 -"-69-203 x ioo = 29'650 parts of oxygen.

A second relationship between oxygen and silver is
afforded by the fact that 100 parts of silver were precipi-
tated by 69*103 parts of potassium chloride, which could be
derived from potassium chlorate containing 60*846 per cent,
of the salt and 39*154 per cent, of oxygen; 69*103 parts
of chloride would therefore combine with

69 • 1 03 x 391? 54 = 44467

60*846
parts of oxygen.

xiv THE ATOMIC THEORY 309

In the sulphate then the ratio of silver to oxygen is
100 : 29*650, the corresponding ratio derived from the study
of the chlorates is 100 : 44*467. These two quantities of

oxygen are in the ratio 44 4 7 = 14997. This differs from
29*650

the integral ratio 3 : 2 by only i part in 5000 or 0-02 per
cent.

D. RECIPROCAL PROPORTIONS

The law of reciprocal proportions. — The LAW OF
RECIPROCAL PROPORTIONS deals with the case in which two
elements combine with a third element, and also with one
another. It states that :

The relative proportions in which two elements combine with
a third element are in a simple ratio to those in which they
combine with a fourth element or with one another.

The significance of the law may be illustrated by a simple
example. In water, i part of hydrogen is combined with 3
parts of oxygen ; in marsh gas it is combined with 3 parts
of carbon, whilst in olefiant gas it is combined with 6 parts
of carbon, just twice as much as in marsh gas. What will
happen if oxygen on the one hand combines with carbon on
the other ? The law of reciprocal proportions suggests that
these elements will combine with one another in the
proportion

oxygen : carbon — 8 : 3 or 8 : 6

or some simple multiple of these numbers. The two pro-
portions suggested by the law are found in fact to be those
in which the elements are united in carbonic anhydride and
in carbonic oxide respectively ; in a recently-discovered sub-
oxide of carbon the proportions are

oxygen : carbon = 8:9.
Again, suppose that oxygen and carbon unite with a fourth

310 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

element such as sulphur, instead of with one another. The
law of reciprocal proportions suggests that equal quantities
of sulphur will combine with 8 parts of oxygen and with 3
parts of carbon, or with some simple multiple of these
numbers. In practice it is found that 16 parts of sulphur
unite with 3 parts of carbon and with 2x8=16 or with
3 x 8 = 24 parts of oxygen.

Like the two preceding laws, the law of reciprocal pro-
portions may be interpreted most easily with the help of
t)alton's atomic theory. But as an experimental law it
had been recognised and used twenty years previously.
Thus Cavendish had recognised in 1788 (Phil. Trans., 78,
178) that the quantities of nitric and sulphuric acid which
neutralised equal weights of potash would also decompose
equal weights of marble. The weights of potash and of
marble which saturated the same quantity of acid he had
described as long ago as 1766 as EQUIVALENT to one another
("On Rathbbne Place Water," Phil. Trans., 1767, 57, 102).

The law of reciprocal proportions was developed and
tested experimentally for a large range of acids and bases by
the German chemist J, B. Richter (1762-1807) in his
Stochiometry (1792-1794) and in a series of eleven volumes
on New Chemical Topics, published between 1791 and 1802.
Richter noticed that :

" When two neutral solutions are mixed, and a decom-
position follows, the new resulting products are almost with-
out exception neutral also."

" The elements must, therefore, have amongst themselves
a certain fixed proportion of mass" (Stochiometry, I. I24).1

Thus muriate of lime and sulphate of ammonia, when
mixed together in solution, were changed into sulphate of
lime and muriate of ammonia. The sulphate of lime

1 Quoted from Angus Smith's Dalton and the Atomic Theory (1856),
p. 190. In this book, twenty pages are given to translations from
Richter's works.

Xiv THE ATOMIC THEORY 311

separated as a white precipitate of gypsum, but the
neutrality of the solution was not affected by the double
decomposition and the liquid above the precipitate con-
tained nothing but perfectly-formed common sal-ammoniac
(Angus Smith, p. 207). No free acid or alkali was produced
because the relative weights of alkali and of earth required
to saturate one acid could be shown to be the same as those
required to saturate the other ; the lime and sulphuric acid
combining to form gypsum, therefore, liberated ammonia
and muriatic acid in exactly the right proportions to form
a neutral solution of sal-ammoniac.1

Richter's observations were collected by one of his con-
temporaries2 into a table which shows the weight of base
which will neutralise 1,000 parts of sulphuric acid or a tab-
ulated quantity of any other acid. A part of this table is
reproduced below.

EQUIVALENT WEIGHTS OF

(l) BASES

(2)

ACIDS

Alumina

• 525

Hydrofluoric .

• • 427

Magnesia

. 615

Carbonic

• • 577

Ammonia

. 672

Muriatic

. . 712

Lime

• 793

Oxalic .

• • 755

Soda

• 859

Phosphoric .

• 979

Strontia .

. 1329

Sulphuric

. 1000

Potash .

• 1605

Nitric .

. 1405

Baryta .

. 2222

Acetic

. 1480

1 Wenzel, who studied (Theory of Affinity, Dresden, 1777) the mixing
of copper sulphate with lead acetate to produce lead sulphate and
copper acetate, concluded from his analyses that the acetic acid set free
from the acetate of lead was not sufficient to dissolve the copper set free
from the sulphate of copper ; he calculated that 9^ out of 124 parts of
copper would remain undissolved in the precipitate of sulphate of lead.
There is therefore no justification for the circumstantial statement of
Berzelius, which attributes to Wenzel the discovery of the law
enunciated fifteen years later by Richter.

2 Fischer, in German translation (1802) of Berthollet's Researches on
the Laws of Affinity ; quoted by Berthollet, Chem. Statics, tr. 1804,
I. 402.

312 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

Berzelius's work on equivalents. — The importance of
Richter's observations was realised, perhaps for the first time,
by Berzelius, who writes :

" It follows from the researches of Richter, that from
good analyses of a few salts, one could calculate with
precision the composition of all the others.

"I formed [in 1807] the project of analysing a series of
salts, with the idea that it would be superfluous to examine
the others." Thus

"It is evident that if analyses are made of all the salts
formed by one acid, for instance, by sulphuric acid with all
the bases, and of those formed by one base, for example
baryta with all the acids, one would have the necessary
data to calculate the composition of all the salts formed by
double decomposition, provided that they retained their
neutrality" (Chemical Proportions, 1819, 16).

To Berzelius belongs the credit of having done more
than any of his contemporaries to provide the data whereby
the composition of a large range of compounds might be
calculated from the results of a few careful analyses, — not
merely in the case of acids, bases and salts, but throughout
the whole field of chemical combination.

The proportions by weight in which elements and com-
pounds interact with one another are described as their

EQUIVALENTS Or COMBINING WEIGHTS. As these weights

are all relative, some standard has to be selected to
take the place of the r,ooo parts of sulphuric acid used
above in summarising Richter's data.

Dalton selected one part of hydrogen, whilst Berzelius
preferred one hundred parts of oxygen. The standard used
at the present time is eight parts of oxygen, so that " The
equivalent of an element is that weight which combines with
eight parts by weight of oxygen"

The choice of oxygen instead of hydrogen as the stan-
dard has the practical advantage that all the elements

Carbon

3'oo

Calcium

Oxygen

exactly 8

Sodium

Nitrogen

(in nitrous

Copper

oxide)

. 14-01

Potassium

Sulphur

. 16-03

Mercury

Chlorine

•' • • 35-46

Lead .

Bromine

79-92

Silver .

Iodine .

. 126-92

xiv THE ATOMIC THEORY 313

except fluorine combine directly with oxygen, whilst the
formation of hydrides, at least in the case of the metals, is
the exception rather than the rule. The figure 8 is chosen
because it gives for hydrogen a value, i *oo8, which approxi-
mates very closely to unity.

Table of equivalents of elements. — The following table
shows the equivalents or combining weights of some of the
common elements.

Hydrogen . . . I'OoS Lithium . . .6*94

. 20-03
. 23 -oo
. - 3178
. • 39'iQ

. 100-3
• • 103-55
. 107-88

The table shows at a glance the composition of lime as
containing 20*03 parts of calcium combined with 8 parts of
oxygen, whilst litharge contains 103-55 parts of lead com-
bined with 8 parts of oxygen. Muriate of lime or calcium
chloride contains 20-03 parts of calcium combined with 35*46
parts of chlorine, whilst common salt contains 23-00 parts of
sodium combined with 35-46 parts of chlorine.

In drawing up a table of equivalents it is customary,
when multiple proportions are observed, to tabulate the
equivalent of the element as deduced from the analysis
of one compound, selected as typical, and to represent the
composition of the others by multiples of the tabulated
equivalents. Thus, in the table, carbon and oxygen are
shown as uniting together in the ratio 3*00 to 8 to form
carbonic anhydride ; but they also unite in the ratio
2 x 3-00 to 8 to form carbonic oxide. Again carbon and
hydrogen unite in the ratio 3*00 to 1*008 in marsh gas, but
in the ratio 2 x 3*00 to roo8 in olefiant gas. Such multiple

3i4 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

proportions are particularly common in compounds of three
or more elements, e.g.

LITHARGE . . lead : oxygen =103*55 : 8

and

SULPHURIC ANHYDRIDE . sulphur : oxygen = 16-03 : 3 x 8
unite in these proportions to form

LEAD SULPHATE . . lead : sulphur : oxygen

= 103-55 : J6'03 14x8

Equivalents and atomic weights. — The possibility of
making use of a series of equivalents to express the propor-
tions in which elements combine together finds an obvious
explanation in the atomic theory. But, whilst the equivalents
and the atomic weights are closew related to one another,
they are not necessarily identicafl. Thus, in the case of
hydrogen and oxygen the weights in which the elements
combine together to form water are approximately i to 8.
These are, therefore, the equivalents, or combining weights,
of the two elements. But whilst Dalton preferred to regard
water as a compound of i atom of hydrogen with i atom
of oxygen, Berzelius regarded it as a compound of 2 atoms
of hydrogen with i atom of oxygen. On Dalton's hypo-
thesis the equivalents and the atomic weights are both
expressed by the numbers i : 8, i.e. water is composed of
one atom of hydrogen of weight i, and one atom of oxygen
of weight 8. On Berzelius's hypothesis the equivalents are
still in the ratio i : 8, but the atomic weights were given as
6*2177 : 100 = i : 16, i.e. water is composed of two atoms
of hydrogen of weight i, and one atom of oxygen of weight
16, the combining ratio being 2 x i : 16 = i : 8 as before.
The numbers expressing the equivalent and the atomic weight
of an element must, however, be in a simple integral ratio to
one another, e.g. in the above instance 2 : i, or, in general,
m : n.

It may be noted that analytical data, such as those em-

xiv THE ATOMIC THEORY 315

ployed by Dalton, can only give the equivalents of the
elements and not their true atomic weights. Dalton's table
of atomic weights was thus really only a table of equivalents,
selected so as to give the simplest possible formulae to the
different compounds, but bearing little resemblance to the
real atomic weights as determined by the methods described
in the following chapter. The only check upon his arbitrary
selection of formulae was that a heavy gas was usually
represented as containing more atoms than a lighter one ;
for this reason he regarded carbonic oxide as a binary
compound, and carbonic anhydride as ternary, and therefore
calculated the atomic weight of carbon from the ratio 6 : 8
of carbon to oxygen in carbonic oxide instead of from the ratio
3 : 8 of the combining weights of the two elements in car-
bonic anhydride.

Stas's determination of equivalents. — A description has
already been given of the methods used by Stas in deter-
mining the combining weights of carbon with oxygen
(pp. 148 — 153). During a period of about 25 years, ending
in 1882, Stas determined with extreme care and accuracy
the combining weights of all the elements quoted in
the above table, with the exception of calcium, copper
and mercury. The method he adopted was as
follows :

(i) Potassium chlorate, which contains six equivalents of
oxygen in combination with one of potassium and one of
chlorine, was converted into potassium chloride

(a) By ignition

(b) By heating with hydrochloric acid.

It was found that 100 parts of potassium chlorate gave
60-846 parts of potassium chloride, so that 6x8 parts of
oxygen were combined with 48-^(100 - 60*846) x 60*846 =
74-592 parts of potassium chloride ; this number is,
therefore, the equivalent of potassium chloride.

316 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

(2) The weight of potassium chloride was determined
which was required to precipitate (as silver chloride) a known
weight of silver dissolved in nitric acid. It was found that
74-592 parts of potassium chloride would precipitate 107 '943
parts of silver ; this number is the equivalent of silver,

(3) By a variety of methods it was found that 100 parts
of silver combined with 32*845 parts of chlorine to form
132*845 parts of silver chloride ; since in these
experiments 107-943 parts of silver combined with

.g x 107-943 = 35454 parts of chlorine, this number

was the equivalent of chlorine.

(4) Subtracting 35-454 (the equivalent of chlorine) from
74-592 (the equivalent of potassium chloride) the equivalent
of potassium is found to be 39138.

At the close of his work Stas had determined the
equivalents of ten elements, as follows :

Nitrogen .' . 14-055 Lithium . . 7-022

Sulphur . . 16-037 Sodium . . 23*0455

Chlorine . . 35'457 Potassium . 39-1425

Bromine . . 79 -955 Lead . . 103-456

Iodine . . 126*848 Silver . . 107-930

A comparison with the list of equivalents given in the table
above (p. 313) will show that Stas's final values (which differ
slightly from those worked out in the four preceding para-
graphs) are almost identical with those accepted at the
present time.

Accuracy of the law of reciprocal proportions. — Perhaps
the best proof of the accuracy of the law of reciprocal pro-
portions is the existence of tables of atomic weights, which
chemists everywhere use to calculate the composition of their
compounds, rather than trust to even the best of their own
analytical data. Further evidence is afforded by the main-
tenance of neutrality when neutral salts interact, as noticed
first by Richter (p. 310). The question was, however, tested

xiv THE ATOMIC THEORY 317

definitely in 1811-1812 by Berzelius, who showed that 10
grams lead sulphide could be oxidised to lead sulphate with-
out leaving any excess of litharge or of sulphuric acid thus
proving " that lead sulphide contains its two constituents in
the exact ratio required for the formation of lead sulphate "
(Fixed Proportions, 1811-1812 ; Ostwald's Klassiker,
XXXV. n).

A still more stringent test was carried out by Stas, in
1865 (Works, I. 481-536), in response to the criticisms of
Marignac. Stas removed the oxygen from silver iodate and
examined the silver iodide produced, in order to see if it
contained any excess of silver or of iodine. The excess was
always less than i part in a million, and was often quite im-
perceptible, showing that the proportion by iveight of silver
to iodine, in the iodate and in the iodide, is invariably the same
( Works, I. 500). Similar experiments were made with silver
bromate, using about 20 grams for each experiment. In the
case of silver chlorate, " 259-4535 grams of this salt were
transformed into chloride without liberating a trace of silver
or of chlorine," thus proving that the proportion by weight of
silver and of chlorine is absolutely the same in th,e two salts
( Works, I. 535). Although fifty years have elapsed since this
test was carried out by Stas, it would scarcely be possible to
devise a more stringent proof of the accuracy of the law of
reciprocal proportions.

SUMMARY AND SUPPLEMENT.

John Dalton, about the year 1802, whilst studying the nature
of the atmosphere, revived the atomic theory of the Greek
philosophers and suggested that the "simple atoms" of an
element (i) are all alike in size and weight, and (2) cannot be
created or destroyed ; but (3) may unite with other atoms in
simple ratios to form " compound atoms," or molecules, as they
are now called. These atoms were represented by symbols, e.g.
hydrogen 0, oxygen Q> water OO? which Berzelius, in 1819,
replaced by letters e.g.H, O, H + O.

3i8 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.
The atomic theory afforded an obvious explanation of—

(1) The Law of Fixed Proportions. " The elements combine
together in fixed proportions by weight."

(2 ) The Law of Multiple Proportions. " When two compounds
are formed, these proportions are in a simple ratio."

(3) The Law pf Reciprocal Proportions. " The proportions
in which two elements combine with a third are in a simple
ratio to those in which they combine with a fourth element or
with one another."

The Law of Reciprocal Proportions as applied to acids and
alkalis was tacitly assumed by Cavendish, when in 1766 he
referred to certain quantities of marble and of potash as
equivalent to one another in their power of neutralising acids.
It was formulated more definitely about 1792 by J, B. Bichter
(1762 — 1807), who showed that neutral salts remain neutral
after double decomposition because the acids and alkalis
unite reciprocally in equivalent quantities, e.g. the quantities
of lime and potash which neutralise a fixed weight of sulphuric
acid are identical with those required to neutralise a (different)
fixed weight of nitric acid. A complete table of equivalents of
acids and alkalis was drawn up by Fischer in 1802.

Berzelius in 1807 recognised the practical value of the Law
of Reciprocal Proportions as consisting in the fact that, after
analysing carefully the compounds of all the bases with (e.g.]
sulphuric acid and of all the acids with (e.g.} baryta, one could
calculate the composition of all other neutral salts. He applied
the same idea to the elements and set to work to determine
their combining weights by the exact analysis of a few typical
compounds of each. This work was continued by J. S. Stas
(1813 — 1891), who determined with extraordinary care the equiv-
alents of carbon, nitrogen, sulphur, chlorine, bromine,
iodine, lithium, sodium, potassium, lead and silver. In recent
years similar work has been carried on by Prof. Guye at Geneva
and by Prof. T. W. Richards at Harvard University.

Lists showing the chief analyses used in determining the
equivalents or Combining weights of the elements have been
given by Ostwald (Outlines of General Chemistry, tr. 1912, pp.
126 to 150), and by Freund (Study of Chemical Composition,
1904, pp. 220 to 224).

xiv THE ATOMIC THEORY 319

It should be noted that Dalton had no trustworthy method of
finding out the numbers of atoms of each element which were
present in the molecule (Dalton's " compound atom " ) of a
compound : his tables of atomic weights were therefore similar
to a modern list of equivalents or combining weights.

CHAPTER XV

THE MOLECULAR THEORY

A. THE PROPERTIES OF GASES

The atomic and molecular theories. — The chief purpose
of Dalton's atomic theory was to explain the laws which
determine the composition of chemical compounds. The
molecular theory, which we owe to Avogadro, was on the
other hand very largely a physical theory, introduced in
order to account for the extremely simple properties which
characterise all substances when in the gaseous state. This
simplicity is specially obvious in the changes of volume which
are produced by

(1) changes of pressure (Boyle, 1662) ;

(2) changes of temperature (Charles, 1787);

(3) chemical combination (Gay-Lussac, 1809).

Boyle (1660) shows that air is an " elastic fluid." — In

his " New Experiments, Physico-Mechanical, touching the
Spring of the Air, and its Effects," published in 1660
(Works, 1725, II. 407-474), Boyle described an air-pump
(Fig. 45) which was an improvement upon the pump used by
Otto von Guericke in the celebrated experiment of the Magde-
burg hemispheres (1654). With the help of this pump Boyle
showed that a bladder, strongly tied at the neck, could be
dilated and even burst by the " spring " of the air which

320

FIG. 45— BOYLE'S PNEUMATIC ENGINE OR AIR PUMP.

The air was exhausted from the glass globe A by means of a sucking-pump
supported on a wooden frame 12. The insets show : i. A brass ring DE cemented
to the lip BC of the large glass A : it carries a glass stopper FG, pierced in the
middle to receive the brass key K. 2. The stopcock N, cemented above to the
glass DA, and below to the top PQ of the sucking-pump. 3. The brass cylinder 3, 3
of the sucking-pump, 14 inches long and 3 inches bore ; the sucker 44 was fitted
with a ring of tanned leather, and was driven by a handle 7 acting through a rack-and-
pinion gear 5, 5. The cylinder was charged with a little oil or water to make the
sucker air-tight. During the downward stroke of the sucking-pump, the valve R
was closed and the tap N opened to admit air from the globe ; during the upward
stroke N was closed and R opened.

321 y

322 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

it contained, if the air outside the bladder was removed by
means of the pump (Works, 1725, II. 413). Quantitative
measurements showed that air could be dilated in this way
to 9, 32, or 60 times its original volume; in an extreme
case, a " bubble of air possessing the space of but one
grain weight of water, ... by its own spring, was rarefied
to one hundred and fifty-two times its former dimensions ;
though it had been compressed only by the ordinary weight
of the contiguous air" (Works, 1725, II. 415). Analogous
experiments in which water was "freed from the pressure
of the atmosphere," revealed numerous little bubbles, pro-
duced by air lurking in the water, but gave no indication at all
" that water uncompressed, has an elastic power " (ibid., p. 428).

A few years later Mayow showed that " air in which an
animal or a lamp had expired, possessed elastic force in an
equal degree with unimpaired air." He also experimented
with gases prepared artificially by the action of acids on
iron (Boyle, 1660; Works, 1725, II. 432), and "ascertained
that air of that kind expands to more than two hundred
times its volume ; and indeed if it had been relieved from
the pressure of the surrounding water, it would have ex-
panded twice as much. Nor will common air, when treated
in the same manner, expand more" (A.C.R. XVII. 115).

By the experiments of Boyle and of Mayow, air was
shown to be typical of a group of " elastic fluids," which
differed from liquids in that they could expand in every
direction when the pressure acting upon them was reduced.
It was, however, left to Cavendish and to Priestley (p. 70)
to show clearly that the different gases or "airs" which
possess this remarkable property are distinct substances and
not merely different varieties of common air.

Boyle's explanation of the "spring of the air." In
order to explain his observations, Boyle suggested :

" That there is a Spring, or Elastical power in the Air we
live in. By which Spring of the Air, that which I mean is

xv THE MOLECULAR THEORY 323

this : That our Air either consists of, or at least abounds with,
parts of such a nature, that in case they be bent or com-
press'd by the weight of the incumbent part of the Atmo-
sphere, or by any other Body, they do endeavor, as much
as in them lies, to free themselves from that pressure, by
bearing against the contiguous Bodies that keep them bent ;
and, as soon as those Bodies are remov'd or reduced to
give them way, by presently unbending and stretching out
themselves, either quite, or so far forth as the contiguous
Bodies that resist them will permit, and thereby expanding
the whole parcel of Air, these elastical Bodies compose.

" This Notion may perhaps be somewhat further explain'd,
by conceiving the Air near the Earth to be such a heap of
little Bodies, lying one upon another, as may be resembled
to a Fleece of Wool. For this (to omit other likenesses
betwixt them) consists of many slender and flexible Hairs ;
each of which may, indeed, like a little Spring, be easily
bent or rolled up ; but will also, like a Spring, be still
endeavouring to stretch itself out again. ..."

" There is yet another way to explicate the Spring of the
Air, namely, by supposing with that most ingenious Gentle-
man, Monsieur Des Cartes, that the Air is nothing but a
Congeries or heap of small and (for the most part) of flexible
Particles ; of several sizes, and of all kind of Figures which
are rais'd by heat (especially that of the Sun) into that fluid
and subtle Etherial Body that surrounds the Earth ; and by
the restless agitation of that Celestial Matter wherein those
Particles swim, are so whirl'd round that each Corpuscle
endeavours to beat off all others from coming within the
little Sphear requisite to its motion about its own Center ;
and (in case any by intruding into that Sphear shall oppose
its free Rotation) to expel or drive it away : So that accord-
ing to this Doctrine, it imports very little, whether the
particles of the Air have the structure requisite to Springs,
or be of any other form (how irregular soever) since their
Elastical power is not made to depend upon their shape or
structure, but upon the vehement agitation, and (as it were)
brandishing motion, which they receive from the fluid
Ether that swiftly flows between them, and whirling about
each of them (independently from the rest) not only keeps

Y 2

324 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

those slender Aerial Bodies separated and stretcht out (at
least, as far as the Neighbouring ones will permit) which
otherwise, by reason of their flexibleness and weight, would
flag or curl ; but also makes them hit against, and knock
away each other, and consequently require more room,
then that which if they were compress'd, they would take up,
" By these two differing ways, . . . , may the Spring of the Air
be explicated. But though the former of them be that, which
by reason of its seeming somewhat more easie, I shall for
the most part make use of in the following Discourse : yet
am I not willing to declare peremptorily for either of them,
against the other (" New Experiments touching the Spring
of the Air," 1660, pp. 22-26;compare Works, 1725, II. 410).

The second explanation, which Boyle borrowed from
Descartes, resembled somewhat closely that which is gener-
ally adopted at the present day under the name of the
KINETIC THEORY OF GASES. According to this theory, gases
consist of elastic particles, moving with a high velocity, and
by their incessant collisions producing a constant pressure
upon each other and upon the walls of the containing
vessel. Their velocity is an expression of the heat energy
which the gas possesses, and increases with the temperature
in such a way that the temperature of a gas may be
measured by the pressure which it produces in a vessel of
given volume.

Boyle (1662) measures the condensation and rarefaction
of the air, Boyle's law.

In a " Defence of the Doctrine touching the Spring and
Weight of the Air .... against the Objections of Francis-
cus Linus," published in 1662, Boyle described "two new
Experiments touching the measure of the Force of the
Spring of Air compressed and dilated" (Defence, pp. 57 68;
Works, 1725, II. 670).

In the first experiment, air was compressed by means of
a column of mercury into one limb of a long glass tube,
shaped like an inverted syphon (Fig. 46). The column of

xv THE MOLECULAR THEORY 325

air was gradually compressed from 1 2 to 3 inches, and the
pressures (including the atmospheric pressure of 29T\
inches) were measured for twenty-five different volumes of
the air. These pressures were found to agree very closely
with " what that pressure should be according to the Hypo-
thesis, that supposes the pressures and expansions to be in
reciprocal proportion."

In the second experiment, on the debilitated force of ex-
panded air, a slender glass tube was immersed in mercury
and closed at the top with sealing wax, in such a way as to
enclose a column of air an inch long. The tube was raised
gradually until the air had expanded to 32 inches and the
position of the mercury was read for a series of nineteen
different volumes. Once again the pressures were found to
agree closely with those calculated from the Hypothesis.

FIG. 46— TUBE USED IN BOYLE'S EXPERIMENTS ON THE CONDENSATION
OF THE AIR.

Boyle's two tables are reproduced on p. 326, where all
the numbers represent inches.

The Hypothesis, now known as BOYLE'S LAW, states that
if the temperature is constant the volume of a gas is
inversely proportional to its pressure. Although applied
at first only to air, it was found to be approximately
true for all other gases and vapours. But in no case is
it accurately true ; thus in the case of both hydrogen
and nitrogen the volume under a pressure of 300 atmos-
pheres is about 25 per cent, greater than that calculated
from Boyle's law.

Gay-Lussac (1802) measures the expansion of gases by
heat. — The fact that air is expanded by heat was well known
in the time of Boyle (see, for instance, Boyle's Works, II.

326 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP, x

A TABLE OF THE CONDENSATION
OF THE AIR.

12

»4
ii

ioj

10

94
9

8
7*

7

6

si

54
Si

5l

44

4i

4

34
3i
3

00

02lf

06 i

I2A

29U

3 2 A

7iA
781S
88ft

D

37
39 r%

47TC

58}!
6iA
64A

37tt | ?

45^
48f| i
53H !

77H

93A

"7A

43H

5°°o
58|?

66^

70

73H

77!

87f
99?

A TABLE OF THE RAREFACTION
OF THE AIR.

i
Ii

2

3

4

6

BCD E

291

9t

1

31

2f

n
It

I4i

9tl

4lf

4i

in
5 X

A. The number of equal spaces in
the shorter leg, that contained the same
parcel of Air diversely extended.

B. The height of the Mercurial
Cylinder in the longer leg, that com-
pressed the Air into those dimensions.

C. The height of a Mercurial Cylinder
that counterbalanced the pressure of the
Atmosphere.

D. The Aggregate of the two last
columns B and C, exhibiting the pressure
sustained by the included Air.

E. What that pressure should be
according to the Hypothesis, that sup-
poses the pressures and expansions to be
in reciprocal proportion.

A. The number of equal spaces at the
top of the Tube, that contained the same
parcel of Air.

B. The height of the Mercurial Cylin-
der, that together with the Spring of the
included Air counterbalanced the pres-
sure of the Atmosphere.

C. The pressure of the Atmosphere.

D. The complement of B to C, ex-
hibiting the pressure sustained by the
included Air.

E. What that pressure should be acr
cording to the Hypothesis.

xv THE MOLECULAR THEORY 327

671). During the next hundred years many attempts were
made to measure this expansion ; but the results were very
irregular, mainly because the air, and the vessels in which it
was contained, were not sufficiently dried.

Still greater irregularities were observed when Priestley, in
1777 ("Experiments on Air," III. 345-348), measured the ex-
pansion by heat of different gases and found that

" The different kinds of air are expanded by the addition
of ten degrees of heat according to Fahrenheit's thermo-
meter, in the following proportion.

Common air I '32

Inflammable air 2 '05

Nitrous air 2'O2

Fixed air 2*20

Marine acid air I '33

Dephlogisticated air 2*21

Phlogisticated air 1-65

Vitriolic acid air 2*37

Fluor acid air 2 '83

Alkaline air 475

(Experiments and Observations, V. 359).

The fact "that oxygen, azote, hydrogen and carbonic
acid, and atmospheric air expand equally between o° and
80° " was discovered by Charles about the year 1787, but
his results were never published (Gay-Lussac, Ann. de
Chimie, 1802, 43. 157). The more detailed experiments
" On the dilatation of gases and vapours," which were
described by Gay-Lussac in 1802 (loc. at. 137-175), were
made with the apparatus shown in Fig. 47, or with a modified
apparatus in which the iron tap was replaced by a mercury
valve. The expansion between the freezing-point and the
boiling-point of water was measured by heating the gas to the
temperature of boiling water, opening the tap to allow the
excess of gas to escape, then cooling in ice and weighing the
water that was drawn in when the tap was again opened under
water. Six experiments with common air showed that

328 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

" From the temperature of melting ice to that of boiling
water, equal volumes of atmospheric air represented by 100

became 137-40, 137*61, 137*44, i37'55> i37'48, i37'57» of
which the mean is nearly 137 -50" (loc. tit. p. 165).

FIG. 47. — GAY-LUSSAC'S APPARATUS FOR MEASURING THE EXPANSION OF
GASES AND VAPOURS.

The globe B is provided with a tap worked by a lever LL. After filling it with
gas, the globe is fixed in the iron cage EFGH and immersed in a copper vessel AD
full of water. The curved tube ID leads to a trap KX filled with mercury.

The increase of volume for four different gases was shown
(loc. cit. p. 167) in a table as follows : —

Expansion. Differences.
Atmospheric air .... 37 '50

Hydrogen 37 -52 +cro2

Oxygen 37 -48 -0-02

Azote 37 '49 -0*01

THE MOLECULAR THEORY

329

In the case of five gases which were soluble in water com-
parative experiments were made with tubes of carefully-dried
gas inverted over mercury (Fig. 48), the expansion observed
being identical in every case with that of atmospheric air.
Gay-Lussac summarises his conclusions as follows :

" The experiments which I have just described and which
were all made with great care, prove incontestably that
atmospheric air and oxygen, hydrogen, azote, nitrous gas,
ammonia, muriatic gas, sulphurous acid, and carbonic acid
are expanded equally by
the same degrees of heat ;
and that, consequently, their
greater or less density . . . ,
their greater or less solu-
bility in water and their
particular nature, have no
influence on their expan-
sion."

" From this consideration
I conclude that all gases,
in general, are expanded
equally by the same degrees
of heat ; provided that they
are all placed under the
same conditions " (loc. cit.
p. 172).

'••-•

- i

"

- T

FIG. 48.— GAY-LUSSAC'S APPARATUS
FOR COMPARING THE EXPANSION OF AIR
WITH THAT .OF SOLUBLE GASES AND
VAPOURS.

The gases were confined over mercury
in two tubes TT standing in a trough of
mercury AC.

A further series of experi-
ments on the expansion of
the vapour of ether between 60° and ioo°C. (loc. cit. p. 173)
carried out in the presence of Berthollet, showed that the
law of equal expansions could be applied to vapours as well
as to gases ; Gay-Lussac therefore concluded that they
would probably be equally compressible, i.e. that the simple
hypothesis of Boyle could be applied to vapours, so long as
they remained uncondensed. All gases and all vapours would
then be influenced in the same way by changes of pressure

330 HISTORICAL INTRODUCTION TO CHEMISTRY CHAII

and of temperature — a remarkable conclusion that has no
parallel in the behaviour of liquids or solids.

B. THE COMBINING VOLUMES OF GASES

Cavendish (1781) measures the combining volumes of
hydrogen and oxygen. — In his experiments on the com-
position of water Cavendish found that 423 measures of
hydrogen were just sufficient to remove the oxygen from
1000 measures of air. If we assume that 20-9% of the
atmosphere consists of oxygen, this would mean that 209
measures of oxygen combine with 423 measures of hydrogen
or 100 measures of oxygen with 202 measures of hydrogen.
He also found that a mixture of 19500 grain measures of
oxygen and 37000 of inflammable air, after a series of
explosions, left a residue of 2950 grain measures of gas, of
which about 1000 consisted of unburnt oxygen. If it be
assumed that the remaining 1950 grain measures consisted
of impurities, shared between hydrogen and oxygen in pro-
portion to their original volumes, we find that 19500 - 1000
-650=17850 measures of oxygen combined with 37000
[ - 1300 = 35700 measures of hydrogen; that is, one volume
of oxygen combined exactly with two volumes of hydrogen.

Gay-Lussac and Humboldt (1805) measure the
combining volumes of hydrogen and oxygen. — In the
year 1805, Gay-Lussac and Humboldt (Jour, de Physique^
1805, 60, 129) attempted to determine the "Ratio of the
Constituents of the Atmosphere" by exploding air with
hydrogen in a Volta's eudiometer and measuring the
diminution of volume which resulted. In order to
interpret their observations they found it necessary to make
a fresh determination of the ratio by volumes in which
hydrogen and oxygen unite. The most trustworthy series
of experiments was one in which 100 parts .of oxygen (of
which 99*6 parts could be absorbed by potassium sulphide)

xv THE MOLECULAR THEORY 331

were exploded with 300 parts of hydrogen and produced a
diminution of 2987 parts by volume. In these experiments
99*6 parts of oxygen united with 298*7-99'6 = 199*1 parts of
hydrogen, or TOO parts of oxygen united with 199*89 parts
of hydrogen, a number that did not differ appreciably from
(that required for the exact integral ratio i : 2 (loc. tit. pp.
146—147).

Gay-Lussac (1809) states that gases combine in simple
proportions by volume. — (i) The combination of ammonia
with acids. This simple relationship between the combining
volumes of hydrogen and oxygen was in such striking
contrast with the complex ratio of the combining weights
that Gay-Lussac in 1809 (A.C.R. IV. 8-24) preceded to
examine the ratios of the combining volumes of other gases.
As the simplest case he studied first the combination of
ammonia with acid gases, including muriatic acid gas and
carbonic anhydride. He found that :

" 100 parts of muriatic gas saturate precisely 100 parts of
ammonia gas, and the salt which is formed from them is
perfectly neutral, whether one or other of the gases is in
excess." " If carbonic gas is brought into contact with
ammonia gas, by passing it sometimes first, sometimes
second into the tube, there is always formed a sub-carbonate
composed of 100 parts of carbonic gas, and 200 of ammonia
gas" (A.CR. IV. 10).

(2) Composition by volume of ammonia, sulphuric acid and
carbonic anhydride. Equally simple relationships were
found to govern the proportions of nitrogen and hydrogen
in ammonia gas, of " sulphurous gas " and oxygen in
sulphuric anhydride, and of the carbonic oxide and oxygen
in carbonic anhydride.

" According to the experiments of M. Ame'dee Berthollet,
ammonia is composed of:

100 of nitrogen,
300 of hydrogen
by volume.

332 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

"I have found (ist vol. of the Societe d'Arcueil) that
sulphuric acid is composed of :

100 of sulphurous gas,
50 of oxygen gas.

" When a mixture of 50 parts of oxygen and 100 of
carbonic oxide (formed by the distillation of oxide of zinc
with strongly calcined charcoal) is inflamed, these two gases
are destroyed, and their place taken by 100 parts of carbonic
acid gas. Consequently carbonic acid may be considered as
being composed of :

100 of carbonic oxide gas,

50 of oxygen gas." (A.C.R. IV. 13).

(3) The oxides of nitrogen. Gay-Lussac next calculated
from Davy's analyses the volumes of nitrogen and oxygen
which would be obtained from the oxides of nitrogen, if
these could be decomposed into their elements, and showed
that the volume ratios were again of the very simplest.

" Davy, from the analysis of various compounds of
nitrogen and oxygen, has found the following proportions
by weight : —

Nitrogen. Oxygen.

Nitrous oxide . . . . 63 '30 3670

Nitrous gas .... 44/05 55'95

[Nitrogen peroxide] . . . 29-50 70-50

Reducing these proportions to volumes, we find :-

Nitrogen. Oxygen.

Nitrous oxide .... 100 49-5

Nitrous gas 100 108*9

[Nitrogen peroxide] . . . 100 204-7

The first and last of these proportions differ only slightly
from 100 to 50, and 100 to 200 ; it is only the second
which diverges somewhat from TOO to 100. The difference,
however, is not very great, and is such as we might expect
in experiments of this sort ; and I have assured myself that it is
actually nil. On burning the new combustible substance
from potash in 100 parts by volume of nitrous gas, there

xv THE MOLECULAR THEORY 333

remained over exactly 50 parts of nitrogen, the weight of
which, deducted from that of the nitrous gas (determined
with great care by M. Be'rard at Arcueil), yields as result
that this gas is composed of equal parts by volume of
nitrogen and oxygen.

We may then admit the following numbers for the pro-
portions by volume of the compounds of nitrogen with
oxygen.

Nitrogen. Oxygen.

Nitrous oxide 100 50

Nitrous gas .... 100 100

[Nitrogen peroxide] . . . 100 200"

(A.C.R. IV. 13-14.)

From the observations recorded above Gay-Lussac
concluded that "gases always combine in the simplest
proportions when they acton one another" (A.C.R. IV. 15).

Extension of the law of volumes to the products
formed by the interaction of gases. — Further experiments
convinced Gay-Lussac that :

" Not only, however, do gases combine in very simple
proportions, as we have just seen, but the apparent con-
traction of volume which they experience on combination
has also a simple relation to the volume of the gases, or at
least to that of one of them."

"I have said, following M. Berthollet, that 100 parts of
carbonic oxide gas, prepared by distilling oxide of zinc and
strongly calcined charcoal, produce 100 parts of carbonic
gas on combining with 50 of oxygen. It follows from this
that the apparent contraction of the two gases is precisely
equal to the volume of oxygen gas added."

" We know, besides, that a given volume of oxygen
produces an equal volume of carbonic acid ; consequently
oxygen gas doubles its volume on forming carbonic oxide
gas with carbon, and so does carbonic gas on being passed
over red-hot charcoal."

" Oxygen gas, in combining with sulphur to form
sulphurous gas, only experiences a diminution of a fiftieth of
its volume, and this would probably be nil if the data I have
employed were more exact."

334 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

" Ammonia gas is composed of three parts by volume of
hydrogen, and one of nitrogen, and its density compared to
the air is 0*596. But if we suppose the apparent contrac-
tion to be half of the whole volume, we find 0-594 for the
density. Thus it is proved, by this almost perfect
concordance, that the apparent contraction of its elements
is precisely half the total volume, or rather double the
volume of the nitrogen.

" We see, then, from these various examples, that the
contraction experienced by two gases on combination is in
almost exact relation with their volume, or rather with the
volume of one of them " (A.C.R. IV. 15-20).

Gay-Lussac's law of volumes. — Gay-Lussac summarises
his conclusions in the following words :

" I have shown in this memoir that the compounds of
gaseous substances with each other are always formed in very
simple ratios, so that representing one" of the terms by
unity, the other is i, or 2, or at most 3. These ratios
by volume are not observed with solid or liquid substances,
nor when we consider weights, and they form a new proof
that it is only in the gaseous state that substances are in the
same circumstances and obey regular laws. . . . The apparent
contraction of volume suffered by gases on combination is
also very simply related to the volume of one of them, and
this property likewise is peculiar to gaseous substances "
(A.C.R. IV. 24).

In order to include all the cases discussed in the memoir,
Gay-Lussac's law of volumes may be stated as follows :

When gases enter or leave chemical combination, the volumes
absorbed, or liberated, are in simple ratios to one another.

Modification of Gay-Lussac's law. — Later experiments
have shown that Gay-Lussac's law does not hold exactly.
Thus, Scott (Phil. Trans., 1893, 184, 566), by exploding
together gases of remarkable purity, found that hydrogen
and oxygen combine together at ordinary temperatures in
the ratio 2*00245 to i. This deviation from Gay-Lussac's
law is confirmed by the work of Morley (p. 131), who found

xv THE MOLECULAR THEORY 335

the ratio at o°C to be 2-00269. *n tne same way it has
been shown that the volume of hydrogen contained in two
volumes of hydrogen chloride is not i, but 1*00790 (Gray
and Burt, Trans. Chem. Soc., 1909, 95, 1656), whilst the
combining ratio of hydrogen and nitrogen in ammonia is
3-00172 (Guye and Pintza, Comptes rendus, 1908, 147, 928).

These deviations from exact whole numbers probably
depend on the fact that gases differ slightly in their com-
pressibility, so that equal volumes under a pressure of i
atmosphere would become unequal under a pressure of 2
atmospheres or \ atmosphere. It is, however, believed
that Gay-Lussac's law would" hold good accurately if the
gases could be examined under very low pressures.

Gay-Lussac's law of volumes explained by the atomic
theory. — Gay-Lussac's observations provided excellent ex-
amples of the laws of chemical combination described in
the preceding chapter. Thus, no better illustration could
be desired of the law of " Multiple Proportions " than the
three oxides of nitrogen, the two oxides of carbon and the
two oxides of sulphur, in which the proportions of oxygen
to the other element were shown by Gay-Lussac to be in
the ratios i : 2 : 4, i : 2 and 2 : 3 respectively. It is there-
fore not surprising that Gay-Lussac was ready to interpret
his observations by means of the atomic theory which
Dalton had suggested a few years previously.

" According to Dalton's ingenious idea, that combinations
are formed from atom to atom, the various compounds
which two substances can form would be produced by the
union of one molecule of the one with one molecule of the
other, or with two, or with a greater number, but always
without intermediate compounds."

"The numerous results I have brought forward in this
Memoir are also very favourable to the theory" (A.C.R.
IV. 23).

The law of volumes rejected by Dalton. — It is, however,
remarkable that Dalton, instead of welcoming the Law of

336 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

Volumes as giving support to his theory, refused to recognise
its validity, and even proceeded to question the evidence on
which it was based. The argument which influenced him
so strongly is summarised in the following paragraphs.

The simplest way of applying the atomic theory to gases
was to assume that the space occupied by the atom (or
molecule) was the same for all gases. This hypothesis was
worked out by Dalton in the first volume of his "New
System," published in 1808, where he writes :

"At the time I formed the theory of mixed gases, I had
a confused idea, as many have, I suppose, at this time, that
the particles of elastic fluids are all of the same size ; that a
given volume of oxygenous gas contains just as many
particles as the same volume of hydrogenous ; or if not,
that we had no data from which the question could be
solved " (A.C.R. IV. 6-7).

" By the size or volume of an ultimate particle, I mean
in this place, .the space it occupies in the state of a pure
elastic fluid" (A. C.R. IV. 6).

But the logical consequences of the hypothesis were such
that Dalton " became convinced that different gases have
not their particles of the same size," and concluded :

" That every species of pure elastic fluid has its particles
globular and all of a size; but that no two species agree in the
size of their particles, the pressure and temperature being
the same " (A.C.R. IV. 7).

This conclusion was reached as the result of an argument
which showed that, if the atoms or molecules of all gases
occupied the same volume, there must be a decrease of 'volume
and an increase of density, whenever the atoms of two simple
gases united together to form the molecules of a compound
gas. But in the case of nitric oxide (nitrous air), Davy's,
observations showed :

(i) That the volume of the gas was equal to that of its

xv THE MOLECULAR THEORY 337

constituents, which did not therefore contract in combina-
tion, and

(2) That whilst the density of the gas was a little greater
than that of nitrogen it was actually less than that of the
oxygen which formed one of its constituents. Davy's actual
figures ( Works, III. 9-10) were :

Nitrogen 3°'45 grains per 100 cubic inches.
Nitric oxide 34*3 „ „ „ „ „
Oxygen 35*09 „ „ „ „ „

Dalton argues as follows :

" If equal measures of azotic and oxygenous gases were
mixed, and could be instantly united chemically, they would
form nearly two measures of nitrous gas, having the same
weight as the two original measures ; but the number of
ultimate particles could at most be one half of that before
the union. No two elastic fluids, probably, therefore, have
the same number of particles, either in the same volume or
the same weight" (A.C.R. IV. 5).

When Gay-Lussac, in 1809, put forward his Law of Vol-
umes, the simple hypothesis to which it seemed to point had
therefore already been tested by Dalton, and found to be
untenable. It is not surprising then that Dalton should
become its foremost opponent, his repudiation extending
even to a refusal to accept the experimental facts to which
Gay-Lussac had called attention. In the second part of
his "New System." published in 1810, he sums up his
criticism as follows :

" The truth is, I believe, that gases do not unite in equal
or exact measures in any one instance ; when they appear
to do so, it is owing to the inaccuracy of our experiments.
In no case, perhaps, is there a nearer approach to mathe-
matical exactness, than in that of i measure of oxygen to
2 of hydrogen ; but here, the most exact experiments I

z

338 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

have ever made, gave 1*97 hydrogen to i oxygen " (A.C.R.
IV. 27).

C. AVOGADRO'S HYPOTHESIS

Avogadro (1811) revives Dalton's discarded hypo-
thesis.— The suggestion as to the spacing of gaseous atoms,
which Dalton had considered and rejected in 1808, was
revived three years later, in 1811, by Amadeo Avogadro,
(1776-1856) Professor of Physics at Turin, as the most
obvious, perhaps even the only possible, method of
explaining Gay-Lussac's Law of Volumes. AVOGADRO'S
HYPOTHESIS, as it is now generally called, states that :

" When their temperatures and pressures are equal, equal
volumes of different gases contain equal numbers of
molecules"

Later researches have proved the substantial truth, as
well as the usefulness of this hypothesis. But it is now
recognised that (like Gay-Lussac's Law of Volumes) it is
strictly accurate only when applied to gases under very low
pressures ; if applied to gases at atmospheric pressures, a
correction must be made for their unequal compressi-
bilities.

Avogadro postulates complex molecules in gaseous
elements. — The merit of Avogadro's paper on the " Relative
Masses of Elementary Molecules," consists in the fact that
he was able to find a remedy for the defects which
had proved fatal to the hypothesis as discussed and
condemned by Dalton. Avogadro suggested that all
difficulties would disappear if the molecules of an elementary
gas consisted of groups of atoms, similar to those which
formed the molecules of a compound gas.

If the molecules of an element were indivisible they
could never give rise to more than an equal number of

xv THE MOLECULAR THEORY 339

molecules of a compound ; a gaseous element could
therefore never give rise to more than its own volume of
a gaseous compound. But Gay-Lussac had shown that
" the volume of water in the gaseous state is ... twice
as great as the volume of oxygen which enters into it,"
i.e. i volume of oxygen produces 2 volumes of steam,
or (on the theory of equal spacing) i molecule of oxygen
produces 2 molecules of steam.

" But a means of explaining facts of this type in con-
formity with our hypothesis presents itself naturally enough :
we suppose namely that the constituent molecules of any
simple gas . . . are not formed of a solitary elementary [atom],
but are made up of a certain number of these [atoms]
united by attraction to form a single [molecule] " (A.C.R.
IV. 31).

When the element enters into combination, this molecule
may be split into two or more parts, giving rise to two or more
molecules of a compound, which would thus occupy a volume
two or more times as great as that of the element.

Avogadro pointed out that the molecule of an element does
not usually split into more than two parts, although
rightly enough he asserts the possibility of further sub-
division.

" On reviewing the various compound gases most
generally known, I only find examples of duplication of the
volume relatively to the volume of that one of the con-
stituents which combines with one or more volumes of the
other. We have already seen this for water. In the same
way, we know that the volume of ammonia gas is twice
that of the nitrogen which enters into it. M. Gay-Lussac
has also shown that the volume of nitrous oxide is equal to
that of the nitrogen which forms part of it, and consequently
twice that of the oxygen. Finally, nitrous gas, which contains
equal volumes of nitrogen and oxygen, has a volume equal
to the sum of the two constituent gases, that is to say,

z 2

340 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

double that of each of them. Thus in all these cases there
must be a division of the molecule into two ; but it is
possible that in other cases the division might be into four,
eight, etc." (A.C.R. IV. 32).

Molecules which can be divided into four atoms have
since been found in the vapours of phosphorus and arsenic,
whilst the vapour of boiling sulphur contains molecules
which are divisible into eight atoms, exactly as suggested
by Avogadro in the preceding paragraph.

Molecular weights determined by means of Avogadro's
hypothesis. — Avogadro pointed out that, on the theory of
equal spacing, the density of a gas is proportional to the
weight of its molecules.

.

" Setting out from this hypothesis, it is apparent that we
have the means of determining very easily the relative
masses of the .molecules of substances obtainable in the
gaseous state ... for the ratios of the masses of the
molecules are then the same as those of the densities of the
different gases at equal temperature and pressure" ( A.C.R.
IV. 30).

This argument may be illustrated by the following
example. Suppose that the weights of i litre of oxygen
and hydrogen are i -42900 and 0*089873 gram respectively, at
o°C. and 760 mm. ; and that there are n molecules in i litre
of either gas. Then :

i '42900 grams Weight of i litre of oxygen,

0*089873 gram Weight of i litre of hydrogen.

Weight of n molecules of oxygen,
Weight of n molecules of hydrogen.

Weight of i molecule of oxygen,
Weight of i molecule of hydrogen

xv THE MOLECULAR THEORY 341

For the purposes of the argument it is not necessary to
know the actual number of molecules in the litre of gas,
but recent experiments have shown that the number n has
the value 3 x io2'2, the probable error in this number being
only about io per cent. The " Avogadro constant," A",
which represents the number of atoms in the gram-atom or
of molecules in the gram-molecule, e.g. oxygen atoms in
1 6 grams of oxygen, or oxygen molecules in 32 grams of
oxygen, has been found to be 67 x io'22, ten different
methods of measurement having given values lying between
62 and 72 x io22.

Complexity of gaseous molecules determined by means of
Avogadro's hypothesis, — The fundamental weakness of
Dalton's Atomic Theory arose from the fact that he had no
way of finding out how many atoms were present in the
naolecule of a given compound. He always chose .the
simplest formulae that were possible ; but this procedure
was quite arbitrary, and it was not possible to say whether
the conclusions arrived at were correct or not. Avogadro's
hypothesis had the merit of supplying a method by which
the relative weights of the atoms, and their relative numbers
in any given compound, could be determined from the
properties of the substances themselves, instead of from the
arbitrary whim of the investigator. The following examples
were selected as illustrations of the new method :

Water. — Dalton had supposed that water was formed by
the union of one atom of hydrogen with one atom of
oxygen. Avogadro recognising that :.

2 volumes of hydrogen + i volume of oxygen yield 2
volumes of steam

and that, therefore :

2 molecules of hydrogen + i molecule of oxygen yield 2
molecules of steam,

342 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

regarded the molecule of water as " composed of a half-
molecule of oxygen with one molecule, or, what is the
same thing, two half-molecules of hydrogen" (A.C.R. IV.
footnote, p. 32).

Nitric Oxide and Ammonia — In the same way Avogadro
was able to show that since :

1 volume of oxygen and i volume of nitrogen are
present in 2 volumes of nitrous gas

the molecule of nitrous gas is " composed of a half-
molecule of oxygen and a half-molecule of nitrogen "
(A.C.R. IV. 36). So too in the case of ammonia,
Berthollet's observation that :

2 volumes of ammonia are decomposed by sparking into
i volume of nitrogen and 3 volumes of hydrogen

led Avogadro to conclude that the molecule of ammonia
is composed of one half-molecule of nitrogen combined with
three half-molecules of hydrogen.

Muriatic Acid. — Another case quoted by Avogadro is
that of gaseous muriatic acid.

" It follows from the experiments both of Gay-Lussac and
Thenard, and of Davy, that muriatic acid gas is formed by
the combination of equal volumes of [chlorine] and
hydrogen, and that its volume is equal to their sum. This
means, according to our hypothesis, that muriatic acid is
formed of these two substances united molecule to molecule,
with halving of the molecule, of which we have had so
many examples." (A.C.R. IV. 44.)

Here since :

i volume of hydrogen and i volume of chlorine give 2
volumes of hydrogen chloride,

the molecule of hydrogen chloride is composed of one
half-molecule of hydrogen and one half-molecule of chlorine.

xv THE MOLECULAR THEORY 343

In the case of each of the gaseous elements (hydrogen,
oxygen, nitrogen and chlorine) discussed by Avogadro in
this section of his memoir, the molecule was found to be
divisible into two parts, but no further division was observed.
The indivisible atom was therefore probably the half-molecule
in the case of each of these gases. It should be observed,
that the facts prove no more than that the molecule can be
divided into some even number of atoms. In a later section
further reasons will be given in support of the belief that in
each of these cases not more than two atoms are contained
in the molecules ; but other cases are known in which the
molecules of a gas can be divided into 3, 4, 6 or 8 atoms.

A vogadro's hypothesis neglected for nearly fifty years. —
It was unfortunate that Avogadro, having found a trustworthy
method of determining the relative weights of gaseous
molecules did not limit his deductions to those volatile
compounds and elements to which alone his hypothesis
could be applied. In attempting to extend his views to
non-volatile compounds, such as the oxides and salts of the
metals, Avogadro was compelled, like Dalton, to fall back
upon mere guess-work. His speculations failed to secure
the support of his contemporaries and may well have pre-
judiced them against the accurate conclusions which he
reached in the case of gaseous elements. But, whatever
'the cause, the fact remains that nearly half a century elapsed
before A vogadro's views, re-stated and applied, in 1858, by
Stanislao Cannizzaro (1826—1910), Professor of Chemistry
at Genoa (A.C.R. XVIIL), finally secured the universal
recognition that they deserved.

D. DETERMINATION OF MOLECULAR AND ATOMIC
WEIGHTS.

Cannizzaro (1858) chooses the half -molecule of hydrogen
as a standard of molecular and atomic weights. — In

344 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

measuring the density of a gas or vapour it is often conve-
nient to calculate the density relatively to some selected gas
as standard, rather than to give the actual density in grams
per litre. Thus whilst the actual densities of air, oxygen,
hydrogen and nitrogen may be compared at o°C. and 760
mm. pressure, vapours such as steam cannot be included
in the comparison because they would condense to liquids
under the conditions selected. But as all gases expand and
contract in much the same way (see Section A of this
chapter), their relative densities are almost independent of
temperature and pressure. A table of relative densities may
therefore include vapours as well as gases ; e.g. steam may
be compared with air at 130°, whilst hydrogen is com-
pared with air at 20° or at o°. Vapour densities, and
gas densities, are often given relatively to air ; but hydrogen,
as being the lightest known gas, is usually preferred as a
standard, since the relative densities of all other gases can
then be represented by numbers greater than unity.

In selecting a standard of molecular weights, the hydrogen
molecule would provide a convenient unit, since the rela-
tive weights of gaseous molecules could then be represented
by the same number as their relative densities. But
since the hydrogen molecule is known to contain two atoms,
its selection as a standard would necessitate the use of a
fraction 0*5 to represent the hydrogen atom ; Cannizzaro
therefore selected the half-molecule or atom of hydrogen as
a unit both of molecular and of atomic weights. Cannizzaro
writes (A.C.R. XVIII. 6-7):

" I prefer to take as common unit for the weights of the
molecules and for their fractions, the weight of a half and
not of a whole molecule of hydrogen, I therefore refer the
densities of the various gaseous bodies to that of hydrogen = 2.
If the densities are referred to air = i, it is sufficient to
multiply by 14*438 to change them to those referred to that
of hydrogen = i ; and by 28*87 to refer them to the density
of hydrogen = 2.

XV

THE MOLECULAR THEORY

345

" I write the two series of numbers, expressing these
weights in the following manner :

Densities or weights

of one volume, the

Densities referred to

volume of Hydrogen

that of Hydrogen =2,

Names of
Substances.

being made=i, i.e.
weights of the mole-
cules referred to the

i.e. weights of the
molecules referred to
the weight of half a

weight of a whole

molecule of Hydro-

molecule of Hydro-

gen taken as unity.

gen taken as unity.

Hydrogen

i

2

Oxygen, ordinary ...

16

32

Oxygen, electrised ...

64

128

Sulphur below 1000°.

96

192

Sulphur above 1000°

32

64

Chlorine

35*5

71

Bromine

80

1 60

Arsenic

150

300

Mercury

100

200

Water

9

18

Hydrochloric Acid . . .
Acetic Acid

18-25
30

36-50
60"

The table shows amongst other things, " that the same
substance in its different allotropic forms can have different
molecular weights," but the data for " electrised oxygen "
are incorrect, the actual densities of ozone being 24 and 48
instead of 64 and 128.

Oxygen as a standard of molecular and atomic
weights. — In Cannizzaro's table, the hydrogen atom = i
and the hydrogen molecule = 2. The relative density of
oxygen is given as 16 and its molecular weight as 32.
Later experiments have shown that the exact figures are
i5'88 and 3176.

In recent years, in accordance with a suggestion made by
Stas, the half-molecule or atom of oxygen has been adopted
as a standard of molecular and atomic weights, its weight

346 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

being taken as exactly 16, instead of 15*88. The atomic
weight of hydrogen then becomes roo8 instead of i and its
molecular weight 2-016 instead of 2 ; all other molecular
and atomic weights are increased in the same proportion.

The choice of oxygen as a standard is justified by the fact
that, in practice, molecular and atomic weights are nearly
always measured by comparison with oxygen rather than
with hydrogen. The actual results of this comparison can
now be tabulated, whereas formerly the accuracy of the
tables was greatly impaired in reducing the numbers from
the oxygen to the hydrogen standard by means of a factor
deduced from the composition of water.

Avogadro's hypothesis only approximately true.—
Cannizzaro's method of deducing molecular weights has also
been modified by the discovery that Avogadro's hypothesis
is only approximately true. Like Gay-Lussac's " Law of
Volumes," it would probably be exact if applied to gases
under very low pressures. Several methods have been used
to eliminate the effects of unequal compressibility on the
relative densities of gases. The following table shows the
nature of the correction, and the large magnitude which it
attains in the case of easily-liquefied gases.

Molecular

Gas.

Ratio of
densities.

Corrected
Molecular
Weights.

Weight
deduced
from Equi-

valents.

Hydrogen

2*0125

2*0150

2 "OI 52

Nitrogen
Carbonic oxide ,

28-007
28-001

28-013
28-003

28-02
28-00

Oxygen ...

32

32

32

Carbonic anhydride

44-267

44-003

44-00

Nitrous oxide

44-284

44-000

44-02

Hydrogen chloride
Sulphurous anhydride

36741

36-484
64. • c6 c

THE MOLECULAR THEORY

347

Cannizzaro's method of determining atomic weights. —

The method used by Cannizzaro in determining the atomic
weight of an element depends on setting out in a table all
those volatile compounds the vapour density and percentage
composition of which are known. From the vapour density
the molecular weight is calculated, whilst the percentage
composition gives the weights of the different constituents
in the molecular weight of the compound.

The substance of Cannizzaro's table1 is set out in the
first three columns of the following series of tables : —

i. HYDROGEN COMPOUNDS. H=i.

Molecular

weight

Name of the

referred
to the

Weights of the constituents
of one molecule, all referred

A

0

Substance.

weight of
a half-

to a half-molecule of Hydro-

£

o

molecule

gen = i.

fe

of Hydro-

gen = i.

Hydrogen

2

2 Hydrogen .

H2

Hydrogen chloride

36-5

I Hydrogen, 35*5 Chlorine

HC1

Hydrogen bromide

i Hydrogen, 80 Bromide

HBr

Hydrogen iodide. . .

128

i Hydrogen, 127 Iodine

HI

Water- vapour ...

18

2 Hydrogen, 16 Oxygen

H2O

Sulphuretted

hydrogen 34

2 Hydrogen, 32 Sulphur

HoS

Ammonia

17

3 Hydrogen, 14 Nitrogen

HjN

Arsine

78

3 Hydrogen, 75 Arsenic

H3As

Phosphine

34

3 Hydrogen, 31 Phosphorus

H3P

1 Cannizzaro gives a general table and a series of special tables for
the individual elements. In the following paragraphs the general table
has been subdivided, and a few additional substances, quoted in
Cannizzaro's special tables, have been added. The values for ozone
and for phosphine have been corrected.

348 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

It is seen that the quantity of hydrogen in the molecule is
always a half-molecule or some integral number of half-
molecules ; the half-molecule of hydrogen appears therefore
to represent the extreme limit to which the subdivision of
the molecule can be carried. Cannizzaro concludes that
"The atom of hydrogen is contained twice in the molecule
of free hydrogen" (A.C.R. XVIII. n).

2. OXYGEN COMPOUNDS. O=i6.

1

Mole-

Name.

cular

Composition of Molecule.

Formula

Weight.

Oxygen

32

32 Oxygen

o.

Ozone
Water vapour ...

48
18

48 Oxygen
1 6 Oxygen, 2 Hydrogen

03
OH2

Nitrous oxide

44

1 6 Oxygen, 28 Nitrogen

ON2

Nitric oxide

3°

1 6 Oxygen, 14 Nitrogen

ON

Carbonic oxide ...

28

1 6 Oxygen, 12 Carbon

OC

Carbonic anhy-

dride

44

32 Oxygen, 12 Carbon

O2C

Sulphurous anhy-

B

dride

64

32 Oxygen, 32 Sulphur

o.2s

(See also under Carbon. )

The smallest weight of oxygen in the molecule is 16.
This is the weight of the oxygen atom and is now taken as
the standard of comparison for all other atomic weights.
The molecule of oxygen gas has weight 32 ; it is therefore
a " diatomic " molecule containing two atoms, as represented
by the formula O2 ; but ozone, prepared from it by electri-
fication, is seen to be a " triatomic " modification, containing
three atoms in the molecule, as represented by the formula
(X

xv THE MOLECULAR THEORY

3. CHLORINE COMPOUNDS. Cl — 35*5.

349

Mole-

t

Name.

cular

Composition of Molecule.

Formula.

Weight.

Chlorine gas

71

7 1 Chlorine

Cl2

Hydrogen

chloride ...

36-5

35 '5 Chlorine, i Hydrogen

C1H

Mercurous

chloride ...

235-5

35 '5 Chlorine, 200 Mercury

ClHg1

Mercuric

chloride ...

271

71 Chlorine, 200 Mercury

Cl2Hg

Arsenic chlor-

ide

181-5

io6'5 Chlorine, 75 Arsenic

Cl3As

Phosphorus

chloride . . .

I38-5

io6'5 Chlorine, 32 Phosphorus

C13P

Ferric chlor-

ide

325

213 Chlorine, 112 Iron

Cl6Fr8

The smallest weight of chlorine in the table is 35*5.
This is therefore taken as the atomic weight of chlorine.
The weight of the molecule is 71; it therefore contains
(like the hydrogen molecule) 2 atoms, and is represented
by the formula C12.

4. NITROGEN COMPOUNDS. N = 14.

Mole-

Name.

cular

Composition of Molecule.

Formula.

Weight.

Nitrogen gas
Ammonia

28
17

28 Nitrogen
14 Nitrogen, 3 Hydrogen

N2
NH3

Nitrous oxide
Nitric oxide ...:

44
30

28 Nitrogen, 16 Oxygen
14 Nitrogen, 16 Oxygen

N0O
NO

The atomic weight of nitrogen is seen to be 14. The
gas is again "diatomic," containing two atoms in each
molecule, and is represented by the formula N2.
1 See later, p. 529.

350 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP
5. SULPHUR COMPOUNDS. 8 = 32.

Name.

Mole-
cular
Weight.

Composition of Molecule.

Formula.

Sulphur under 1000°
Sulphur above 1000°
Sulphuretted
hydrogen
Sulphurous
anhydride
Carbon sulphide ...

192
64

34

64
76

192 Sulphur
64 Sulphur

34 Sulphur, 2 Hydrogen

32 Sulphur, 32 Oxygen
64 Sulphur, 12 Carbon

if

SH2

SO2
S2C

The atomic weight of sulphur is seen to be 32. At
temperatures above 1000° the vapour has a molecular weight
64 ; the molecules of the gas are then diatomic like those
of hydrogen, chlorine, nitrogen and oxygen. But at lower

6. . CARBON COMPOUNDS. C = 1 2.

Name.

Mole-
cular
Weight.

Composition of Molecule.

Formula.

Carbonic oxide

28

12 Carbon, 16 Oxygen

CO

Carbonic anhydride

44

12 Carbon, 32 Oxygen

COo

Carbon bisulphide...

76

12 Carbon, 64 Sulphur

cs;

Marsh gas
Ethylene

16

28

12 Carbon, 4 Hydrogen
24 Carbon, 4 Hydrogen

CH4
C2H4

Propylene

42

36 Carbon, 6 Hydrogen

C3H6

Formic acid

46

12 Carbon, 2 Hydrogen,

CH202

32 Oxygen

Acetic acid

60

24 Carbon, 4 Hydrogen,

C2H402

32 Oxygen

Acetic anhydride ...

1 02

48 Carbon, 6 Hydrogen,

C4H603

48 Oxygen

Alcohol

46

24 Carbon, 6 Hydrogen,

C2H60

16 Oxygen

Ether

74

48 Carbon, 10 Hydrogen

C,HinO

/ T"

1 6 Oxygen

*-'41 10^

1 See footnote on p. 351.

THE MOLECULAR THEORY

351

temperatures the molecules POLYMERISE, i.e. collect together
into larger groups of atoms, the average number of atoms in
the molecule at the boiling point being about 6. l

The smallest weight of carbon in the table is 1 2 ; this is
therefore the atomic weight. In the case of carbon, Can-
nizzaro points out that " since we cannot determine the
vapour-density of free carbon, we have no means of knowing
the weight of its molecule, and thus we cannot know how
many times the atom is contained in it." But this does not
in the least affect our knowledge of the atomic weight of
the element, since " the knowledge of the weight of the
molecule of this last would merely add a datum more to
those which are already sufficient for the solution of the
problem" (A.C.R. XVIII. 14).

7. PHOSPHORUS, ARSENIC AND MERCURY. P = 3i, As
Hg= 200

75,

Mole-

Name.

cular

Composition of Molecule.

Formula.

Weight.

Phosphorus

124

124 Phosphorus

P4

Arsenic

300

300 Arsenic

As4

Mercury

200

200 Mercury

Hg

(See also under hydrogen and chlorine. )

The smallest quantities of these elements which are found
in the molecules of their compounds are 31, 75 and 200
respectively. The vapours of phosphorus and arsenic are
therefore composed of " tetratomic " molecules which are
subdivided into quarter-molecules by combination with
hydrogen or chlorine. The vapour of mercury on the other
hand is composed of " monatomic " molecules, which

1 Recent work has indicated that the vapour is a mixture of S8 and
S2, the existence of S6 being very doubtful.

352 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

enter as a whole into combination with other elements and
are not capable of being further. subdivided. "Diatomic"
molecules, although abundant amongst the commoner gases,
are by no means universal ; in a complete table of volatile
elements the diatomic molecules are seen to be less abun-
dant than the monatomic :

Monatomic Na, K, Zn, Cd, Hg, He, Ne, Ar, Xe, Kr, I
Diatomic H2, N2, O2, F2, C12, Br2, I2, S2

TriatomiC) 6°<r. O3, P4, As4, Ss.

Formulae of volatile compounds deduced by Cannizzaro.—

tlaving established a method of determining the atomic
weights of their constituents, Cannizzaro was able to assign
formulae to a large number of volatile compounds.

Thus in the case of ether, the molecular weight, 74,
deduced from its vapour density, was shown by the analy-
tical data to contain

Carbon 48, Hydrogen 10, Oxygen 16 parts.

Assuming the weights of the atoms to be
Carbon 12, Hydrogen i, Oxygen 16

it was clear that the numbers of atoms in the molecule were

Carbon i- = 4 Hydrogen — = 10 Oxygen — = i

12 I IO

and that the formula of the substance was
C4H100.

Formulae deduced in this way, being based directly upon
Avogadro's hypothesis, were entirely independent of the
personal bias which had produced so much confusion
.during the preceding half-century; it is therefore not sur-
prising that they soon met with universal acceptance.

The complexity of gaseous elements confirmed by
measurements of their heat capacity.— The conclusions of

XV THE MOLECULAR THEORY 353

Avogadro and Cannizzaro as to the number of atoms
present in the molecules of various gases and vapours have
been confirmed in a very striking way by independent
physical methods. It is well known that the heat required
to raise the temperature of a gas is greater when the gas is
allowed to expand than when its volume is kept constant ;
the ratio y of the two heat-capacities can be calculated to be
5/3=1-67 for an ideal or "perfect gas." Experiment has
shown that this value is rarely attained in the case of
elements, and never in the case of compounds, and further
that the value of y diminishes as the number of atoms in
the molecule increases. Typical values are as follows :

Substance. Ratio 7.

Hg, He 1-67

H2, N2, Oo, CO, HC1, HBr, HI 1-41 V

C12, Br2, I.,, C1I I -29 to I -32 /

C02, N20,N02 i'3i

It will be seen from the table that the value 1*4 is
characteristic for the lighter diatomic gases and the figure
i '3 for the heavier diatomic and the lighter triatomic gases,
whether elementary or compound. This observation supplies
a remarkable proof of the correctness of Avogadro's
fundamental proposition, that the molecules of the
elements nitrogen and oxygen are of the same complexity as
those of the compound nitric oxide, that hydrogen and
chlorine are as complex as hydrogen chloride, and so
forth.

Special interest attaches to the case of mercury, which falls
into a special class, giving a value of y which is much higher
than the figure i '4 characteristic of the diatomic gases. In
the case of so heavy a molecule this can only mean that the
number of atoms is less than two ; the molecule must

1 The value of 7 seems to depend on the moment of inertia of the
molecules : the hydride HI has probably a smaller moment of inertia
than C12, although its weight is greater.

A A

354 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

therefore be monatomic, a conclusion that is in agreement
with the rather scanty data for the vapour density of
mercury and its compounds (pp. 349 and 351). The same
method of investigation has shown that the inert gases
HELIUM, NEON, ARGON, XENON, and KRYPTON, discovered by
Sir William Ramsay in atmospheric "azote," are mon-
atomic gases; their atomic weights are therefore identical
with the molecular weights calculated from their densities.

Accurate determination of atomic weights. Valency.—
The vapour-density of an element or compound can as a
rule be determined only approximately. Moreover, even an
accurate measurement must be corrected for compressibility
before it can be used to determine the true molecular weight
of a volatile substance. It is therefore evident that the
method described above does not lend itself readily to the
accurate determination of atomic or molecular weights.

On the other hand the equivalent of an element or
compound can often be determined with extraordinary
accuracy by gravimetric analysis : thus, the figures given by
Stas for the weight of silver chloride produced from 100
parts of silver show an average deviation from the mean of
only i part in 40,000.

I32'84i 132-840 132-846 132-842
132-843 132-849 132-848 13^845 = Mean.

It is therefore important to consider how these accurate
analyses may be utilised in drawing up a table of atomic
weights based upon Avogadro's hypothesis.

Dalton's Atomic Theory finds expression in tables of
combining weights or equivalents, based solely upon the
chemical data supplied by analysis, e.g. :

H=roo8, C = 3'oo, O = 8 , 8=16-03, 0 = 35-46, etc.
Avogadro's hypothesis, on the other hand, finds its

xv THE MOLECULAR THEORY 355

expression in tables of molecular and atomic weights, based
largely upon physical measurements of density, e.g :

H=i, C=i2, 0 = i6, 8 = 32, Cl = 35i

A glance at the two lists shows that the atomic weights
calculated from the vapour densities are either identical
with, or are simple multiples of, the equivalents determined
by analysis. The ratio of the atomic weight to the equiva-
lent is called the VALENCY of the element. As elements
combine together in simple atomic proportions the valency
must obviously be a simple number. In practice, therefore,
measurements of vapour density are used to determine the
valency of an element (i.e. the factor by which the equivalent
must be multiplied to give the atomic weight) rather than
the atomic weight itself. The accurate determination of an
Atomic weight therefore includes the following operations :

(1) By means of a series of exact analyses, the equivalent
or combining weight of the element is determined with the
greatest accuracy possible.

(2) Vapour-density measurements are made with as many
compounds as possible, but not necessarily attaining a very
high order of accuracy. From these the approximate value of
the atomic weight is easily deduced by Cannizzaro's method.

(3) The ratio of the (approximate) atomic weight to the
(exact) equivalent approximates so nearly to an integer that
no doubt can exist as to the correct value of the valency.
The equivalent multiplied by the integral number repre-
senting the valency gives the exact atomic weight.

These operations may be illustrated in the case of arsenic.
Exact analyses show that 24/99 parts of arsenic combine
with 8 parts of oxygen or with 35*46 parts of chlorine ; the
equivalent of arsenic is therefore 24*99. From the vapour-
density of its volatile compounds its atomic weight is seen
to be approximately 75. Its valency is therefore 3 and the
exact atomic weight is 3 x 24*99 = 74'97-

A A 2

356 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

The exact molecular weight of a compound of known
composition may be calculated in a similar way from its
approximate vapour-density and the exactly-known atomic
weights of its constituents.

SUMMARY AND SUPPLEMENT.

A. THE PROPERTIES OF GASES.

Boyle, in 1660, with the help of a new air-pump, showed that
air was an "elastic fluid" which could be expanded to at least
152 times its original volume by merely diminishing the pressure
of the atmosphere upon it. To explain the " spring of the air"
he suggested that air was composed of elastic fibres, or (following
Descartes) of elastic spheres in rapid motion under the influence
of heat. In 1662, he showed that the pressure which air could
exert was inversely proportional to the volume which it occupied
(Boyle's Law).

Mayow. in 1674, showed that air in which a candle or lamp
had expired, and "air" prepared by the action of acids on
metals, were just as elastic as common air and could be expanded
to an equal extent.

Gay-Lussac, in 1802 (following Charles, 1787), showed that
air, oxygen, nitrogen and hydrogen all expand to theextent of
37-50 per cent, between o° and 100° C. : nitric oxide, carbon
dioxide, sulphur dioxide, ammonia, hydrogen chloride and the
vapour of ether were also shown to expand at the same rate as
common air. Gay-Lussac therefore concluded that the volumes
of all gases and vapours are influenced to the same extent by
changes of temperature and of pressure.

B. THE COMBINING VOLUMES OF GASES.

Gay-Lussac and Humboldt, in 1805, confirmed the obser-
vations of Cavendish (1781) that two volumes of hydrogen unite
with one volume of oxygen to form water ; their experiments
gave the ratio 199^89 : 100.

Gay-Lussac, in 1809, concluded that "gases always combine
in the simplest proportions when they act on one another."

THE MOLECULAR THEORY

357

The cases which he quotes in support of this view may be
represented by the following equations :

2H2 + O2 = 2H2O

I VOL

2 VOls.

NH

2NH3

2 vols.

N2

i vol.

O2

i vol.

O2

i vol.

2N

N

HC1 - NH4C1

1 vol.

C02 = (NH3)2CO

i r/0/.

3H2 - 2NH3
2SO3
2CO2

02 - 2N2O

»<?/.

02 = 2NO

+ 2SO2

2 vols.

+ 2 CO

2 zW.y.

N2

I Vol.

2O2 =2NO2orN2O4

He found further that the change of volume when gases
combine has a simple relation to their volume and quotes the
following cases :

2CO + O2 = 2CO2

2 vols. i vol. 2 vols. (contraction i vol.)

O2

1 vol.

CO2

i vol.

O2

i vol.

N2

ivo(,

+ 2C

2CO

2 vols. (expansion i vol.)

4- C = 2CO

z vols. (expansion i vol.)

+ S = SO2

i vol. (contraction nil)

+ 3H2 = 2NH3

3 vols., 2 vols. (contraction 2 vofc.)

358 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.
Gay-Lussac'sZ<TZ£/0/' Volumes may be summed up as follows :

" In every chemical change in which gases are concerned,
the volumes absorbed and liberated are in simple ratios to one
another."

This law is not exact, as the following examples show :

Hydrogen : oxygen = 2*00268 : i

Hydrogen : hydrogen chloride = 1*00790 : 2
Hydrogen : nitrogen = 3*00172 : i

It would probably be true at zero pressure, the deviations
shown above being due to the lesser compressibility of hydro-
gen as compared with the other gases.

C. AVOGADRO'S HYPOTHESIS.

Dalton, about 1802, whilst trying to form a mental picture of
the atmosphere, concluded that the particles of nitrogen, oxygen,
water- vapour, etc., were " all of the same size," in the sense that
each occupied an equal space in the gas. Later he was obliged
to reject this view, because nitric oxide occupied as much space
as the nitrogen and oxygen of which it was composed ; the con-
traction, which must accompany the production of one " com-
pound atom " from two or more " simple atoms " of equal volume,
did not in fact occur.

Avogadro, in i8ir, got over this difficulty by suggesting that
the molecules of a simple elementary gas might contain two or
more atoms. Thus, if the decomposition of nitric oxide were
represented by the equation

2NO -> N2 + O2

2 mols. i inol. i mol.

it would produce no change in the number of molecules and no
alteration in the volume of the gas.

The possibility of halving the molecules of oxygen and
nitrogen is confirmed by the production of 2 volumes of steam
from i volume of oxygen and of I volume of nitrogen from 2
volumes of ammonia.

The halving of the hydrogen molecule was proved by the
presence of i^ volumes of hydrogen in i volume of ammonia,

xv THE MOLECULAR THEORY 359

whilst the presence of ^ volume of hydrogen and \ volume of
chlorine in i volume of hydrogen chloride proved the divisi-
bility of both hydrogen and chlorine.

2H2 + 02 -> 2H20

I Vol. 2 Vols.

2NH3 -> N2 + 3H2

2 vols. i vol. 3 vols.

H2 + C12 -> 2HC1

I VOL I »0/. 2 ZX7/.T.

In each of the cases studied by Avogadro, H2, O2, N2, C12,
the molecule proved to be divisible into two parts : his sugges-
tion that subdivision into four or eight parts was possible has
been realised in the molecules, P4, As4, Sg ; cases are also known
in which the molecule contains only a single atom and cannot
be subdivided, e.g. Hg, Zn, Cd, Ar ; in the case of ozone, O3,
the molecule contains three atoms.

D. CANNIZZARO'S EXPOSITION.

Avogadrds Hypothesis states that

" If the temperature and pressure are the same, equal volumes

of different gases contain equal numbers of molecules."

From this hypothesis, which is strictly true only at zero
pressure, it follows that the molecular weights of different gases
are in the same ratios as their densities. Caimizzaro, in 1858,
selected the half-molecule or atom of hydrogen as the standard
for all molecular and atomic weights. The molecular weight of
a gas is then equal to its density relative to hydrogen = 2, or, in
the modern system, relative to oxygen = 32. The atomic weight
of an element is the smallest weight which is found in the mole-
cular weights of its volatile compounds.

The molecular and atomic weights deduced from vapour-
densities are not exact. But the ratio of atomic weight to the
equivalent (the valency of the element) is a whole number, and
it is therefore easy to deduce the exact atomic weight from the
equation

Atomic weight =• valency x equivalent.

After the atomic weights have been determined, the formula
of any volatile compound can be deduced from its vapour density
and percentage composition by using the method illustrated
on p. 353.

CHAPTER XVI

ATOMIC WEIGHTS OF THE METALS

A. VAPOUR DENSITIES

The vapour-density of metallic compounds. — Avogadro's
hypothesis, which enabled Cannizzaro to establish on a
firm basis the atomic weights of the non-metals, failed him
in the case of the metals, because of the lack of volatile
compounds. Very few of the metals can be vaporised,
either as elements or compounds, even under the wide range
of conditions now available for vapour-density determina-
tions. The most favourable cases-*re those of mercury and
arsenic, two elements which in some ways serve to bridge the
gap between metals and non-metals. The data already given
for these two elements, with others derived from their
volatile organic compounds, are sufficient to establish the
atomic weights, Hg = 2oo, As = 75.

The following table contains almost all the data that are
available for other metals l :

1 The references are : — Dumas, Ann. Chim. Phys., 1826, 88,385;
Dumas quoted by Bineau, Ann. Chim. Phys., 1838, 68, 427-428 ;
Frankland, Phil. Trans., 1855, 145, 266; Deville and Troost, Ann.
Chim. Phys., 1860, 58, 257-299; Wanklyn, Jour. Chcm. Soc., 1861,
13, 128; Frankland and Duppa, Jour. Chem. Soc., 1864, 17, 33;
Buckton and Odling, Proc. Roy. Soc., 1865, 14, 19-21 ; Ladenburg,
Liebig's Annalen, 1872, 8, Suppl. 55-80; V. and C. Meyer, Ber.,
1879, 12, 1195-1200 ; Menschingand Meyer, Ber., 1886, 19, 3295-3298,
and Ber., 18,87, 20, 582-583; Scott, Proc. Roy. Soc. Ed., 1887, 14.
410 ; Friedel and Crafts, Comptes rendus, 1888, 107, 306 ; Nilson and
Pettersson, Trans. Chem. Soc., 1888, 53, 814-831 andZizV. Phys. Chem.,
1889, 4, 206-225 5 Biltz and Meyer, Ber., 1889, 22, 725 or Zeit. Phys.
Chem., 1889, 4, 249-269 ; Mond, Langer and Quincke, Trans. Chem.
Soc., 1890, 57, 752 ; Dewar and Jones, Proc, Roy, Soc,, 1905, 76, 567,

360

xvi ATOMIC WEIGHTS OF THE METALS

V?-* ,, a a IB a ,sac oo

o

JS

i 2*

i I
S

1

'w £. en £. R. So*o 2 t^vS R. 2 R.^ 2. 5- 5- » O « bv"«« 0 'cs O V

i-iUUOOWUOWH-iUoOljvjOo KJlJffiffiffiffiKOO

c^ p" p" p~ p" p" p" p" p" p"

o 000000000

CONOVOOMWlOlOlOlOMU->lOlOfr>*-' VOvOVOvNNCOMlo^

p p ;•*• .•* O p _•*• ro O N _io p ^ tx rx TJ- pv TJ

Vo c^ "10 "o vo oo Vovb oo w rrj V^ *ro »o vb ^*- ^-oo 10 o^ O
<M mvo M vo o -«• co t^ rx rovo ro w IN CN to 10 N r^vo

« w ro o oo tx

rovo M vo oo vo rovo * ^txrotxwroioioro r^vo c>oi Ooo row f^Ovt^

O °n °Q °Q °O °O °p

°" ^°o ^ |° ° °o °o °o 2"°N °tx £ f?0^0*- °0 oo OLO°O °-*0|^°5 °° °

i • i • i : ; i ; ; •: %£ • • z. , • •

O £ -^ "M "

~> fx irt

.

1 STJJ sr i£M 8-5 - -c«sM. 5*3

rs.1 ***"« --ili^i

0) V

v -a -a

T3 — —

ow ^^p-am B * le*a a

. . . ."2."S 0 'g ^ o'^.-o u ."2 ."S .2 -c -o "g • •

•>,

I !?

g CD

« rt rt x

i 1

ilia

II

pj g*a « ^

•1 Mil • £

I III ?

llli 1

Illiil

2 w^^g

PJ l

1 S.SSOSSSB (J^^ u co o

111 'III 11*11 1 11 1.1TI&H

362 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

Data are also available for antimony, bismuth and
tellurium, but no vapour-density determinations appear to
have been made in the case of magnesium, calcium,
strontium, barium, gold and platinum. It is clear that the
data are far too scanty to give trustworthy values for the
atomic weights of the metals. Thus the data for iron

are :

Molecular Weight of Iron in
Compound. Weight. Molecular Weight.

Ferrous chloride 125*6 ... 56 (about)

Ferric chloride 332 '2 ... 112 ,,

Iron carbonyl 199*0 ... 56 ,,

From so short a table we cannot tell whether 56, or a
sub-multiple of 56 (say 28), is the atomic weight of iron ; in
other words, we cannot be sure whether the molecule of
ferrous chloride or of iron carbonyl contains one, or two or
more atoms of iron. The facts show only that 56 is the
maximum value for the atomic weight of the metal.

The formulae given in the last column of the large table are
based upon atomic weights deduced by other methods. They
show that the smallest weight of metal found in the molecules
of its volatile compounds is usually a single atom. But
pairs of atoms are frequently found

(Al2Cl6,Al2Br6,Al2IG,Fe,Clc, etc.).

In the case of copper the only compound examined is of
this type, the atomic weight of the element being 63 '6 and
not 127*1.

B. THE LAW OF ATOMIC HEATS.

Dulong and Petit's law of atomic heats (1819).— The

difficulty of finding the atomic weights of the metals by
means of density determinations renders very important the
fact, discovered in 1819 by the French physicists Dulong
and Petit (Ann. Chim. Phys., 1819, 10, 405), that

" The atoms of all simple substances have the same capacity
for heat."

XVI

ATOMIC WEIGHTS OF THE MP:TALS

363

Their table shows :

(1) The elements which they examined.

(2) Their values for the heat-capacity of these elements,
compared with that of an equal weight of water.

(3) The values which they selected for the atomic weights
of the elements, O= i.

(4) The product of these numbers.

A fifth column has been added which shows Dulong and
Petit's atomic weights reduced to the modern standard,
O=i6.

ATOMIC HEATS (DULONG AND PETIT).

Atomic

Atomic

Element.

Heat
Capacity.

Weight.
0=i

Product.

Weight.
O=i6

Bismuth

0-0288

1 3 '3°

0-3830

212-8

Lead
Gold

0*0293
0-0298

12-95
12-43

Q'3794
0-3704

207-2
198-9

Platinum

0-0314

I2-I61

0-3740

194-6

Tin

0-0514

7'35

0-3779

117-6

Silver

0-0557

675

0-3759

1 08-0

Zinc

0-0927

4-03

0-3736

64'5

Tellurium

0-0912

4 "03

0-3675

64-5

Copper

0-0949

3'957

0-3755

633

Nickel

0-1035

3-69

0-3819

59-o

Iron

O'lIOO

3'392

o-373i

543

Cobalt

0-1498

2-46

0-3685

39 '4

Sulphur

0-1880

2 -01 1

0-3780

32-2

1 The value Ii"i6 given in Dulong and Petit's paper appears to be
a misprint for 12 'i6 (Berzelius's number for the atomic weight) ; but
there is some further error in the table, as 12*16 x 0-0314=0-3818.

In Dulong and Petit's table the atomic weights of the
metals were reduced to one-half of the values given by
Berzelius, whilst the value for silver was reduced to one-
quarter. The new values were ultimately accepted by
Berzelius, and are substantially the same as those employed

364 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

at the present time ; tellurium and cobalt (modern values,
127-5 and 58*97) are exceptions to this statement, on
account of errors in the values for their specific heats.

Significance of the law of atomic heats. — The SPECIFIC
HEAT of a substance is the amount of heat in calories
required to raise the temperature of i gram through i° C.
The product of the specific heat and the atomic weight of
an element is called its ATOMIC HEAT and represents the
amount of heat required to raise the temperature of i
" gram-atom " (i.e. the atomic weight expressed in grams)
through i° C. Dulong and Petit's LAW OF ATOMIC HEATS
can be expressed algebraically in the form :

Specific heat x atomic weight — atomic heat = constant = 6'o.

The average value of Dulong and Petit's product was
0*3753 (O = i). If reduced to the modern standard, O = 16,
it becomes 6;oo, a number which is practically the same as
the average value 6 '05 for the atomic heat over the range
from 20° to 1 00° of the 25 solid elements tabulated on
page 365.

If the specific heat of an element at ordinary temperatures
is known, its atomic weight may be calculated by dividing the
specific heat into the figure 6'o. The atomic weights
obtained in this way are even less accurate than those
deduced from vapour densities ; but they are sufficiently
exact to give the valency of the element, i.e. the factor which
must be usecl in deducing the atomic weight from the
equivalent, as described in the preceding chapter.

The law of atomic heats can be applied most easily in the
case of the metals in which the direct application of
Avogadro's hypothesis is most difficult. It was much used
by Berzelius, and was regarded by him as a more trust-
worthy guide than Avogadro's hypothesis ; but, as Cannizzaro
showed, it is not independent of that hypothesis, and merely
provides a method for extending it from hydrogen and

XVI

ATOMIC WEIGHTS OF THE METALS

365

oxygen to silver and gold ; this is done with the help of
elements such as iodine and mercury, the atomic weights of
which can be compared both by means of their vapour
densities and by means of their specific heats.

Abnormal atomic heats. — The following table shows the
atomic heats, over different ranges -of temperature, of 25

ATOMIC HEATS OF. SOLID ELEMENTS

Specific

Atomic

Atomic

Atomic

Element.

Atomic
Weight.

Heat.
+ 20° to

Heat.
+ 20° to

Heat,
-i 88° to

Heat.
- 253° to

+ 100°.

+ 100°.

+ 20°.

- I95J C.

C

12-00

0-1990

2*4

I'I5

0-03

Mg ...

24-32

0-2492

6-1

5'06

174

Al ...

27*1

0-2173

5'9

473

I '12

Si ...

28-3

0-1833

3'34

077

P

31-04

0-1981

6'2

5'24

i '34

s

32-07

0*1780

57

4-20

175

Cr .

52-0

0-I2IO

4-14

0*70

Mn .

54 '93

O*I2II

67

5-12

1-26

Fe ...

0*1146

6-4

4-80

0-98

Ni ...

58-68

0*I092

64

5-09

1*22

Co ...

58*97

O-IO3O

6-2

4-88

I "22

Cu ...

63 '57

0*0936

6-0

5-01

I'56

Zn ...

65 '37

0*093I

6-1

5]53

2-52

As ...

74-96

o 0827

6-2

1-94

Mo ...

96-0

0-0647

6-2

5 '33

1-36

Pd ...

106-7

0-059

6"3

2-03

Ag ...
Cd ...

107-88
112-4

o "0566

0*0549

6-1
6-2

579

2-62

3 '46

Sn ...

119-0

0*0556

6-6

5'97

3-41

Sb ...

I2O'2

0*0503

6*1

5-63

2-89

Pt ...

195*2

0-0320

6*3

5'45

2-63

Au ...

I97-2

0-0316

6-2

5-86

3*16

Tl ...

204-0

0-0326

67

6-04

4-80

Pb ...

207-I

0*031

6*4

6-21

4-96

Bi ..

208 *o

o 0303

5'9i

4 '54

Average

—

—

6-05

S-C6

2-12

solid elements. Over the range from +20° to +100° the
average atomic heat is 6*05 calories. At lower temperatures,

366 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

the atomic heats are much diminished. Between 20° and
- 1 80° the average value for the 25 elements is 5-06
(Richards and Jackson, Annual Reports, 1910, 16). Be-
tween — 195° and — 253°C. it is 2*12 calories (Dewar, Proc.
Roy. Soc., 1913, A. 89, 168). The values for different
elements are then not even approximately equal, elements
which have a large atomic volume (at. wt. -r density) giving
large atomic heats, whilst elements with small atomic
volumes have small atomic heats. It is only as the tem-
perature is raised that Dulong and Petit's Law becomes
valid.

In the range from 20° to 100° all the solid elements
except two give values lying between 5*2 and 67. The
two exceptions are :

Boron... At. wt. II Sp. ht. 0*2616 At. ht. 2*9

Carbon. At. wt. 12 Sp. ht. O'i532 (diamond) At. ht. I '84 (diamond)
0-1990 (graphite) 2 -39 (graphite)

In the case of diamond, the atomic heats at different tem-
peratures are as follows (Nernst, Annual Reports, 1912, 9) :

Temperature... 896° 85° -41° -64° - 181° -231° - 243°C.
Atomic heat ... 5-45 2-12 O'86 O'66 0-03 O'oo o-oo

the values diminishing from a normal figure at about
1000° almost to nothing at temperatures below the boiling
point of liquid air. In the case of boron the atomic heats
are (Weber, Phil. Mag., 1875, [ivj ^9, 293 ; Dewar, loc. at.) :

Temperature... 233*2° 177-2° 125-8° 767° 26*6° -39*6 -224°
Atomic heat ... 4*03 3-72 3-38 3-01 2*62 2'ii 0-24

These exceptions are confined to elements of small atomic
weight, to which Avogadro's hypothesis can easily be applied •;
they do not, therefore, affect the usefulness of Dulong and
Petit's Law as a means for determining the atomic weights
of those elements which do not form volatile compounds.

xvi ATOMIC WEIGHTS OF THE METALS 367

Molecular heat of compounds — Neumann, in 1831 (Pogg,
Ann. der Physik, 1831, 23, 1-39), found a constant value
for the MOLECULAR HEATS (sp. ht. x mol. wt.) of a series of
seven carbonates and another constant value for a series of
four sulphates ; in each of these cases the atomic heat of
the combined metals was constant throughout the series.

Cannizzaro, in 1858 (A.C.R. XVIII. 23), showed that mer-
cury and iodine obey Dulong and Petit's Law when combined
together, and used this fact as a further proof of his view that
the atomic weight of mercury was 200 and not 100, as had
been supposed previously. The figures he quotes are as
follows :

Atomic heats :

Solid bromine, Br1 80 x 0-0843 — 6746

Iodine, I1 127x0-0541= 6*873

Solid mercury, Hg 200x0*0324 = 6*482

Molecular heats :

Mercurous chloride, HgCl1 (2004- 35-5) xo*o52i = 12-258 = 2 x 6-129
Mercurous iodide, Hgl1.., (200+127 ) xo*O395= 12*913 = 2 x 6 457
Mercuric chloride, HgCl2. (200 + 71 ) x 0*0689 = 18*669 = 3 x 6*223
Mercuric iodide, HgI2 ...(200 + 254 ) x 0-0420= 19*054 = 3 x 6*351

If the atomic weight of mercury were 100, the molecular
heats of the mercurous salts should have been 3x6, and those
of the mercuric salts 4 x 6, for the quantities shown in the
table.

A more detailed study of the molecular heats of com-
pounds was made in 1865 by Kopp ("Investigations of the
Specific Heat of Solid Bodies," Phil. Trans., 1865, 155,
71-202), who showed that "Each element has the same
specific heat in its solid free state and in its solid com-
pounds " (loc. cit. p. 83).

Whilst Cannizzaro assumed that in all cases the
molecular heat of a compound '= number of atoms x 6,

1 The argument is not affected if these formulae are written Br.2, I2,
Hg2Q2, and HgJ2-

368 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

Kopp's rule gives expression to the fact that lower values
are found for the molecular heats of compounds containing
elements such as boron, carbon and sulphur, of which the
atomic heats are abnormally low.

C. THE LAW OF ISOMORPHISM.

Mitscherlich (1819) discovers isomorphism. — In select-
ing atomic weights, Berzelius adopted the rule that com-
pounds which resemble each other closely should be repre-
sented by similar formulae. A remarkable illustration of
this close resemblance between compounds of similar com-
position was discovered in 1819 by Mitscherlich, a pupil
of Berzelius. " He examined the acid phosphate and arsenate
of potassium, which are now represented by the formulae
KH2PO4 and KH2AsO4, and found that

" these salts are composed of the same number of atoms . . . ,
and only differ from one another in that the radical in one
is phosphorus and in the other arsenic. The crystalline
form of these two salts is the same . . . ; for not only the
primitive form, but all the varieties, resemble each other so
closely in the size, number, and angles of the faces, that it
is quite impossible to find any difference, even in the
characters which appear to be quite accidental " (" On the
Relation between Crystalline Form and Chemical Propor-
tions," Ann. Chim. Phys., 1820, 14, 172 and 173).
A similar analogy of form was detected in

the sodium salts { ^J^O } and

the ammonium salts j /^S^S2 A°n I
I ViNrlJrlgAS^ J

In each of these cases 75 parts of arsenic could take the
place of 31 parts of phosphorus, without producing any alter-
ation of crystalline form.

In the second part of the memoir (Ann. Chim. Phys.,
1821, 19, 350) Mitscherlich described as ISOMORPHOUS

xvi ATOMIC WEIGHTS OF THE METALS 369

those groups of elements which could take the place of one
another in compounds of identical crystalline form. In the
case of isomorphous elements, "The same number of
atoms combined in the same way produces the same crys-
talline form • the crystalline form is independent of the
chemical nature of the atoms, and is determined only by
the number and relative position of the atoms " (ibid., p.
419).

The law of isomorphism. — The LAW OF ISOMORPHISM
as now enunciated states that : " Substances which are similar
in crystalline form and in chemical properties should be
represented by similar formula"

This law is limited in two directions : (i) On one hand,
mere identity of crystalline form is not sufficient to prove
that the formulae of two compounds should be similar.
In the case of crystals of a high degree of symmetry, such as
the cube and octahedron, the angles can be calculated as
simple fractions or functions of a right angle, and are identical
for all the elements and compounds (iron, silver, iron pyrites,
salt, sal-ammoniac, etc.) which crystallise in these forms.
(2) On the other hand, if the crystalline form is less symmetri-
cal, the angular measurements may differ considerably, even
in the most closely-related compounds; thus, the mineral
carbonates crystallising in simple rhombohedra show the
following angles :

Calcite, CaCO3 74°55'

Magnesite, MgCO3 72°36'

Chalybite, FeCO8 73°o'

Calamine, ZnCO3 72°2o'

Rhodochrosite, MnCO3... 72°s8'

Isomorphous mixtures or solid solutions. — Some more
certain test than mere geometrical form is evidently
required before " isomorphism " can be used as trustworthy
evidence of analogous chemical composition. Such a test is
supplied by the fact (investigated by Mitscherlich) that many

B B

370 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

compounds, which are similar both in chemical. type and in
crystalline form, possess the property of forming crystals
in which the two constituents are mixed as intimately as
the salt and water in an aqueous salt solution. Crystals of
this kind are described as ISOMORPHOUS MIXTURES or SOLID
SOLUTIONS. They are formed whenever two or more closely-
related substances are allowed to crystallise together from
solution or from the fused state. Thus if

potash alum ..... KAl(SO4)2,i2H2O
' soda alum ..... NaAl(SO4)2,i2H2O
and chrome alum .... KCr(SO4)2,i2H2O

are dissolved together in water, the crystals which separate
will contain all three substances, the colour ranging from a
pure white to a deep purple according to the proportion
of chrome alum which they contain. So also in the case of
minerals, garnets of the type
Ca3Al2Si3O12 or

which have crystallised from a magma of complex com-
position, are all mixtures of isomorphous substances, in
which a part of the calcium has been replaced by magnesium,
ferrous iron, or manganese, and part of the aluminium by
ferric iron, or by chromium. But in all this complex
substitution (known to mineralogists as VICARIOUS REPLACE-
MENT) the type remains constant and the general formula

3[Ca,Mg,Fe,Mn]0,[Al,Fe,Cr]203,3Si02

may be used to represent every member of the series.

The formation of isomorphous mixtures is a valuable test
of chemical similarity, because they cannot be produced from
substances of different type, even when absolutely identical
in crystalline form ; on the other hand, a difference of 5° in
the angles of analogous compounds need not prevent the
formation of an isomorphous mixture, since composite

xvi ATOMIC WEIGHTS OF THE METALS 371

crystals may still be formed, the angles of which are a
compromise between those of the constituents.

Applications of the law of isomorphism. — In the
isomorphous phosphates and arsenates, 75 grams of arsenic
replace 31 grams of phosphorus without altering the crystal-
line form. Since 31 is the atomic weight of phosphorus,
the Law of Isomorphism indicates that 75 is the atomic
weight of arsenic, thus confirming the value arrived at by
other methods. A similar resemblance appears in the
crystalline properties of arsenic, antimony and bismuth.
In the isomorphous compounds of these elements, 75
grams of arsenic are replaced by 120 grams of antimony,
and 120 grams of antimony by 208 grams of bismuth,
without producing any large alteration of crystalline form.
In this way the Law of Isomorphism leads up from P = 3i,
through As = 75 to Sb = 1 20 and Bi = 208.

In 1828 Mitscherlich showed (Pogg. Ann., 1828, 12, 138)
that the selenates and sulphates of sodium and silver formed
an isomorphous series in which 23 parts of sodium were
replaced by 108 parts of silver, whilst 32 parts of sulphur
were replaced by 79 parts of selenium. This isomorphism
served to establish the atomic weight of selenium.

Isomorphism of manganese and chromium with
chlorine and sulphur. — Two interesting cases of isomor-
phism were described in detail by Mitscherlich in 1832
(Pogg. Ann., 1832, 25, 287-302). He showed that the red
potassium permanganate was isomorphous with potassium
perchlorate. In these compounds 55 parts of manganese
replace 35-5 parts of chlorine; 55 is therefore the atomic
weight of manganese. On the other hand, the green
potassium manganate was isomorphous with the chromate,
selenate and sulphate; the equivalent quantities, S 32, Se 79,
Cr 52, Mn 55, are therefore the atomic weights of these
elements.

Mitscherlich had already proved the isomorphism of the

B B 2

372 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

divalent metals Ca, Mg, Mn (manganous), Fe (ferrous),
Cu, Zn, Co, Ni in their simple and double sulphates (Ann.
Chem. Phys., 1821, 19, 416), and of the tervalent metals
Cr, Mn, Fe, Al in the spinels MgAl2O4,Fe3O4, etc. The way
was therefore open to link up the atomic weights of nearly all
the metals with those deduced by Avogadro's hypothesis
for the non-metals chlorine and sulphur.

Mitscherlich's observations on isomorphism, confirmed as
they were by Dulong and Petit's Law of Atomic Heats, led
Berzelius in 1827 (Jahresbericht, 1828, 7, 67-78) to write
the oxides of iron as FeO and Fe2O3 instead of FeO2 and
FeO3, and to publish a new table of atomic weights in which
the values for nearly all the metals were halved. He argued
from the isomorphism of the sulphates and chromates that
chromic acid must be CrO3, compare SO3 ; chromic oxide
was therefore Cr2O3. The isomorphism of the salts of
chromium, manganese, ferric iron and aluminium gave the
formula Mn2O8, for manganic oxide, Fe2O3 for ferric oxide,
and A12O3 for alumina. But if ferric oxide was Fe2O3,
ferrous oxide must be FeO, and this type of formulae must
also be given to all those oxides the salts of which were
isomorphous with the ferrous salts.

Exceptions to the law of isomorphism. — There are many
exceptions to the law of isomorphism. On one hand,
many substances of similar structure crystallise in totally
different forms. Thus Mitscherlich found that the usual
form of acid sodium phosphate, NaH2PO4, was quite different
from that of the acid arsenate, NaH2AsO4. This difference
is due to the POLYMORPHISM of the phosphate, which forms
two types of crystals, only one of which is isomorphous with
the arsenate.

On the other hand, perfect isomorphism is often found in
substances of dissimilar composition. The most important
case is that of the ammonium salts in which the compound
radical NH4 may be displaced by a single atom of sodium

xvi ATOMIC WEIGHTS OF THE METALS 373

or potassium without producing any change of crystalline
form. Again, silver sulphide, Ag2S, as argentite, is isomor-
phous with lead sulphide, PbS, as galena, and is nearly
always present as an isornorphous impurity in this mineral,
in spite of the unequal numbers of atoms in the two
molecules. Cases of this sort show that the law of
isomorphism, although of great value, cannot be used
as a final test in establishing the atomic weights of the
elements.

D. FORMULA OF METALLIC COMPOUNDS.

Empirical and molecular formulae. — In the preceding
chapter (p. 352) a method has been described by which the
formula' of a volatile compound may be deduced from its
vapour-density and percentage composition, when the atomic
weights of its constituents are known. This method can-
not be applied to substances, such as sugar or potassium
chlorate, which decompose when heated, or to substances,
such as silica, which boil at so high a temperature that their
vapour-densities cannot be determined. There is, however,
no difficulty in deducing from the analysis of the substance
an EMPIRICAL FORMULA, which expresses its composition in
the simplest way, but gives only a minimum value for its
molecular weight.

The method used in calculating an empirical formula
may be illustrated by using Stas's figures for the analysis of
cinnamic acid (Works, I. 281).

On combustion, 0*900 gram of acid gave 0*444 gram
of water and 2*402 grams of carbonic anhydride. The
composition of the acid was therefore :

C 7278 H 5-47 O 2175%.

Taking the atomic weights as

C 1 2 'oo H i '008 O 1 6,

374 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.
the numbers of atoms are in the proportion

c H _ o or

I2*OO I'OOS IO

C 6-06 H 5-43 O 1-36.

Dividing by the smallest number, these proportions become

C 4-46 H 3-99 O i,

or yery nearly

C 4^ H 4 O i.

Multiplying by two in order to remove fractions, the
simplest empirical formula for the compound is seen to be
C9H8O2. The composition calculated for this formula is

C 72-94 H 5-45 O 21-61

and agrees closely with the composition found by com-
bustion. The MOLECULAR FORMULA, which should express
both the composition and the molecular weight of cinnamic
acid, is [C9H8O2]?j., where n is an integer, the value of which
must be determined by other methods, e.g. from the chemical
properties of the acid or from the boiling-points and freezing-
points of its solutions.

Empirical formulae of metallic salts. — The following
examples show how the composition and the empirical
formulae of some typical metallic salts have been determined.
The analytical data are taken from Berzelius's1 papers on
Fixed Proportions, 1811 — 1812 (Ostwald's Klassiker,
No. 35). The atomic weights are substantially those of Stas.

i. Lead oxide, sulphide and sulphate. — Berzelius found
(Fixed Proportions, 1811-1812; Klassiker, XXXV, 6-13)
that :

1 In quoting Berzelius's data it has been necessary to distinguish
between the actual analyses, and the compositions which he calculated
on the assumption that both chlorine and nitrogen were not elements
but oxides.

/i ATOMIC WEIGHTS OF THE METALS 375

(1) 10 grams of lead, dissolved in nitric acid and ignited,
gave 1078 grams of the oxide (litharge).

(2) 10 grams of lead, when fused with sulphur, gave 1 1*56
grams of sulphide.

(3) Io grams of lead sulphide, when digested with aqua
regia, gave 12*65 grams of sulphate, no excess of lead or of
sulphuric acid being left uncombined.

(4) 10 grams of lead, dissolved in nitric acid and pre-
cipitated with sulphuric acid, gave 14*635 grams of lead

sulphate, in close agreement with the weight -*-^ — —— — 5

10

= 14*623 grams calculated from (2) and (3).

From these experiments Berzelius concluded :

(5) That lead and sulphur in lead sulphide are combined
together in exactly the same proportions as when their
oxides unite to form lead sulphate.

(6) That 10 grams of lead sulphide in forming lead
sulphate unite with 2*65 grams of oxygen ; the 10 grams of

sulphide contain IQO = 8*651 grams of lead, yielding on
11*56

oxidation — ^r x IO ' = 9*326 grams of lead oxide, thus
10

accounting for 9*326 - 8*651 = 0*675 gram of oxygen ; the
remainder, 2*65 - 0*675 = r975 of tne oxygen in the
sulphate was therefore combined with 10 - 8*651 = 1*349
grams of sulphur to form 3*324 grams of sulphuric anhy-
dride. The composition of sulphuric anhydride was there-
fore :

Sulphur 1*349 grams = 40'58%
Oxygen 1*975 grams = 59*42%

i.e. sulphur : oxygen = 2:3 nearly.

(7) In the sulphate, litharge containing 0*675 gram of
oxygen was combined with sulphuric anhydride containing
1*975 grams of oxygen. The weights of oxygen in the acid
and in the base were therefore in the ratio 3 : I, a ratio that
held good for the sulphates generally (Chemical Proportions,
1819, p. 122).

376 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

By the analysis of barium sulphite (Klassiker, XXXV, 18), the
composition of sulphurous anhydride was found to be

Sulphur = 50'55%
Oxygen = 49 '45%

i.e. sulphur : oxygen = I : i nearly. The weights of oxygen in
the two oxides were thus in the ratio 3 : 2 as deduced from Gay-
Lussac's volumetric analysis (p. 169). Berzelius found that when
barium sulphite was oxidised to barium sulphate, no excess of
baryta or of sulphuric acid was produced. The ratio of barium
to sulphur was therefore the same in the two compounds. The
ratio of oxygen in the acid to oxygen in the base was 2 : i for the
sulphites, as contrasted with 3:1 for the sulphates (Chemical
Proportions, p. 122).

Taking the modern values for the atomic weights, the
empirical formulae of these compounds can be deduced

at once.

LITHARGE

Formula PbO.
LEAD SULPHIDE =

Formula PbS.
SULPHURIC ANHYDRIDE

Formula SO3.

Lead
10 grams :

TO

Oxygen
0-78 gram

atoms

207*1
0-0483 :
i :

0-0487 atom
i atom

Lead

10

Sulphur
: i -56 grams

TO
207*1

0*0483

: -L-1- atoms
32-1

. 0*0486 atom

I

: i atom

Sulphur
I-349
r349
32-1
0*0420
i

Oxygen

i -9 7 5 grams

1221 atoms

16

0-12 34 atom
3 atoms

n ATOMIC WEIGHTS OF THE METALS 377

LEAD SULPHATE =

Lead

Sulphur

Oxygen

8-651

: r349

: 2*65 grams

8-651

• _11349

: 216-5 atoms

207*1

32-1

16

0*0418

: 0*0420

: 0*1656 atom

I

: i

: 4 atoms

Formula PbSO4.

The corresponding formula for lead sulphite is PbSO3.

2. Potassium Chlorate. — In this case Berzelius weighed
the oxygen directly (by loss of weight) and the chlorine as
silver chloride, and determined the metal by difference.

(1) 3*9850 grams of potassium chlorate, heated in a small
retort, lost i'5475 grams of oxygen and left 2*4375 of potass-
ium chloride (Klassiker, XXXV. 122).

(2) 10 grams of potassium chloride, precipitated with silver
nitrate, gave 19*21 grams of fused "horn-silver" (silver
chloride) (Klassiker, XXXV. 93).

(3) 3 grams of pure silver, dissolved in nitric acid, pre-
cipitated with muriatic acid, evaporated and fused, gave
3*98 grams of " horn-silver " (Klassiker, XXXV. 24).

It follows

(4) From (i), that 3*9850 grams of potassium chlorate con-
tain 1'5475 grams of oxygen.

(5) From (2) and (3), that since 3*98 grams of silver
chloride contain 3 grams of silver and 3*98-3*00 = 0*98
grarn of chlorine, therefore 19*21 grams of horn-silver con-

tain - .-4- x °'98 = 473 grams of chlorine, derived from

10 grams of potassium chloride ; thus, 3*9850 grams of
potassium chlorate or 2*4375 grams of potassium chloride

contain x ^^ _ 1*1529 grams of chlorine.

10

(6) Finally, 3*9850 grams of potassium chlorate must con-
tain 3*9850- 1*5475 - 1*1529 = 1*2846 grams of potassium.

Chlorine
1*1529
1*1529

Oxygen
: i "5475 grams
• T'5475 atoms

35'46
0*0325

I

16
: 0*0967 atom
: 3 atoms

378 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.
The formula of the salt is deduced as follows :

Potassium

POTASSIUM CHLORATE = 1*2846
1-2846

= 39'i°
= 0-0329
= i

Formula KC1O3.

3. Chalk. As Berzelius was not able to obtain pure
metallic calcium, he could not determine the composition
of lime by direct oxidation ; an indirect method of analysis
was therefore adopted :

(1) 10 grams of chalk when ignited gave 5*64 grams of
lime and 4-36 grams of carbonic anhydride containing
28-4% of carbon.

(2) 10 grams of chalk dissolved in muriatic acid gave
10-96 grams of fused muriate of lime (calcium chloride).

(3) 3'QI grams of this fused muriate, when dissolved and
precipitated with silver nitrate, gave 775 grams of " horn-
silver" (fused silver chloride), containing 7 5 '3% of silver and
therefore 24-7% chlorine (Klassiker, XXXV. 77).

From these experiments it followed

(4) That 10 grams of chalk contained 0-284 x 4'36 = 1'238
grams of carbon, united with 4*36- 1*238 = 3*122 grams of
oxygen to form carbonic anhydride.

(5) That 3-01 grams of muriate of lime contained 0*247 x
7-75 = 1*914 grams of chlorine united with 3-01-1*914 =
1-096 grams of calcium ; therefore the 10-96 grams of
muriate of lime derived from 10 grams of chalk contained

I0'96 x 1*096 = 3*99 grams si calcium.
3*01

(6) That the 5*64 grams of lime obtained by igniting 10
grams of chalk contained this 3*99 grams of metal united
with 5*64-3*99 = 1'65 grams of oxygen.

xvi ATOMIC WEIGHTS OF THE METALS 379

The composition of chalk is therefore :

f LIME, 5-64 grams= - f

CHALK, ,o grams = CARBONIC ANHYDRIDE}8™,™ % ™* "

I 4-36 grams = \CARBON... 1*24 ,,

The oxygen in the acid is seen to be double the amount
in the base, as in the case of the sulphites.
The proportions of the three elements are :

Calcium

Carbon

Oxygen

3*99

: 1*24

: 4*77 grams

= 3'99

i'24

4*77

: 2_l_f atoms

40*1

12*0

16

= 0*099

: 0*103

: 0*298 atom

= i

: i

3 atoms

Formula CaCO,,.

4. Lead Nitrate. — Berzelius found that

(1) 20 grams of dry, powdered lead nitrate, ignited in a
weighed platinum crucible, left 13*445 grams of oxide ;
therefore 100 grams of the salt contain 67*22 grams of lead
oxide, composed of 4*85 grams of oxygen and 62'37 grams
of lead (Klassiker, XXXV. 129).

(2) Gay-Lussac had proved (p. 203) that nitric anhydride
was composed of 2 volumes of nitrogen of density 0*969 and
5 volumes of oxygen of density 1*104 (air = 0 ' the com-
position by weight was therefore 0*969 x 2 : 1*104 x 5 =
i "938 : 5*520. The weights of the two gases in 100 - 67*22 =
32*78 grams of nitric anhydride are therefore

— fx 32*78= 8 '52 grams of nitrogen
and 32*78 - 8*52 = 24 '26 grams of oxygen.

From these analyses it is clear that the composition of
the salt is :

( LITHARGE, ^LEAD 62*37 grams

LEAD NITRATE, J 67^22 grams = \OXYGEN... 4*85

ANHYDRIDE, fOXYGEN... 24*26

32 '78 grams = \ NITROGEN 8*52

ioograms= | NITRIC ANHYDRIDE, | OXYGEN... 24-26

380 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

and that the oxygen in the acid is five times1 that in the
base (Berzelius, Chemical Proportions •, 1819, p. 124).
The proportions of the three elements are :

Lead Nitrogen Oxygen

LEAD NITRATE = 62*3 7 : 8*52 : 29*11 grams

62*37 8'ij2 2Q-II

: : atoms

207*1 14*01 16

= 0*301 : 0*608 : 1*819 atoms
= i : 2 : 6 atoms

Formula PbN2O6.

SUMMARY AND SUPPLEMENT
A. VAPOUR-DENSITIES

Vapour-densities have been determined for the following
metals and metallic compounds :

Elements : Na, K, Zn, Cd, Hg.

Haloid Salts: KI, Cu2Cl2, AgCl, ZnCl2, PbCl2, 2A1C13Z±
Al2Clfl, Al2Brr>, A12IC, 2FeCl2^Fe2Cl4, Fe2Cl6^±2FeCl3^±
2FeCl2 + Cl2, CrO2Cl2, CrCl3, MnCl2, SnCl2=^Sn2Cl4, SnCl4,
HgCl2, Hg2Cl2^Hg + HgCl2, etc., in addition to com-
pounds of As, Sb, Bi.

Carbon Compounds: Zn(CH3)2, Zn(C2H5)2, Zn(C5Hn)2,
A1(CII3)3, A1(C2H5)3, Sn(CH3)4, Sn,(C2H6)c, Fe(CO)6,
Ni(CO)5.

The data are too scanty to give trustworthy values for the
atomic weights of the metals.

B. ATOMIC HEATS

The chief method for determining the atomic weights of the
metals is by means of the Law of Dulong and Petit (1819),

specific heat x atomic weight = atomic heat = constant.
The value of the "constant" is about 6 for the range from 100°

1 In 1812 Berzelius regarded nitrogen as an oxide, and concluded
that this ratio was six times ; in 1819 he gave the ratio as six times or
five times, according as nitrogen was regarded as a compound or as an
element.

xvi ATOMIC WEIGHTS OF THE METALS 381

to 20°, but falls to 5 in the range from 20° to - 188° and to 2
between -188° and -253°. In the lowest range the atomic
heat varies very widely and is roughly proportional to the atomic
volume ; but at ordinary temperatures the atomic heats of all the
elements except boron and carbon fall between 5*2 and 67 ;
even these two elements give " normal " values at higher
temperatures.

In the case of solid compounds, Kopp (1865) found that "each
element has essentially the same specific or atomic heat in com-
pounds as it has in the free state."

The Law of Atomic Heats only leads to very rough values for
the atomic weight, but is sufficiently exact to give the valency,
i.e. the integral ratio by which the equivalent of an element
must be multiplied in order to give its atomic weight.

C. ISOMORPHISM

Mitscherlich, in 1819, discovered that identical crystalline
forms were shown by the following pairs of isomorphous salts :

J KH2PO4 1 /NaH2PO4,H2O \ /(NH4)H2PO4 \
\KH2AsO4J tNaHjAsO^HjO/ \ (NH4)H2AsO4 / '

The sodium phosphate also crystallised in a second form and
was therefore dimorphous.

In series of isomorphous compounds the isomorphous elements
replace one another atom by atom ; the analysis of isomorphous
salts can therefore be used to determine atomic weights. Thus
in the phosphates and arsenates 75 parts of arsenic take the
place of 31 parts of phosphorus ; therefore if P =31, then
As = 75.

Isomorphous salts of similar composition possess the property
of forming isomorphous mixtures or solid solutions, i.e. compo-
site crystals in which the different constituents are mixed as
intimately as in a liquid solution. If these mixtures can be formed
in any proportions the isomorphism is complete (compare water
and alcohol): if one substance will only tolerate a limited amount
of the other (compare water and ether) the isomorphism in incom-
plete. An example of incomplete isomorphism is found in the
potassium and ammonium phosphates formulated above, which
will only mix with one another to the extent of about 20%.

382 HISTORICAL INTRODUCTION TO CHEMISTRY CH. xvi

In the cubic system isomorphism is perfect, i.e. the angles for
the different substances are exactly equal. Examples are

(1) the alums MIMIII(SO4)2,i2H2O.
where M1 == K, Na, NH4, &c.

M'ni =A1, Fe(ferric), Cr, &c.

(2) Spinel MgAl2O4 Also MgFe2O4
Gahnite ZnAl2O4 MnFe2O4
Magnetite FeFe2O4 FeAl2O4
Chromite FeCr2O4 &c.

In less symmetrical crystals the isomorphism is imperfect, but
mixed crystals may be formed if the angles do not differ more
than 5°. The rhombohedral angles of a series of isomorphous
carbonates are given on page 369.

Isomorphism often occurs in compounds of dissimilar structure,
e.g. CaCO3 and NaNO3, Mg2SiO4 and Al2BeO4, and even in
compounds which contain unequal numbers of atoms in the
molecule, e.g. PbS and Ag2S or (NH4)Al(SO4)2,i2H2O and
KAl(SO4)2,i2H2O. A remarkable collection of cases of isomor-
phism (in which fluorine replaces oxygen, water replaces chlorine,
&c., without producing any change of crystalline form) has been
made by T. V. Barker (Trans. Chem. Soc., 1912, 101, 2484).

D. FORMUL/E OF SALTS

In the case of salts it is usually possible to give only an em-
pirical formula, which expresses its composition, but not its
molecular weight. Such a formula can be deduced readily from
the analytical data by methods indicated on pp. 373 to 380.

CHAPTER XVII

MOLECULAR ARCHITECTURE

A. THE RISE OF ORGANIC CHEMISTRY

The development of inorganic chemistry, 1766-1816. —

It can scarcely fail to be noticed that almost the whole of
the work described in the sixteen preceding chapters falls
within a very narrow period of time. Occasionally it has
been necessary to go back to study the origins of things, or
forward to study their more mature developments ; but by
far the greater part of the experiments that are described
were carried out during the latter part of the eighteenth and
the early years of the nineteenth century.

This narrow period, undoubtedly the most fertile in
the history of chemistry, begins with Black's work on
" Magnesia Alba," published in 1755. If this investigation
be set aside as merely a herald of the dawn, it will be found
that almost every chapter is filled with accounts of experi-
ments made in the fifty years from 1766 to 1816. The
former date saw the publication of Cavendish's papers
" On Factitious Air," and was followed almost immediately
by the wonderful development of Pneumatic Chemistry
which we owe to the successful labours of Priestley. The
latter date saw the completion of Gay-Lussac's work on the

383

384 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

oxides of nitrogen and his final solution of the problem of
the composition of nitric and nitrous acids. On the practi-
cal side, it may be noticed that by the year 1816 all the
commoner gases had been prepared and analysed, some fifty
different elements had been recognised clearly, water had
been decomposed and recomposed, the alkalis and most of
the earths had yielded up their metals, and the composition
of all the important mineral acids had at last been estab-
lished. On the theoretical side, it will be sufficient to
notice that, by the year 1816, the atomic and molecular
theories had been enunciated by Dalton and by Avogadro,
and that the Laws of Chemical Combination to which they
lead had been tested by Berzelius and proved to be correct.
The brief period of fifty years had thus sufficed, not merely
to lay the foundations of Inorganic Chemistry, but also to
reveal all the main outlines of the edifice.

The rise and development of organic chemistry, 1815-
1865. — It now remains to condense into a single chapter the
work of the succeeding fifty years, which witnessed a corres-
ponding development in the daughter science of Organic
Chemistry. For the sake of convenience the two fifty-year
periods may be made to overlap by a single year. ^3Phe second
of the periods will then extend from 1815, when Gay-Lussac
discovered the cyanogen radical and isolated cyanogen gas,
up to 1865, when Kekule first put forward his structural
formula for benzene.

The period begins with some of the earliest accurate
analyses of organic compounds, leading at once to correct
formulae, based upon the theories of Dalton and of Avogadro,
for compounds such as alcohol and ether. Unfortunately,
a few obstinate exceptions threw discredit upon Avogadro's
hypothesis, with the result that, in the matter of formulae,
darkness and confusion reigned almost all through the fifty
years that we are now considering. The marvellous energy
of the many workers, coupled with the complex and un-
familiar formulae which they used, render this period

xvn MOLECULAR ARCHITECTURE 385

only less difficult to study than the alchemistic period
itself.

The supreme feature of the period was the gradual
development of structural chemistry, leading at last to a
clear knowledge of the way in which the atoms are linked
together, even in the more complex products of vegetable
and animal life. But apart altogether from all progress
in the region of theory, there was a marvellous growth in
the experimental knowledge of organic compounds. In
this period we find recorded the discovery, or the first
analysis, of methyl alcohol and its higher homologues, of a
series of fats and fatty acids, of acetone and the ethers, of
chloral and chloroform, glycerol and glycol, aniline and the
amines, camphor and camphoric acid, benzene and naphtha-
lene, and the many products obtained by chlorinating,
brominating, nitrating and sulphonating these different
materials. The organic chemist of to-day will recognise in
this imperfect list evidences of advance along most of the
principal routes by which the growth of knowledge is pro-
ceeding in his science at the present time ; and the structural
formulae assigned by Kekule to ethyl acetate and to benzene
are but typical of those which are now being assigned, fifty
years later, to the more complex terpenes and alkaloids.

Lavoisier (1784) shows that organic substances are
compounds of carbon and hydrogen.— The formation of
fixed air and of water as products of the combustion of
organic materials had been recognised vaguely from the
time of Van Helmont (p. 64). In the case of fixed air the
information first became definite in the hands of Black,1

1 In his Lectures on the Elements of Chemistry, published in 1803,
after his death, Black says that in the year 1757 he had discovered or
proved : —

(1) That fixed air "is deadly to all animals that breathe it by the
mouth and nostrils together."

(2) That " the change produced on wholesome air by breathing it
consisted chiefly, if not solely, in the conversion of part of it into fixed

C C

386 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

of Cavendish (p. 95), and of Priestley (p. 95). The pro-
duction of water during the burning of spirit of wine was
frequently recorded, right up to the time of Lavoisier, who
in 1784 obtained 18 ounces of water by burning 16 ounces
of spirit (Works, II. 358). But these observations could
not be used to explain the composition of organic compounds
until the experiments of Lavoisier (p. 97) and Cavendish
(p. 113) had disclosed the composite character both of fixed
air and of water. When these substances had been proved
to be " oxide of carbon " and " oxide of hydrogen " respec-
tively, Lavoisier was in a position to describe correctly the
nature of combustible organic substances as compounds of
carbon and hydrogen ; indeed, some years before these
names were introduced, he had already (pp. 145-148) made
analyses to determine the proportions of carbon and hydro-
gen in spirit of wine, in wax, and in olive oil. In his
Elementary Treatise, published in 1789, Lavoisier was able
to enumerate the chief elements present in organic com-
pounds, usually in combination with oxygen :

" I have already observed that, in the mineral kingdom,
almost all the oxidisable and-acidifiable radicals were simple ;
that, in the vegetable kingdom on the contrary, and above
all in the animal kingdom, there were scarcely any which
were composed of less than two substances, hydrogen and
carbon ; that often nitrogen and phosphorus were united
with them, and that there were produced radicals with four
bases" (Works, I. 147).

Thus, "the acetic radical is formed by the union of

of caustic alkali, the lime was precipitated, and the alkali was rendered
mild."

(3) That "fixed air is the chief part of the elastic matter which
is formed in liquids in the vinous fermentation."

(4) That the " deadly vapour of burning charcoal . . . must be fixed
air," since a piece of red-hot charcoal, inserted in the nozzle of a pair
of chamber-bellows, imparted to the air which passed over it the power
to render lime-water milky (Lectures, 1803, II. 87-88).

These observations were not published for nearly fifty years, though
manuscript copies of the lectures (which can still be purchased) appear
to have been sold to the students by the lecture-assistant.

xvn MOLECULAR ARCHITECTURE 387

carbon and of hydrogen brought to the state of acid by the
addition of oxygen. This acid is, consequently, composed
of the same principles as tartaric acid, malic acid, etc. but
the proportion of the principles is different for each of
these acids, and it would seem that acetic acid is the most
oxygenated of all " ( Works, I. 218).

Alcohol and ether. — It would be difficult to exaggerate
the importance to organic chemistry of the process of vinous
fermentation. Fermented liquors have been known from
time immemorial. The spirit or " aqua vitse " prepared
from them by distillation was known in the time of Geber.
A minute description of the methods of purifying it is given
by Raymond Lully (1235-1315). The spirit, prepared
from dark wine, was subjected to repeated fractional distil-
lation ; in the first distillation only one-tenth of the distillate
was collected, then one-fourth, one-third, one-half, and
finally (in 14 further distillations) almost the whole of the
volatile distillate. The spirit was also dried by distilling
from calcined tartar (dry potassium carbonate). There
can be little doubt that by such methods Lully would
obtain almost pure ALCOHOL. Lully also tried the action of
oil of vitriol on alcohol, and was one of the first to convert
it into the still more volatile ETHER. The further action of
oil of vitriol on alcohol, whereby it is converted into a gas,
OLEFIANT GAS or ETHYLENE, was referred to by Becher in
1669 andhas already been described in detail (pp. 154-156).

In view of its unique importance, it is not surprising that
alcohol was selected by Lavoisier as the subject of his first
organic' analysis (pp. 146-148). Both alcohol and ether
were analysed by de Saussure in 1807 (Ann. de Chimie,
1807, 62, 225-241), by a variety of methods, but with
no great measure of success; in 1814, however (Ann. de
Chimie, 1814, 89, 273-305), he obtained a very good
analysis of alcohol, by first decomposing it in a red-hot
tube, and a less accurate analysis of ether, by exploding

c c 2

388 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

it directly with oxygen in a eudiometer. The results are
shown below :

ALCOHOL. ETHER.

Lavoisier, de Saussure. Calculated. de Saussure. Calculated.
Carbon . 28*5 52*0 52*2 68*0 64*9

Hydrogen 17-4 137 13-0 14-4 13-5

Oxygen . 54-1 34-3 34*8 17-6 21 '6

The vapour densities of alcohol and ether were deter-
mined by Gay-Lussac in 1815 (Ann. de Chimie, 1815, 95,

Lavoisier's work on vinous fermentation (1789).—
More important than the isolated analysis of spirit of wine
is Lavoisier's complete investigation of the process of vinous
fermentation, as described in his Treatise, in 1789 (Works,
I. 100-108) Lavoisier was able to show that

" The effects of fermentation reduce themselves to a sep-
aration of the sugar, which is an oxide, into two portions,
oxygenating one portion at the expense of the other to form
carbonic acid ; deoxygenating the other in favour of the first
to form a combustible substance, which is alcohol ; so that,
if it were possible to recombine these substances, alcohol and
carbonic acid, sugar would be reproduced " ( Works, I. 107).

Lavoisier took 100 Ib. of sugar, 400 Ib. of water and 10
Ib. of yeast-paste containing about 7 Ib. of water. At the
end of some days 35 Ib. of carbonic acid had escaped,
carrying with it 14 Ib. of water. There remained 460 Ib.
of liquor, containing 409 Ib. of water, 58 Ib. of alcohol,
2 Ib. of acetic acid, 4 Ib. of sugar and i Ib. of (dry) yeast.
Lavoisier worked out a balance-sheet as follows :

Before Fermentation : —

Hydrogen. Oxygen. Carbon. Nitrogen. Total.

Water . 6i.i.3l 346. 2.4 = 407. 3.7

Sugar . 8.0.0 64. o o 28. o.o = 100. o.o

Yeast . 0.4.5 1. 10. 2 0.12.5 0.0.5 = 2.12.1

69.6.0 411.12.6 28.12.5 0.0.5 = 510. 0.0

1 8 gros=i ounce, 16 ounces = i pound; the grains have been
omitted from the table.

xvn MOLECULAR ARCHITECTURE 389

After Fermentation : —

Hydrogen.

Oxygen.

Carbon.

Nitrogen.

Total.

Carbonic acid
Water .

61.5.4

25. 7.1

347. 10. i

9.I4-3

z

= 35- 5-4
= 408.15.5

Alcohol .

9.9.2

31. 6.2

16.II.6

—

= 57-H.2

Acetic acid

0.2.4

I.II-4

0.10.0

—

= 2. 8.0

Sugar . r

0.5.2

2. 9-7

I. 2.3

—

= 4- 1-4

Dry yeast

0.2.3

O.I3.I

0. 6.2

0.0.3

= i. 6.1

71.8.7 409.10.0 28.12.6 0.0.3 =510. 0.0

Lavoisier's chapter on " Vinous Fermentation " will always
be remarkable for the fact that it contains one of the first
clear statements of the law of CONSERVATION OF MASS during
chemical change, sometimes described also as the INDE-
STRUCTIBILITY OF MATTER, and one of the first descriptions
and illustrations of a CHEMICAL EQUATION. The inde-
structibility of the atom had already been postulated by
Newton (p. 292). The proof that the total mass is unaltered
by chemical changes is closely bound up with the proof
that heat is imponderable ; this proof had been given by
Cavendish for the explosion of hydrogen and oxygen (p. 113)
and by Lavoisier himself for the calcination of tin in closed
vessels (pp. 35-38). Lavoisier was therefore able to enunciate,
as a guiding principle for his work on fermentation the law
that :

" Nothing is created, either in the operations of art, or in
those of nature, and it may be set out as a principle that, in
every operation, there is an equal quantity of matter before
and after the operation ; that the quality and the quantity
of the principles is the same, and there are only changes,
only modifications."

"On this principle is based the whole art of making ex-
periments in chemistry : in all of them one must suppose a
true equality or equation between the principles of the body
which one examines and those which one obtains from them
by analysis. Thus, since grape-juice gives carbonic acid gas
and alcohol, I can say that grape-juice = carbonic acid +
alcohol" (Works, I. 101).

390 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

Organic analyses of Gay-Lussac and Thenard (1810), —

The foundations of organic analysis were laid by Lavoisier
in the two papers that have already been quoted (pp. 145-
147) ; but it was not until many years later that simple and
accurate methods became generally available. Some of the
earlier workers, including Berth ollet, relied upon decompos-
ing the organic compounds by heat into charcoal, which
could be weighed, and gaseous products which could be
analysed in the eudiometer.

In 1810 a general method of analysis was worked out by
Gay-Lussac and Thenard (Phy si co- Chemical Researches,
1811, II. 265-350) and applied to four substances contain-
ing nitrogen and fifteen substances free from nitrogen. The
substance was mixed with a known weight of potassium
chlorate, made into pellets and burnt by dropping into a hot
tube (Fig. 49). The gases produced by burning the pellets
(oxygen, carbonic anhydride, and nitrogen, if present) were
collected and analysed. From the volume of carbonic
anhydride the weight of carbon in the substance could be
calculated ; a similar calculation gave the weight of nitrogen.
Blank experiments showed what volume of oxygen was set
free by igniting the chlorate alone. In several cases the
deficiency of oxygen was exactly equal to that due to the
formation of carbonic anhydride ; any part of the weight
of the substance not accounted for after adding together the
carbon and nitrogen, together with other easily recognised
elements, such as chlorine, phosphorus and the metals, was
then regarded as water. A greater or less deficiency of
oxygen indicated the presence of an excess of hydrogen or
of oxygen, above the proportions required to produce water.

The analyses of Gay-Lussac and Thenard revealed the
remarkable fact that six of the substances analysed (sugar,
gum, starch, milk-sugar, oak-wood, and beech-wood) con-
tained carbon united with hydrogen and oxygen in just the
proportions in which they are required to form water. Such
compounds are now known as CARBOHYDRATES. These six

XVII

MOLECULAR ARCHITECTURE

391

substances were all neutral. Five acids (mucic, oxalic,
tartaric, citric, acetic) contained an excess of oxygen. Four
resinous and oily substances (resin, copal, wax, olive-oil)
contained an excess of hydrogen.

5 DECIMETRES

FIG. 49. — GAV-LUSSAC AND THENARD'S APPARATUS FOR THE COMBUSTION OF
ORGANIC COMPOUNDS.

The tube A A', of hard glass, 20 cm. long and 8 mm. wide, carried a side-tube
JSB', leading into a trough of mercury. A copper cap CC', cemented to the top
of the tube A A', carried a glass tap DD' (shown on a larger scale as an inset)
and a cup of broken ice Ff, to prevent the grease on the tap from melting down
into the hot tube. The tap DD' was not pierced in the usual way, but had a
small cavity, in which a pellet could be inserted, and dropped into the hot tube
without opening it to the air. Several pellets were burnt to displace the air
from the apparatus, and the gases set free by burning a known weight of the dry
pellets were then collected, measured and analysed.

Berzelius (1815) assigns formulas to organic compounds. —

In 1815 Berzelius described a new series of organic analyses
which he carried out " in order to determine how far the
laws which I had established in inorganic nature could be

392 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

hn -w ti

no.s a

applied to organic
bodies " (Ann. de
Chimie, 1815, 94,
1-28, 170-190, 296-
323, 95, 51-90).

These analyses
were made by a
modified form of the
method of Gay-
Lussac and Thenard,
in which the carbon
and the hydrogen
were determined
directly, by weighing
the water and the
carbonic anhydride
produced by combus-
tion. The water was
caught in a small
glass bulb, followed
by a tube filled with
calcium chloride ; the
carbonic anhydride
was driven forward
into a bell-jar filled
with mercury, where
it was absorbed in a
small bulb containing
solid potash (Fig. 50).
The substance was
burnt by mixing it
with potassium chlor-
ate, copiously diluted
with common salt
to restrain the viol-
ence of the combus-

XVII

MOLECULAR ARCHITECTURE

393

tion; potassium chlorate, not mixed with the substance, was
placed at both ends of the combustion-tube, in order to
provide an ample supply of oxygen (i) to complete the burn-
ing, (2) to drive the products of combustion forward into
the absorption-apparatus. Extreme care was needed to dry
perfectly all the materials with which the combustion-tube
was charged.

Berzelius analysed nine organic acids in the form of
their lead salts, and four carbohydrates, thus covering much
the same ground as Gay-Lussac and Thenard. But, unlike
them, he was able to check his analyses by the requirements
of the newly-established atomic theory and to express the
composition of his compounds by means of formulae.
Berzelius's formulae for the acids are usually those of the
lead salt minus lead oxide, or of the acid minus water ;
apart from this, they agree very well with those adopted
to-day, the few differences that occur being due to an over-
estimation of the hydrogen. This is shown by the following
comparison :

Rerzelius's

Modern formulae.

C6H807 -H10 = C.H606

C4H606 -H20 = C4H405

C2H2O4 -H2O = C2O3

C4H604 -H,0 = C4H40.

2C2H402 -H20 = C4H603
C6H603 (pyrogallol)
C6H100

Berzelius's

formulae.

Citric acid

CHO

Tartaric acid

C4H505

Oxalic acid
Succinic acid

C12H018
C4H403

Acetic acid

C4H603

Gallic acid

C6H603

Mucic acid

C6H1008

Benzoic acid

Tannin .

QHJA

Sugar

Cj2H21Oj0

Milk-sugar

CH2O

Gum

CioH24O12

Starch (-. :

. C7H1306

C7H602

C12H22O
CBH100B

wso
= [CH20]12

Combustion by means of metallic oxides, — It should be
noticed that metallic oxides were used for the quantitative
combustion of organic substances by Lavoisier, who used red
lead (p. 146), as well as by Berzelius, who used the brown
peroxide in his earlier analyses. The oxides of leads were
discarded in favour of potassium chlorate, because they

394 HISTORICAL INTRODUCTION TO CHEMISTRY CH. xvn

could not easily be freed from moisture and from carbonic
anhydride. The black oxide of copper was first used by Gay-
Lussac, in 1815, for the combustion of mercuric cyanide
(Ann. de Chimie, 1815, 95, 184), and in the hands of
Liebig (Pogg. Ann. der Physik, 1831, 21, 1-43) soon became
the standard oxidising agent in organic analysis (see Fig. 51).

B. THE STRUCTURE OF SALTS.

Lavoisier's oxygen-theory of the structure of salts.
— The idea that compounds possess a definite structure
originated from observations on the interactions of acids
and bases to form salts. Mayow, in considering this action,
concluded (A.C.R. XVII. 160) that "although [acids] and
alkalis pass into a neutral substance when they meet, yet
they do not, as is generally supposed, entirely destroy each
other." Each grain of salt must then contain, intimately
united within it, a certain portion of acid and a certain
portion of alkali. Lavoisier elaborated this idea by regard-
ing the acid and base as themselves compounded of a non-
metal or a metal united with oxygen, so that the structure
of the salt might be represented as :

SALT = ACID -f BASE

= {NON-METAL + OXYGEN} + {METAL + OXYGEN}

" The acidifiable substances, when combined with oxygen
and converted into acids, acquire a great tendency to
combination ; they become capable of uniting with earthy
and metallic substances, and it is by this union that
neutral salts are produced" (Treatise, 1789; Works, I.
"5).

Berzelius's dualistic theory (181&). — Lavoisier's theory
broke down when it was discovered that acids (such as
muriatic acid) and bases (such as ammonia) existed, in
which no oxygen was present, but its essential features were
revived by Berzelius in his "Theory of Chemical Propor-

Ill I
111 *

£b= fi

is-g

.

8«
*

.5

8 <n

o
il

«! «

£ ,S rtJ? 2 s £2 >'H y

g His II 5*1S|

2 ^ ag O-Q "rt-5 ajs o

S .§3 J, ,5 J° 0'$ So g

i life

1.9 I --3^

111338 ll-sl3*

|||]|| j.;|||j:

w-a'o.o « S ^3 c5T S2

-S2

396 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

tions " 1819. Speaking in terms of Dalton's Atomic Theory
he suggests that

" When atoms of two different substances combine, a
compound atom is produced, in which we suppose the force
producing combination to be infinitely stronger than all the
conditions which might tend to separate the united atoms.
This compound atom must be considered to be as indivisible
by mechanical means as an elementary atom."

" These compound atoms combine with other compound
atoms, giving rise to atoms still more compound in their
nature. When these in their turn combine with others,
they produce atoms of still more complex composition.
These different atoms must be distinguished. We will
divide them into atoms of the first, second, third order, etc."
(Chemical Proportions, 1819, p. 26). As an example he
quotes the case of alum :

Fourth Order.

VAlum crystals

Oxygen \ Water

Hydrogen / J

The chief feature of Berzelius's theory of the structure of
salts is the combination of elements and compounds in
pairs ; it is therefore frequently described as the DUALISTIC
THEORY. Although it was applied in the first place only
to the combination of elements with oxygen, and^of basic
with acidic oxides, it was easily extended to include the
combination of metals with chlorine, even when the chlorides
were proved to be binary compounds, analogous with the
oxides, instead of ternary compounds analogous with the
sulphates. Further, in his " Memoir on the Sulpho-salts,"
(Ann. Chim. Phys. 1826, 32, 60) Berzelius had no difficulty
in showing that these salts, formed by the combination of

ELEMENTS.

First Order.

COMPOUNDS.

Second Order. Third Order.

Oxygen
Potassium

Oxygen
Sulphur

)

!

Potash

Sulphuric
anhydride

(.Sulphate of "\
j Potash

rk

>ry Alum

xvn MOLECULAR ARCHITECTURE 397

two sulphides, were strictly analogous with the oxy-salts
formed by the combination of two oxides.

Berzelius's electro-chemical theory (1819). — Berzelius's
theory included one other unifying feature to which
attention must be directed. Berzelius arid Hisinger (p. 273)
had been the first to show that the electric battery could
produce, not only a decomposition of water, but also a
resolution of the dissolved salts into acid and base. This
observation^ was soon followed by Davy's discovery of the
decomposition of the alkalis and earths into metal and
oxygen when the same agency was applied to these sub-
stances in the absence of water. As the electric current
was thus proved to be the most powerful agency for resolv-
ing chemical compounds, Berzelius concluded that chemical
affinity was electrical in character.

" We are now confident that substances on the point of
combining, exhibit opposite electric charges, increasing in
strength as they approach the temperature at which com-
bination takes place, until, at the moment of union, the
charges disappear with a rise of temperature which is often
so great that fire blazes out. On the other hand, we are
equally confident that compound substances, exposed in a
suitable form to the action of the electric fluid produced
by the discharge of the pile, are separated and recover their
original chemical and electrical properties, whilst the charges
acting on them disappear."

" In the actual state of our knowledge, the most probable
explanation of combustion and of the ignition resulting
therefrom, is then : that in all chemical combination there is
a neutralisation of opposite electric charges, and that this
neutralisation produces fire in the same way as in the dis-
charge of a Ley den jar, an electric pile or lightning, although
not accompanied, in these last phenomena, by a chemical com-
bination" (Chemical Proportions •, 1819, 72-73).

Berzelius's ELECTRO-CHEMICAL THEORY, as thus enunciated,
accounted not only for chemical combination, but also for
the liberation of light and heat, thus destroying the necessity

398 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

which Lavoisier had felt for postulating the existence of
" matter of fire " and " matter of heat " as imponderable
constituents of substances, set free from them when they
combined together. In this attempt at unification Berzelius
anticipated, not only the science of Thermo-chemistry, which
attempts to measure the energy of chemical combination in
units of heat, but also the later and more accurate method
of expressing energy of combination in terms of electro-motive
forces.

C. THE THEORY OF COMPOUND RADICALS.

Structure of organic compounds. Presence of compound
radicals. — Berzelius's dualistic electro-chemical theory was
based upon his observations of the decomposition of mineral
substances by the electric current, and was applied primarily
to salts, acids and bases, which yielded readily to this treat-
ment. But it threw very little light upon the structure of
organic substances, such as sugar, alcohol, or olive-oil, which
would not conduct the current and were not decomposed
by it. Lavoisier, who regarded the mineral acids as binary
compounds of oxygen with a simple elementary radical such
as carbon, sulphur or phosphorus, had been obliged to
recognise that the organic acids contained oxygen in combin-
ation with both carbon and hydrogen, and sometimes also
with nitrogen and phosphorus ( Works, I. 147). These com-
pound radicals could not be resolved by the dualistic theory
and were frankly recognised by Berzelius as demanding
exceptional treatment.

" [Compounds] of the first order are composed of simple
elementary atoms ; they are of two kinds, organic and
inorganic. /The latter never contain more than two
atoms; the latter always contains at least three" {Chemical
Proportions, 1819, p. 26).

" In inorganic nature all oxidised bodies contain a simple
radical, while all organic substances are oxides of compound

xvn MOLECULAR ARCHITECTURE 399

radicals. The radicals of vegetable substances consist
generally of carbon and hydrogen, and those of animal
substances of carbon, hydrogen and nitrogen" (Textbook,
1817, I, 544).

Gay-Lussac on the cyanogen radical (1815). — An excel-
lent example of a compound radical of organic origin was
available in the case of prussic acid, which Gay-Lussac had
described in 1815 as a compound of hydrogen with the com-
pound radical CYANOGEN (p. 245). The analogy between the
compound organic radical cyanogen and the simple inorganic
radical chlorine was remarkable. It could combine, not
only with hydrogen, but with chlorine and with metals such
as potassium, silver and mercury; it could also exist alone
in the form of a gas, which was evidently the analogue of
chlorine gas. The similarity of the two series of compounds
may be shown by writing their formulae, as is often done
with the symbol Cy (Berzelius, Jahresbericht, 1839, 18,
120), to represent the compound-radical CN. Thus we
have

Cyanogen, CN or Cy. Chlorine, Cl.

Cyanogen gas C2N0 or Cy2 ^ r^ui • r-i

rt LI -j /^i/^-ivT r*\r* f Chlorine gas . . vJL

Cyanogen chloride C1CN or CICy /

Prussic acid . . HCN or HCy Muriatic acid . . HC1

Potassium cyanide KCN or KCy Potassium chloride KC1

Silver cyanide . AgCN or AgCy Silver chloride . AgCl

Mercuric cyanide . HgC2N2 or HgCy2 Mercuric chloride . HgCl3

r ^ Wbhler and Liebig (1832) on the benzoyl radical, —
The importance of the theory of radicals became evident
whenWohler1 and Liebig, in 1832, published their "Re-
searches on the Radical of Benzoic Acid " (Liebig's Ann.
der Pharm.) 1832, 3, 249-282). In these researches they
showed that oil of bitter almonds could be converted into a

1 Wohler was already well-known for his preparation of aluminium
(Pogg. Ann. der Physik, 1827, 11, 146-161) and for his synthesis of
urea, a typical animal product, from inorganic materials (Pogg. Ann.
der Physik, 1828, 12, 253-256) ; the urea was obtained in attempting
to prepare ammonium cyanate, and is formed as a product of
"isomeric change," (NH4)CNO — CO(NII2).2.

400 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

series of compounds, all of which might be regarded as
derived from a compound radical, which they proposed
to call BENZOYL (loc. tit. p. 279). This radical combined
with hydrogen to form oil of bitter almonds, with oxygen
and hydrogen to form benzoic acid, with chlorine, bromine,
iodine, sulphur and cyanogen to form benzoyl chloride,
bromide, iodide, sulphide and cyanide. These compounds
may by represented by the following formulae : l

BENZOYL = C7H8O.

Oil of bitter almonds = benzoyl hydride . . = C^HgO *H

Benzoic acid = benzoyl hydroxide . . = CyHgO'OH

Benzoyl chloride = C7HBOC1

Benzoyl cyanide - C7H5O'CN

Wohler and Liebig also prepared :—

Benzamide . = C7H5O-NH2

Ethyl benzoate = C7H5O'OC2H5

Benzoin, a solid having the same composition as
oil of bitter almonds, but now regarded as a
polymer (C7H5OH)2 = C14H12O2

Berzelius was delighted with the discovery made by
Wohler and Liebig. The compound radicles which had
been studied hitherto had been binary compounds of two
elements, e.g., cyanogen, CN, ammonium, NH4 ; jthe
benzoyl group was the first example of a compound radical
containing three elements, which nevertheless showed very
many of the properties of a simple substance. In a well-
known letter to Wohler and Liebig, dated from Stockholm,
Sept. 2, 1832, and published at the conclusion of their paper
(Liebig's Ann. der Pharm., 1832, 3, 282-287), Berzelius
writes :

"The facts that you have ascertained suggest so many
considerations, that they may well be regarded as the begin-
ning of a new day in vegetable chemistry. I would therefore
propose, to call the first example of a compound radical
containing more than two substances Proin (from Tr/awi', the

1 Wohler and Liebig doubled these formulae and wrote benzoyl =
C14H10O2 ; chlorobenzoyl (benzoyl chloride) = C14H]0O2C12, etc.

xvil MOLECULAR ARCHITECTURE 401

beginning of the day. . .) or Orthrin (from opOpbs, dawn of
the morning) " (loc. cit. p. 285).

As in the case of the ammonium radical, which he had
represented by the symbol Am, Berzelius proposed to repre-
sent the benzoyl radical by a single symbol, Bz, thus

Oil of bitter almonds = BzH
Benzoyl chloride = BzCl, etc.

He also proposed to describe by the name AMIDE the
compound radical NH2, which is formed by removing an
atom of hydrogen from ammonia as in

C7H5ONH2 = Benzamide,

K'NH2 = Potassamide,

Na'NH2 = Sodamide.

The structure of alcohol and ether. Dumas and
Boullay (1827) on the aetherin radical. — The action of
sulphuric acid on alcohol had long been known to give rise
on one hand to olefiant gas (Chap. VIII, p. 155), and on
the other hand to a volatile, inflammable liquid, which was
generally described as " sulphuric ether," but is now known
simply as "ether." Gay-Lussac in 1815 (Ann. de Chimie,
JL&LS, 95, 311-318), by measuring the vapour-densities of
alcohol and of ether, showed that their composition might
be represented as follows :

Alcohol = Olefiant Gas + Water-vapour
i vol. i vol. i vol.

Ether = Olefiant Gas + Water-vapour

I Vol. 2 Voh. I Vol.

In an important memoir " On the Compound Ethers " *
(Ann. Chim. P/tys., 1828, 37, 15-53), Dumas and Boullay
showed further that the ethereal substances formed by the
action of acids on alcohol could all be represented as com-

1 Read at the Paris Academy of Sciences on Dec. 24, 1827, and
published in the Memoires de F Institut, 1838 (!), 15, 457-494.

D D

402 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

pounds of the acids with olefiant gas formed by the removal
of water from the alcohol. They concluded :

"That olefiant gas plays the part of a very powerful
alkali, endowed with a capacity of saturation equal to that
of ammonia, of which it would probably show most of the
reactions, if it were, like ammonia, soluble in water.

2. That alcohol and ether are hydrates of olefiant gas.

3. That the compound ethers are salts of olefiant gas,
etc." (loc. tit. p. 52).

The analogy of composition between the compound
ethers and the salts of ammonia was shown in a table, a
part of which (using modern atomic weights and formulae)
would read as follows :

Compound Ethers.* Ammonium Salts.

Alcohol, C2H4,H2O 3 Ammonium hydroxide, NH3, H2O

Sulphuric ether, 2C2H4,H2O 3 Ammonium oxide, 2NH3,H2O
Hydrochloric ether, Q,H4,IIC1 Ammonium chloride, NH3,HC1

Hydriodic ether, C2H4,HI Ammonium iodide, NH3,HI

Nitric ether, C2H4,HNO2 Ammonium nitrite, NH3,HNO2

Acetic ether, G,H4,C2H4O2 Ammonium acetate, NH3,C2H4O2

Sulphovinic acid2 C2H4,H2SO4 Ammonium bisulphate, NH3,H2SO4

Liebig (1834) on the ethyl radical. — Berzelius in dis-
cussing Wohler and Liebig's paper "On the Radical of
Benzoic Acid," adopted the view of Dumas and Boullay in
reference to the composition of alcohol and the ethers. He
suggested (Ann. der Pharm., 1832, 3, 286) that the group
of atoms present in olefiant gas should be described as
AETHERIN and represented by the symbol Ae. Liebig him-
self (" On the Constitution of Ether and its Compounds,"
Ann. der Pharm., 1834, 9, 1-39) adopted a different view.

1 The compound ethers were named after the acids used in preparing
them but " sulphuric, phosphoric, arsenic ethers ... are identical
amongst themselves" (loc. cit. p. 15), being merely alcohol deprived of
water.

2 Examined just before by Dumas and Boullay ("Memoir on the
Formation of Sulphuric Ether," Ann. Chim. Phys., 1827, 36, 294-
310) as the first product of the action of sulphuric acid on alcohol.

3 Isolated in 1909, see p. 287.

MOLECULAR ARCHITECTURE 403

Just as Ampere in 1816 had regarded the salts of ammonia
as formed by the union of " ammonium " (ammonia and
hydrogen) with chlorine, iodine, sulphur, etc., so Liebig re-
garded the compound ethers as formed by the union of ETHYL
(aetherin and hydrogen) with chlorine, bromine, iodine, etc.
If aetherin be represented by the modern formula, C2H4, of
olefiant gas, then ethyl must be represented as C2H5, thus :

ammonium = NH3 + H = NH4 = Am
ethyl = C2H4 + H = C2H5 = Et

On this view

alcohol = ethyl hydroxide = (C2H5)OH or EtOH

ether = ethyl oxide = (C2H5)2O or Et2O

muriatic ether = ethyl chloride = (C2H5)C1 or EtCl

just as, according to Ampere's view,

sal-ammoniac = ammonium chloride = (NH4)Cl.or AmCl
Actually, Berzelius and Liebig doubled the formulae of these
radicals and wrote

aetherin = C4H8 = Ae
ethyl = C4H10 = E

ether = C4H10O = Ae + H2O, or EO

alcohol = C4H12O2 = Ae + 2H2O or EO + H2O

ethyl chloride = C4H10C12 = Ae + H2Cl2 or E + C12.

It will be seen that in doing this they broke away finally
from the idea that the relative sizes of the molecules could
be deduced from measurements of vapour density ; thus,
alcohol, with a lighter vapour than ether, was now repre-
sented as a compound of ether and water, and all their
derivatives were regarded as containing at least four atoms
of carbon.

Dumas and Peligot (1834) discover a new alcohol in
spirit of wood. Methylene and methyl as radicals.—
In 1834 Dumas and Peligot read to the Academy a paper
" On a New Alcohol and on the different Ethereal Com-
pounds derived from it" (Mem. Inst^ 1838, 15, 557-632).

D D 2

404 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

They showed that spirit of wood 1 contained a compound in
which they " recognised all the characters of a true alcohol."
In order to describe the new alcohol and its derivatives they
proposed to give the name METHYLENE (from " /xe'flv, wine,
and vA.^7, wood ; that is to say, wine or spirituous liquor of
wood") to the radical which it contained (loc. tit. p. 561).
They showed that
" i. Spirit of wood corresponds with alcohol.

2. By losing half its water, it forms a gaseous ether.

3. Its radical unites volume by volume with the
hydracids to form neutral anhydrous ethers, etc." (loc.cit.
p. 620).

They regarded the radical methylene, which they were
not able to isolate, as a hydrocarbon having the same com-
position as ethylene, but containing only half as many atoms ;
in just the same way, ethylene in its turn had the same
composition as the hydrocarbon (butylene) which Faraday
had isolated from oil-gas, but contained only half as many
atoms. Using modern formulae'2 the three compounds are
Methylene (not known) CH2

Ethylene (defiant gas) C2H4

Butylene (Faraday's hydrocarbon) C4H8
All the ethers derived from spirit of wood could be con-
sidered as formed by the addition of water or of acids to
methylene, but it was pointed out (loc. at. p. 625) that they
might also be regarded as oxides, chlorides, etc., of a
hypothetical radical, analogous with ethyl, to which the name
METHYL is now given. Thus, again using modern formulae,2

1 Discovered by Philip Taylor in 1812, and described by him ten
years later in a letter to the editors of the Philosophical Magazine, 1822,
60, 315-317-

2 Dumas formulated the three hydrocarbons as CH, C2H2, C4H4.
But these were multiplied by four in their compounds ; methylene
hydrate (methyl ether, C2H6O) was C4H4,H2O or C4H6O (C = 6);
spirit of wood (methyl alcohol, CH4O) was methylene bihydrate,
C4H4,2H2O or C4H8O2; ordinary alcohol (C2H6O) had now grown to
C8H]2O2, Dumas having followed Berzelius and Liebig in regarding
it as a hydrate of ether.

xvn MOLECULAR ARCHITECTURE 405

spirit of wood, CH4O

= methylene hydrate, CH2,H2O,

or methyl hydroxide, CH3'OH ;
its gaseous ether, C2H6O

= methylene hydrate, 2CHf,H9O,

or methyl oxide, (CH3)2O;
its gaseous chloride, CH3C1

= methylene hydrochloride, CH£,HC1,

or methyl chloride, CH3C1.

Dumas and Liebig united, — The discovery of METHYL
ALCOHOL, CH4O, and its derivatives greatly strengthened the
theory of radicals, since it showed that when applied to a
new group of ethers "in a series of very complicated
phenomena, one could predict everything, explain everything
and submit everything to calculation" (loc. tit. p. 621). It
had further become evident that the difference between
Dumas and Boullay's " ethylene " or " aetherin " and
Liebig's " ethyl " theory of the ethers was only of secondary
importance ; both radicals might, in fact, be regarded as
hydrides of the still simpler radical, C2H3, which Liebig
called "acetyl."1 In just the same way ammonia and
ammonium could be regarded as two hydrides of Berzelius's
"amide" radical, NH.^. Thus, in modern formulae

Amide = NH2 = Ad "Acetyl"1 =C2H3 = Ac

Ammonia =NH3 = AdH Aetherin (ethylene) = G>H4 = AcH
Ammonium =NH4 = AdH2 Ethyl = C2H5 = AcH2

(Liebig, Ann. der Pharm., 1839, 30, 139).

Having thus settled their main points of difference Dumas
and Liebig decided in 1837 to combine their forces in a great
campaign to establish a " natural classification of organic
compounds" by a detailed study of their component
radicals. A joint manifesto in reference to their programme
was issued in the form of a " Note on the present position
of organic chemistry" (Comptes rendus, 1837, 5, 567-572),

1 The word "acetyl" is now used to describe the radical C2H3O
(see page 431).

406 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

a joint report on the subject was to be prepared in response
to a request from the Liverpool meeting of the British
Association, and from 1838 to 1841 Dumas and Graham
appeared as collaborators with Liebig in the publication of his
Annalen. This happy collaboration was to be made effec-
tive by training a new generation of chemists ; it was there-
fore no mere coincidence that resulted, ten years later, in
the association together of Auguste Laurent (1807-1853,
pupil of Dumas) and Charles Gerhardt (1816-1856, pupil
of Liebig) in another campaign to establish a rational system
of atomic weights and formulae in chemistry ; only the early
death of these two workers prevented them from seeing the
triumph of their ideas through the masterly exposition of
Cannizzaro.

D. THE THEORY OF SUBSTITUTION.

Dumas (1834) on substitution or metalepsy. — In the

second part of his Researches in Organic Chemistry, " On
the Action of Chlorine on Alcohol. The Law of Substitu-
tions or Metalepsy," read before the Academy of Sciences
at Paris, on Jan. 13, 1834, Dumas directed attention to the
fact that

"Chlorine possesses the remarkable power of seizing
hold of the hydrogen of certain substances, and replacing
it atom by atom; This law of nature, this law or theory of
substitutions, has seemed to me to be worthy of a special
name. I propose to call it metalepsy, from //,€TaAr?i/as, which
expresses well enough that the body on which one acts, has
taken one element in the place of another, chlorine in place
of hydrogen, for example" (Mem. Inst., 1838,! 15, 548).

The laws of substitution were set out in the following
year in his Treatise on Chemistry as follows :

" i. When a hydrogenated compound is submitted to
the dehydrogenating action of chlorine, of bromine, of

1 This passage does not appear in the paper as published four years
earlier in the Ann. Chim. Phys., 1834, 56, 1-13-154.

xvn MOLECULAR ARCHITECTURE 407

iodine, of oxygen, etc., for each atom of hydrogen which
it loses, it gains an atom of chlorine, of bromine or of
iodine, or half an atom of oxygen ;

2. When the hydrogenated compound contains oxygen,
the same rule applies without modification ;

3. When the hydrogenated compound contains water,
this loses its hydrogen without replacement, and hereafter,
if a further quantity of hydrogen is removed, it is replaced
as in the preceding cases" (Treatise on Chemistry ', 1835, V.
99).

The following examples of simple substitution of chlorine
for hydrogen are given in his Memoir (loc. cit. pp. 549-556)
and quoted again in his Treatise (loc. cit. pp. 99-102).

(a) Gay-Lussac, in 1815, (Ann.de Chimie, 1815, 95, 210)
had shown that hydrogen cyanide, in passing into cyanogen
chloride " loses one volume of hydrogen and gains exactly
one volume of chlorine."

HCN + C12 = C1CN + HC1.

(b) Wohler and Liebig, in 1832 (see above p. 400), had
shown that oil of bitter almonds "treated with chlorine,
loses two volumes of hydrogen and gains precisely two
volumes of chlorine."

C7H60 + C12 = C7H5OC1 + HC1.1

(c) Faraday, in his experiments on the chlorination of
Dutch liquid^(ethylene chloride) (Phil. Trans^ 1821, 47-74),
had found that " chlorine, acting on it in sunlight, produces
a chloride of carbon, absolutely free from hydrogen." The
composition of this solid perchloride of carbon can be pre-
dicted by the law of substitution and agrees exactly with
Faraday's analysis.

C2H4C12 + 4C12 = C2C16 + 4HC1.

Removal of hydrogen without replacement. — The re-
servation in the third law of substitution was introduced to

1 These formulae were doubled: hence the "two volumes" of
hydrogen and chlorine.

408 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

account for the fact that hydrogen was sometimes removed
(wholly or in part) without replacement. Thus, in the con-
version of alcohol into chloral by the action of bleaching
powder it was seen that "the ten volumes of hydrogen,
removed from the alcohol, have been replaced by only six
volumes of chlorine," or, in terms of modern formulae, five
atoms of hydrogen have been replaced by only three atoms
of chlorine.

C2H60 + 4C12 = C2HC130 + 5HC1.

Dumas explained this on the view that alcohol was com-
posed of ethylene and water; the ethylene obeyed the
ordinary law of substitution, whilst the water (as Berthollet
had observed in 1785) gave oxygen and hydrogen chloride,
thus :

24 2 = C2HC13 +3HC1
H2O +C12 = O +2HC1

= C2HC130 + 5HC1

This removal of hydrogen without replacement was regarded
as a test for the presence of water, as distinct from other
forms of hydrogen, in the molecule. Thus, oxalic acid,
which loses hydrogen on oxidation to carbonic acid,

H2C204-H2=20)2,

was regarded as containing the whole of its hydrogen in the
form of water, and was written as a hydrate, CgO^HgO.1
Berzelius (1838-1843) objects to the presence of
oxygen and chlorine in organic radicals and to the whole
theory of the substitution of oxygen and chlorine for
hydrogen, — Berzelius, who had welcomed Wohler and
Liebig's discovery of the benzoyl radical as the dawn of a
new day in organic chemistry, soon repented of his rashness
in admitting oxygen as a constituent of an electro-positive
organic radical. In the particular case of oil of bitter

1 C4O3,H2O in the original (C-6).

xvii MOLECULAR ARCHITECTURE 409

almonds (benzaldehyde) he proposed to remedy this error
by regarding the compound, not as benzoyl hydride,

(C7H50)H,

but as an oxide, (C^H^O, of a radical PiCRAMYL=Pk =
C7H0,1 from Trixpo?, bitter, d/x,uySaA.r7, almond (Jahresbericht,
1843, 22, 328). This policy of robbing the radicals of their
oxygen had already been adopted in 1830 by Hermann, who
had suggested that

" The vegetable acids all share the characteristic that they are
compounds of oxygen with hydrocarbons. They are thus
all to be regarded as different stages of oxidation of hydro-
carbons. Thus vegetable acids = CH + #0 " (Poggendorfs
Annalen der Physik, 1830, 18, 396).

This view had been favourably received by Berzelius (fahres-
bericht) 1832, 11, 210) and was afterwards adopted by
Liebig (p. 405) when he gave to the radical C2H3 the name
"acetyl," thus implying that the radical of acetic acid was
a simple hydrocarbon.

But, if Berzelius thus objected to the presence of oxygen
and chlorine in the organic radicals, he had a still greater
objection to the idea that these elements could creep
in by a process of substitution for hydrogen.

. " An element so eminently electro-negative as chlorine,
can never enter an organic radical : this idea is contrary to
the first principles of chemistry ; its electro-negative nature
and its powerful affinities would prevent it from entering
except as an element in a combination peculiar to itself "
(Comptes rendus, 1838, 6, 633; Ann. Chim. Phys., 1838, 67,
309)-

In order to maintain this opinion, Berzelius was obliged,
in every actual case of substitution, to extradite the chlorine
from the electro-positive to the electro-negative part of the
molecule and to invent some new method of constructing
the radical. One example will suffice to illustrate his
1 C14H12, in the original.

4io HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

method. Malaguti (Comptes rendus, 1837, 5, 334) had
succeeded in introducing four atoms of chlorine into
ordinary ether.

C4H100 + 4C12 = C4H6C140 + 4HC1.

The simplest explanation was to suppose that chlorine
had entered the ethylene or ethyl radicals ; but this would
imply the co-existence in the electro-positive radical of hydro-
gen and chlorine. Berzelius preferred, therefore, to recon-
struct the whole molecule and to write it thus :

the molecule then no longer contained any trace either of
ethyl or of ethylene, but was now composed of methyl oxide
and carbon chloride (Comptes rendus, 1838, 6, 634; Ann.
Chim. Phys., 1838, 67, 310).

Chlorine, in organic compounds, " takes the place " and
" plays the part " of hydrogen. — Dumas at first accepted
Berzelius's view that the chlorination of an organic compound
might involve a reconstruction of the radicals which it con-
tained and in replying to Berzelius' letter emphasised the
fact that the theory of substitution " is an empirical rule "
(Comptes rendus, 1838, 6, 702).

"The theory of substitutions expresses then a simple
relation between the hydrogen that goes out and the chlorine
that enters. In the majority of cases this relation is one of
volume for volume. In announcing it, I believe that I
rendered a real service to science. Indeed, before it had
been signalised, there existed hardly a single exact analysis
of a compound formed by the action of chlorine on an
organic substance. Since it has fixed the attention of
chemists on this kind of reactions, the facts have multiplied,
the analyses have received a precision of which the impor-
tance has been recognised " (Comptes rendus, 1838, 6, 699).

It was not long, however, before the amazing complexities
of Berzelius's formulae, and the extreme simplicity of the
theory of direct replacement, compelled him to adopt the

xvn MOLECULAR ARCHITECTURE 411

view of his colleague Laurent that (as he expressed it some
years later)

" In the phenomena of substitution the type is conserved,
that is to say that not only does the chlorine take the place of
the hydrogen, but that it plays the same part" (Dumas,
"Note on Substitutions," Ann. Chim. Phys., 1857, 49,
487-496, p. 496).

At the same time he repudiated, definitely and finally,
Berzelius's electrochemical theories, as being neither " based
on evident facts," nor even valuable as a hypothesis in
" explaining and predicting facts " (Comptes rendus, 1839, 8,
621).

Liebig also was soon compelled to break away from the
complexities into which Berzelius's electrochemical theories
had led him and, in footnotes to two of Berzelius's papers,
added that :

" I do not share the views of Berzelius, since they rest
upon a mass of hypothetical assumptions, for the correct-
ness of which proof of every kind is lacking."

" I do not share the views, by which he explains the com-
position of the compounds discovered by Malaguti,1 I
believe on the contrary, that these materials are produced by
simple substitutions" (Ann. Chem. P/iqrm., 1839, 31, 119,
32, 72).

E. ORGANIC TYPES, NUCLEI AND RESIDUES

Dumas (1839) adopts the theory of types, — Dumas was
led to believe in the direct replacement of hydrogen by
chlorine mainly through his own experiments on TRICHLORO-
ACETIC ACID. He added small quantities of acetic acid to
a series of fifteen to twenty five-litre flasks filled with chlorine,
exposed them to sunlight, and on the next day found the
interiors frosted over with a crystalline acid, readily soluble
in water and deliquescent on exposure to air. The acid
1 Chlorinated ethers, e.g. C4H6C14O from C4H10O (p. 410).

412 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

was very caustic to the skin and its vapour was irritating and
suffocating even in small doses ; it was a strong acid, but
had none of the bleaching properties of chlorine. Its
chemical properties were almost identical with those of
acetic acid, from which it differed only in the replacement
of hydrogen by chlorine ; using modern formulae,

C2H402 + 3C12 = C2HC1302 + 3HC1.

Dumas prepared and analysed its silver, ammonium and
potassium salts, and its methyl and ethyl ethers, and showed
that these were similar in type to the corresponding deriv-
atives of acetic acid, and that the composition of each could
be predicted from the simple theory of substitution. A
similar fixity of type was observed in the conversion of
aldehyde 1 into chloral

C2H40 + 3C12 = C2HC130 + 3HC1,
in Malaguti's chlorination of ether

C4H100 + 4C12 = C4H6C140 + 4HC1,

and in Regnault's conversion of ethylene into chloroethylene
or vinyl chloride (Ann. de Ctiimie, 1855, 58, 301-320)

C2H4-HC12 = C2H3C1 + HC1.
Dumas sums up these observations as follows :

" Acetic acid, aldehyde, ether, olefiant gas, losing hydrogen
and taking an equal volume of chlorine, produce substances
belonging to the same type as themselves, chloracetic acid,
chloraldehyde, chlorether and chlorolefiant gas."

" In all these substances, chlorine, taking the place of
hydrogen, did not change the properties of the compound at
all, whether it was acid, base or neutral substance, for it
remained acid, neutral substance or base, and even retained
exactly its power of saturation ". . . .

" From the conversion of acetic acid into chloracetic acid,
from that of aldehyde into chloraldehyde, from the fact that

1 Prepared by Liebig as an oxidation- product of alcohol, C2H6O 4-
O = C2H4O + H2O, and described by him as aldehyde, abbreviated from
alcohol dehydrogenatns (Ann. Chirn. Phys., 1835, 59> 29°)«

xvii MOLECULAR ARCHITECTURE 413

all the hydrogen of these substances is replaced by chlorine
in equal volume, without changing their fundamental
character, we must conclude :

" That in organic chemistry there exist certain TYPES which
persist even ivhen in place of the hydrogen which they contain
one introduces equal volumes of chlorine, of bromine or of iodine"

"That is to say that the theory of substitutions rests upon
facts, and on the most striking facts of organic chemistry "
(Comptes rendus, 1839, 8, 620-622; also Ann. Chem.
Pharm., 1839, 32, 117-119).

Dumas (1840) on chemical and mechanical types, — In a

fuller memoir " On the Law of Substitutions and the Theory
of Types " {Comptes rendus, 1840, 10, 149-178 ; Ann. Chem.
Pharm., 1840, 33, 259-300), Dumas distinguished as belong-
ing to the same CHEMICAL TYPE " substances which contain
the same number of equivalents, united in the same way and
showing the same fundamental chemical properties" whilst,
following Regnault, he included under the heading of
MECHANICAL TYPES " substances having the same formula,
produced by substitution, but essentially different in their most
salient chemical properties" (Comptes rendus, 1840, 10, 158,
162). In the latter class he included a large number of cases
of oxidation, in which a neutral substance was converted
into an acid by substituting oxygen for hydrogen, etc. In
the same paper the system of naming substitution-products
by means of the prefix CHLORO-, etc., was formally set out
( Comptes rendus, 10, 169). In both memoirs ( Comptes rendus,
8, 621 ; 10, 169) Dumas referred to the analogy between sub-
stitution without change of type in organic chemistry and
isomorphism in inorganic chemistry. The two phenomena
were equally fatal to Berzelius's theories, since, as Liebig
pointed out (Ann. Chem. Pharm., 1839, 31, 119), the replace-
ment of the hydrogen of acetic acid by chlorine without
change of type was no more remarkable than the replacement
of the manganese of the permanganates by chlorine in the
isomorphous perchlorates. Liebig, however, regarded both
cases as exceptions, rather than as examples of a general

414 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

rule (ibid. 33, 301), and had an obvious motive in append-
ing to Dumas's paper a still more remarkable illustration of
the possibilities of substitution without change of type.

Wohler (1840) on substitution without change of type.—
The substance selected was manganous acetate, and the
action of chlorine upon it was described as follows :

" I passed a current of chlorine through a solution of
manganous acetate, under the direct influence of sunlight.
After 24 hours I found in the liquid a superb crystallisation
of a violet-yellow salt." This was manganous chloroacetate,
the whole of the hydrogen of the acetic acid having been
displaced by chlorine.

"This salt, heated at 110° in a current of chlorine, was
converted with liberation of oxygen gas into a new golden-
yellow compound." In this compound, the oxygen of the
base, as well as the hydrogen of the acid, had been dis-
placed by chlorine.

"The new substance was dissolved with the aid of heat in
pure chloral, and this liquid, which is not attacked by chlorine,
was used to continue the treatment by this agent. I passed
dry chlorine into it, during 4 days, keeping the liquid
always very near its boiling-point. During this time a white
substance was constantly deposited, which on careful exami-
nation was recognised as manganous chloride. I cooled the
liquid some time after, when there was no longer any more
precipitate and I obtained a third substance in small, green-
ish-yellow, silky needles." It contained no manganese, as
this in its turn had now been displaced by chlorine.

" On acting again with chlorine on an aqueous solution of
this substance, carbonic acid was set free and on cooling the
liquid to + 2° it deposited a yellow mass formed of small
plates, closely resembling chlorine hydrate. The carbon
had, in fact, been replaced by chlorine, which was now
the only element present in the compound. The mangan-
ous acetate had been converted by substitution into pure
chlorine, which, however, had a vapour-density which indi-
cated the presence in the molecule of not fewer than twenty-
four atoms of chlorine, MnO,C4H6O3, having been converted
into C12C12,C18C16C16."

xvii MOLECULAR ARCHITECTURE 415

Such was the graphic description of one of the most
surprising discoveries ever made in organic chemistry as set
out in a letter to Liebig (" On Substitution and the Theory
of Types," Liebig's Ann. Chem. Pharm., 1840, 88,308-310)
signed by S. C. H. Windier. The " s(ch)windler " who,
wrote this skit on Dumas's theories was Wohler. Liebig was
perhaps responsible for a footnote, appended to the state-
ment, that " in the decolorising action of chlorine there is a
replacement of hydrogen by chlorine, and that the fabrics
which are now bleached in England according to the laws of
substitutions preserve their types." The footnote runs :

" I have just learnt that there are already in the shops in
London fabrics of spun chlorine, much sought after in the
hospitals and preferred to all others for night-caps, calec.ons,
etc." (loc, cit. p. 310).

Laurent (1837) on nuclei. — Laurent, working in Dumas's
laboratory on lines suggested by Dumas, assimilated most
of Dumas's ideas ; but, with the enterprise of youth, he
developed them more quickly than Dumas, and was able
therefore to claim priority on certain important points.
These claims were, however, put forward in language which
did not err on the side of moderation. "I could scarcely
restrain my indignation on seeing certain chemists tax my
theory at first with being absurd, then later, when they
saw that the facts agreed better with my theory than with
all the others, pretend that I had taken possession of
the ideas of Dumas" (Ann. Chim. Phys., 1837, 66, 326).
If so, Dumas borrowed from Liebig, Liebig from the chemist
who first showed that oxide of potassium could exchange its
oxygen for an equivalent of chlorine, this latter from Richter
and Wenzel, and so to Hermes and Tubal Cain. His
dominant idea of simplicity in the combination of atoms
would probably be attributed to Berzelius, and consequently
to Moses ! (ibid. p. 330).

416 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

Dumas acknowledged that Laurent had been the first
to suggest that chlorine not merely took the place but played
the part of hydrogen (Ann. Chim. Phys., 1857,49, 496).
Laurent claimed further (Comptes rendus, 1840, 10, 412) that
he had anticipated in all but the name Dumas's discovery of
the permanence of types. He had recognised from the first
that the law of substitutions was often incorrect ; when it
happened to be true this was simply because of the strong
tendency of the radical attacked by the chlorine to conserve
its type, and not to any inherent accuracy of the law. These
stable groupings of atoms he had described in his Thesis
(1837) as " fundamental radicals " before substitution, and as
" derived radicals " after substitution (ibid. 412 and 416) ; in
a book published after mYdeath they are described as FUNDA-
MENTAL NUCLEI and as DERIVED NUCLEI (Chemical Methods,
tr. Odling, 1854, p. 195). To render his theory more
intelligible, he translated his idea into a geometrical figure :

" Let us imagine a four-sided prism, of which the eight
angles are occupied by eight atoms of carbon, and the
centres of the twelve edges by twelve atoms of hydrogen.
Let us call this prism \.\\Qform or fundamental nucleus, and
let us represent it by C8H12.

" Let us suppose that chlorine, put in presence of this
simple prism, removes one of the edges or hydrogen atoms ;
the prism deprived of this edge would be destroyed, unless
it were supplied with some other edge, whether of chlorine,
bromine, zinc, etc. ; no matter what the nature of the edge,-
provided it succeeds in maintaining the equilibrium of the
other edges and angles. Thus will be formed a new or
derived nucleus similar to the preceding, and of which the
form may be represented by C8(H11C1) " (Chemical Method,
1854, 195 : compare Comptes rendus, 1840, 10, 416, where
a similar but more complex figure is described in an extract
from his Thesis, 1837).

Addition-products could be derived from the fundamental
nucleus by adding pyramids to the ends of the prisms, to

xvn

MOLECULAR ARCHITECTURE

417

represent the added atoms of hydrogen, hydrogen and
chlorine, hydrogen and oxygen, etc.

Gerhardt (1839) on residues. — A third aspect of the
theory of substitutions, differing both from Dumas's empirical
conceptions and from Laurent's ideas of conservation of
type, was developed in 1839 by Charles Gerhardt
(1816-1856), a pupil of Liebig, who left Germany to
become Professor of Chemistry at Montpellier and after-
wards at Strassburg.

Berzelius had laid stress on the formation of compounds
by direct addition, e.g.

= BaSO4.

Dumas had been impressed by the action of chlorine in
displacing hydrogen from organic compounds, e.g.

C1CN + HC1.

Gerhardt directed attention to the formation of new com-
pounds by COPULATION, i.e. by the combination of two
substances with the simultaneous "production of a very
simple compound, such as water, hydrochloric acid, hydro-
bromic acid, etc." (Ann. Chim. Phys., 1839, 72, 196).
From a series of over twenty examples of copulation, the
following may be selected for representation by modern
equations :

Original
compound. Reagent.

2C6H6 +S03

Benzene.

C6H6 +HN03

Benzene.

inorganic
product.

= H20 -

residue of compoum
residue of reagent

h (C6H5)2S02

Sulphobenzene.

h (C6H5)N02

N itroben zene .

C2H6O + C2H4O2

Alcohol. Acetic acid.

C2H60 + C7H602

Alcohol. Benzoic acid.

= H20

= H20

h (C2HP)C2H302

Ethyl acetate.

f (C2H5)C7H502

Ethyl benzoate.

C7H5OC1 + NH3

Benzoyl
chloride.

= HC1

1- (C7H50)NH2

Benzamide.

or NH3 +C7H5OC1 = HC1 + (NH2)C7H5O

Benzoyl chloride.

Benzamide.

E E

4i8 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

It will be seen at once that this conception is much
broader than Dumas's Law of Substitutions, which was
restricted to those special cases of copulation in which the
reagent was an element, the original compound a hydride,
and the product a chloride, etc., derived from it by substi-
tution, e.g.

(C7H5C))H + C12 = HC1 + (C7H50)C1

Oil of bitter almonds. Benzoyl

Benzoyl hydride. chloride.

Gerhardt (loc. tit.} expressed the contrast as follows :
"If the reagent is an element in the free state, the
material removed is replaced, as Dumas has said, equivalent
by equivalent ; if, on the other hand, it is a compound, the
element removed is replaced by the residual elements of the
reagent" e.g.

SO2 from SOS minus O ; NH2 from NH3 minus H ;
NO2 from HNO3 minus OH ; C7H5O from C7H5OC1 minus Cl.

Gerhardt's conception of copulated compounds led him
very far in the direction in which a solution of the complex
problem of structural chemistry was ultimately found. It
showed him that the products of copulation (or CONDENSA-
TION, as it is now called) must be regarded as formed
from the mutilated RESIDUES of two molecules — residues
which might well be incapable of separate existence.
His residues (or radicals) thus differed essentially from the
radicals of Dumas and Boullay and of Wohler and Liebig,
which were always regarded as real substances, capable
of separate existence, as in the case of ammonia and
ammonium amalgam. Again, the older radicals were con-
sidered to have a real existence in their compounds ; but
Gerhardt was content to regard his residues or radicals as
expressing merely the possible methods of formation and
decomposition of a compound ; he was therefore prepared
to write barium sulphate indifferently as BaSO4 or BaO + SO3,
or BaO2 + SO2 or BaS + O4, according as he wished to
interpret one of its chemical changes or another (Treatise on
Chemistry, 1856, IV. 561).

XVII MOLECULAR ARCHITECTURE 419

Gerhardt (1843) halves the formulae of organic com-
pounds. — Further important developments were made in
Gerhardt's " Investigations on the Chemical Classification of
Organic Compounds " (fourn. prakt. Chem., 1842, 27, 439-
464; 1843, 28, 34-53, 65-100, 30, i-io). His theory of
copulation had compelled him to bring together into one
equation both complex organic products and simple in-
organic substances. He therefore soon discovered "that
the formula of most organic substances are too great by half
in comparison with the formula of inorganic chemistry"
(Journ. prakt. Chem., 1843, 30, 8). In inorganic chemistry.
(following Berzelius), water, ammonia and hydrogen chloride
were written as H2O, NH3, and HC1 ; but as products of
the copulation of organic compounds they always appeared
as H4O2, N2H6, and H2C12. Gerhardt was therefore led to
fall back again upon Avogadro's hypothesis as a valid guide
to the molecular formulae of organic compounds, and re-
wrote them as follows :

Acetic acid
Silver acetate .
Chloracetic acid
Silver chloracetate

New formula.

C2H402

C2(H3Ag)02
C2(HC13)02
C2(AgCl3)02

Old formulas.

CTJ r\ /"" rj /")
g/2gC/4= Cg"4(-/2'''

etc.

These formulae were only rendered possible by the fact that
Gerhardt had abandoned the older form of the theory of
radicals, according to which silver acetate was a binary
compound of silver oxide and "anhydrous acetic acid"
(acetic anhydride), whilst " hydrated acetic acid " (acetic
acid) was a binary compound of water with the anhydrous
acid, thus (in modern formulae)

Silver acetate = C4H6O3,Ag2O = Ag2C4H6O4.
Acetic acid = C4H6O3,H2O = C4H8O4.

To the student of this bewildering period of chemical
history nothing is more refreshing than to find once more
in Gerhardt's papers the simple formulae, based upon

E E 2

420 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

Avogadro's hypothesis, which had been abandoned even by
Dumas and were only established finally when Deville's
experiments on dissociation (Chapter XX) had cleared the
ground for Cannizzaro's epoch-making exposition.

Gerhardt (1856) distinguishes the hydrogen and chlorine
radicals from hydrogen and chlorine gas. — It is typical of
the difficulties that had been introduced into chemistry by
Berzelius's dualistic theories that Gerhardt appeared to be
breaking new ground when he wrote the formula of hydrogen
gas as H2 and the formula of chlorine as C12. Berzelius
would not recognise that hydrogen could combine with
hydrogen or chlorine with chlorine, since there could
be no electrical polarity to produce an attraction between
similar atoms. He therefore discarded Avogadro's hypo-
thesis as applied to elements, even when he was prepared
to apply it to compounds. Dumas, on the other hand,
fell into error when he attempted to apply Avogadro's
hypothesis to the elements on the assumption that all
elementary molecules were diatomic, whence Hg=ioo,
P = 62 instead of Hg = 200, P = 3i. Gerhardt, however, saw
that a distinction must be drawn between the hydrogen
radical and hydrogen gas, between the chlorine radical and
chlorine gas. Thus he writes :

" In opposition to most chemists, / regard the expression
radical in the sense of a relatiotiship, and not in that of a body
that may be or has been isolated, I distinguish, therefore,
the hydrogen radical from hydrogen gas, the chlorine radical
from free chlorine ; or better, if free hydrogen or chlorine
is to be represented by rational formulas, a study of their
reactions shows that hydrogen gas must be represented by
the two radicals HH, and chlorine gas by the two
radicals C1C1. In the usual nomenclature, hydrogen gas
would then be the hydride of hydrogen, and chlorine gas
would be the chloride of chlorine ; that is to say that
chlorine gas and hydrogen gas may be formed by, or may
give rise to, double decompositions precisely similar to those
which have caused oil of bitter almonds to be described as

xvn MOLECULAR ARCHITECTURE 421

benzoyl hydride, and the chlorinated oil as benzoyl chloride "
(Treatise, 1856, IV. 568).

The formation of hydrogen chloride from its elements was,
therefore, not a mere combination of an electro-positive
with an electro-negative element to form a compound of the
first order,

= HC1

but rather a double decomposition, similar to those which
had been studied by Dumas in developing the theory of
substitution

H2 +C12 = 2HC1
compare HCN + C12 = HC1 + C1CN

As Laurent expressed it in 1846 (Ann. Chim. Phys.,
1846, 18, 295) :—

" The molecule of hydrogen, of chlorine .... is formed
of two atoms which constitute a homogeneous compound,
(HH), (C1C1), (MM), etc. These homogeneous compounds,
placed in presence of one another, may give rise to a double
decomposition or to a substitution, and thus form a hetero-
geneous compound ;

(HH) + (C1C1) = (HC1) + (C1H)
(MM) + (C1C1) = (MC1) + (C1M)."

F. SIMPLE INORGANIC TYPES.

Inorganic types. — One of the most far-reaching conse-
quences of Gerhardt's theory of copulation was his recognition
of the essential equality of the two residues which are united
together in a copulated compound. This is shown by the
fact that he classified benzamide as a product of the action,
on one hand, of the reagent ammonia on the substance
benzoyl chloride, and on the other of the reagent benzoyl
chloride on the substance ammonia

C7H50-C1 + H-NH2 = C1H + CrH5O'NH2.

benzoyl chloride, ammonia. benzamide.

422 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

The way was thus opened up for a development of the
theory of types, which soon culminated in the modem
theory of valency.

Dumas's " types," like Laurent's " nuclei," were complex
organic structures, in which hydrogen could be displaced
by chlorine, or by a small group of atoms (such as the nitro-
group, NO2), without causing the structure to collapse. To
Gerhardt mainly belongs the credit of introducing a series
of simple INORGANIC TYPES, from which complex organic
compounds could be derived by substituting organic radicals
(ethyl, benzoyl, etc.) for the hydrogen atoms of the simple
type substance.

The ammonia-type (Hofmann, 1850).— Of these simple
types, the AMMONIA TYPE was the first to be recognised
clearly. Wurtz in 1849 (Ann. Chem. Pharm., 1849, 71,
331) discovered the simple bases, methy famine t CH3'NH2,
and ethylamine, C2H5'NH2. These two volatile organic
bases showed a remarkable resemblance to ammonia, from
which they could be derived by replacing a hydrogen atom
by methyl or ethyl. Hofmann, in the following year, showed
that all the hydrogen atoms of ammonia could be displaced
by alkyl radicals : —

" I have indeed found, that aniline and the bases which
are analogous to it, under the influence of methyl, ethyl, or
amyl bromide, lose one or two equivalents of hydrogen, which
is replaced by the corresponding alcohol-radical. Under the
same conditions ammonia loses one, two, or three equivalents
of hydrogen, which are likewise replaced by a corresponding
number of equivalents of the radical " (Preliminary note in
Ann. Chem. Pharm., 1850, 73, 91. Published in full Phil.
Trans., 1850, 93-131 ; 1851, 357-398).

Using modern formulae and writing

C2H6 =Et= Ethyl
C5Hn = Ay = Amyl
C6H5 =Ph = Phenyl

xvii MOLECULAR ARCHITECTURE 423

HI

H VN
PhJ

the substances tabulated by Hofmann are

1. Amide-bases (primary amines)

H
Aniline or phenylamine H [• N Ethylamine

Et

2. Imide-bases (secondary amines)

H ) H \ H

Et [N Ethylphenylanrine Ay VN Amyl- Et VN Diethylamine

PhJ PhJ phenylamine Et
(ethylaniline) (amylaniline)

3. Nitrile-bases (tertiary amines)

Et ) Et ) Et

Et [ N Diethylphenylamine Ay VN Ethylamyl- Et \ N Triethylamine

Ph) PhJ phenylamine Et
(diethylaniline) (ethylamylaniline)

The classification into PRIMARY, SECONDARY, and TERTIARY
AMINES was introduced in 1856 by Gerhard t, who suggested
that

" One might call the nitrogen-compounds primary ', second-
ary, and tertiary, according as they represent the ammonia-
type with substitution of one, of two, or of three atoms of
hydrogen" (Treatise, 1856, IV. 592).

Williamson (1852) on etherification, — As recently as
1852 it was still uncertain whether or not alcohol "contains
ether and water," i.e. whether the formation of ether from
alcohol should be represented by the equation

C4H100,H20 = C4H100 + H20,

or, as Dumas had represented it twenty-four years before, by
the equation

Williamson ("Theory of Etherification, " Journ. Chem. Soc.>
1852, 4, 106-112, 229-239) tested this point by acting on
alcohol with potassium, forming potassium ethoxide and
hydrogen,

C2H6O + K - C2H5OK + JH2

Potassium ethoxide.

424 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

and then with ethyl iodide, when ordinary ether (boiling at
37° C.) was formed

C2H5OK + IC2H5 = KI + (C2H5)2O.

Ether was thus proved to be a copulated compound con-
taining two ethyl groups, C2H5, and not merely the oxide
of a single radical, C4H10. This view was confirmed by acting
on potassium ethoxide with methyl iodide, CH3I, when

C* TT "i

ethyl methyl ether, p|j 5\O, was produced and with amyl

f TT ^
iodide, C5HUI, when ethyl amyl ether ^ ^2TT5 \Q, was pro-

^'5rill-'

fTrT ~i

duced ; methyl amyl ether ; ^ -rl [ O, was also prepared

by the action of amyl iodide, C5HnI, on potassium meth-
oxide, CH3OK.

Williamson concluded that :

" Alcohol is therefore water in which half the hydrogen is
replaced by carburetted hydrogen,1 and ether is water in
which both atoms of hydrogen are replaced by carburetted
hydrogen, thus :

HQ C2H5Q

HU H L

" From the perfect analogy of properties between the
known terms of the alcoholic series, it was to be expected
that similar substitutions might be expected in the others ;
and this expectation has been verified by experiment ....
Methylic alcohol is, therefore, expressed by the formula

r^T-T (~* TT

TT3O. as common alcohol is Vr5 O: and in the same
IT. xl

(~* TT

manner amylic alcohol is 5-rr11O, and the same of the

higher ones." (loc. tit. pp. 107 and 108).

Williamson (1852) on the water-type. — In a paper " On
the Constitution of Salts" (Journ. Chem. Soc., 1852, 4,
1 i.e. by the hydrocarbon-radical, C2H5.

xvn MOLECULAR ARCHITECTURE 425

350-355) published later in the same year, Williamson
suggested that :

" Formulae may be used ... as an actual image of what
we rationally suppose the arrangement of constituent
atoms in a compound, as an orrery is an image of what we
conclude to be the arrangement of our planetary system ;
and decompositions may be actually effected between them
by the exchange of a molecule of one group for a molecule
in another."

" The adoption of such a method will of course neces-
sitate the adoption of types, from which, by the replacement
of certain elements or molecules, we can deduce the con-
stitution of more and more complex groups. I believe that
throughout inorganic chemistry, and for the best known
organic compounds, one single type will be found sufficient ;
it is that of water, represented as containing 2 atoms of

hydrogen to i of oxygen, thus rrO. In many cases a multi-
ple of this formula must be used, and we shall presently see
how we get thereby an explanation of the difference between
monobasic and bibasic acids, etc." (loc. at., pp. 351-352).

Gerhardt (1856) suggests four inorganic types.— In the

fourth volume of his Treatise, published in 1856, Gerhardt
accepted Williamson's " water - type " as of dominant
importance in chemistry, but suggested that the number
of types should be increased to four, namely, WATER,

HYDROCHLORIC ACID, AMMONIA, and HYDROGEN. Unlike

Williamson, however, he preferred to regard these types as
expressing the double decompositions which a compound
might undergo, rather than the actual arrangement of the
atoms in the molecule.

Gerhardt expressed his views as follows : —

" My types are types of double decomposition. Water, in a
large number of double decompositions, may exchange its
oxygen and its hydrogen for other elements (simple radicals)
or for groups of elements (compound radicals). But I refer

426 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

substances to the type water, when one can effect similar
changes in them, and when the products are related in the
same way as those which result from the substitution of
other radicals for one of the radicals of water. For example,
I derive ether from the type water, because one can, by
double decomposition, replace the oxygen of ether by its
equivalent of chlorine, of bromine, of sulphur or of nitrogen,
to produce ethyl chloride, ethyl bromide, ethyl sulphide or
ethyl nitride (ethylamine) and because the products are
related in the same way as the hydrogen chloride, hydrogen
bromide, hydrogen sulphide and hydrogen nitride (ammonia),
which result from the substitution of the radicals chlorine,
bromine, sulphur and nitrogen, for the radical oxygen of
water" (Treatise, 1856, IV. 586).

" To facilitate the classification of substances according
to their functions, one may, instead of taking water alone as
a formula-type, join to it, as derived types, compounds
which result from the displacement of the radical oxygen
from water, such as hydrogen chloride, hydrogen nitride,
etc. . . . The study of organic compounds, as will be seen
later, proves that the four types water, hydrochloric acid,
ammonia, hydrogen, suffice for a methodical classification.
These four formula-types may be noted as follows :

Water OH2

Hydrochloric acid C1H
Ammonia NH3

Hydrogen HH

equal volumes.

" The water type includes the oxides (bases, acids, salts,
alcohols, etc.), the sulphides, the selenides and the
tellurides.

" The hydrochloric acid type includes the chlorides, the
bromides, the iodides and the cyanides.

" The ammonia type includes the nitrides and the
phosphides.

" The hydrogen type includes the metallic hydrides and
the metals " (ibid. pp. 588-589).

The following examples (with others) are given by
Gerhardt as illustrations of his theory of types : —

XVII

MOLECULAR ARCEIITECTURE

427

Type.
Water <

}{g

Benzoyl derivatives.
Benzoic o ( C7H5O
acid t H

Ethyl
alcohol

Ethyl derivatives.
O \ TT lather O \ /^ TT

Hydrogen
chloride

/ H

ici

Benzoyl / C7H5O
chloride \ Cl

Ethyl
chloride

{aHc

Ammonia '.

fH

V{H
IH

Benz- -i^- i TJ *
amide \ pj

Ethyl-
amine

v**

Triethyl- N
amine,

ICaHg

Diethyl-
amine

N{8S

Hydrogen

f H

IH

Benzoyl / C7H5O
hydride \H

(oilof bitter almonds)

Ethyl
hydride
(ethane]

{&H5

Diethyl
(butane)

1 C2H5
\C2H5

Williamson (1852 to 1855) and Odling (1855) on multiple
types. — Gerhardt's types indicated one stage, but only one
stage, in the construction of the molecule of a compound.

f TT "^

Ether, for instance, when written as £2£[5fO> was shown to

be composed of two ethyl-groups, C2H5, held together by
an oxygen atom ; but no indication was given as to how
the atoms of carbon and hydrogen were held together in
the ethyl radical. Until the position of every atom in the
scheme of combination could be shown, the structure of
/the compound was still only partly known.

The first stage in the subdivision of the compound into
.smaller units came through the introduction of "multiple
types " and of " mixed types," i.e. of structural formulae
;, showing not one bracket only, as in Gerhardt's simple types,
\but two or three brackets, linking together different portions
of the molecule.

MULTIPLE TYPES (as Kekule called them, Ann. Chem.
Pharm., 1857, 104, 133) arose primarily from the study of
dibasic and of tribasic acids (Williamson, Journ. Chem. Soc.t
1852, 4, 352; already quoted on p. 425). Nitric acid,
HNO3, and potassium nitrate, KNO3, could be derived very
easily from the simple water type, by writing them as

N&}°

428 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

But sulphuric acid, H2SO4, potassium hydrogen sulphate
(potassium bisulphate, KHSO4), and potassium sulphate,
K2SO4, could not be represented adequately in this way.
Williamson, therefore, who had begun by writing sulphuric
acid and ethyl hydrogen sulphate (in his two papers " On
Etherification," 1852) as :

H

H)S04

and

was soon led to regard all the sulphates as derived from
two molecules of water and to write their formulae (in his
paper " On the Constitution of Salts," Journ. Chem. Soc.,
1852, 4. 353) as :

The significance of these formulae is shown clearly in a reply
to certain criticisms of Kolbe (Journ. Chem. Soc.^ 1855, 7,
in), in which Williamson derives the formula of sulphuric
acid from that of water by allowing a molecule of sulphur
dioxide to " replace two atoms of hydrogen in two of water,"
thus

'°

(Journ. Chem. Soc., 1855, 7, 137). In the last formula, the
molecule of sulphuric acid is subdivided into five portions,
namely, four single atoms and one small radical, SO2, the
exact structure of which is still doubtful even at the present
day.

Williamson's multiple types were developed by Odling
(" On the Constitution of Acids and Salts," 'Journ. Chem. Soc.,

1 In the original papers Williamson writes all his formulae without
brackets, whilst Odling writes 2O instead of O2 in the formulae of
multiple types ; for the sake of convenience, one method of representa-
tion only is used in this paragraph.

XVII MOLECULAR ARCHITECTURE 429

l855> 7, 1-22), who showed that complex products might
result from the displacement of several atoms of hydrogen
by an acid or basic radical. Thus three molecules of water
are required to produce phosphoric acid and bismuth
hydroxide,

Ho) ~
Bi}°3

or the salts, such as potassium hydrogen phosphate, and
bismuth nitrate, which are derived from them,

H,

Kekule (1857) on mixed types, — Very similar to the
multiple types of Williamson and Odling are the MIXED
TYPES of Odling and Kekule.

Odling (like Williamson) had derived sodium sulphate,
Na2SO4, from two molecules of water, but found it
necessary to derive sodium hyposulphite, Na2S2O3 (ordinary
" hypo," now called sodium thiosulphate), from a molecule
of water and a molecule of sulphuretted hydrogen (Journ.
Chem. Soc., 1855, 7, 8), thus

gives 2o2 (sodium sulphate)

but 2O + S gives 2O + S (sodium thiosulphate).

If written out in full (compare Williamson, p. 428) these
formulae would appear, thus :

«}o Na}o > N>

gives S02 gives SO>

H}° Na}° H}S Na)S

These MIXED TYPES differ from the MULTIPLE TYPES only
in that the types which are linked together are different
instead of being identical. Kekule, who gave them their
name (Ann. Chem. Pharm., 1857, 104, 133), made use very

430 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

largely of mixed types to represent the structure of com-
pounds, e.g. (loc. cit. 135, 136, 137)

H| H \

gave SO2 { (sulphurous acid, H2SO3)
H

cn ci \

H/ cj-v / (chlorosulphonic acid, C1SO3H, prepared

H

Qfx j \Lftiur umtpnunit uciu, v^iov^3iA, jji

2 \ r> by combining SO3 + HC1)

H ;°

Hl

H >N jj VN (carbamic add, the parent substance

H j gave of ammonium carbamate,

CO

H 1
jj VN

CO \n
H /U

Gerhardt (1855) on conjugated radicals.— A parallel
method of subdividing the molecules of organic compounds
was devised by Gerhardt, who found that the process of
" copulation " could be applied, not only to complete
molecules, but also to the radicals of which those molecules
are composed. In his Treatise (1852, IV. 604-610) he
makes use of CONJUGATED RADICALS to represent simul-
taneously " two or several systems of double decomposition
of the same body." This method of representation was
applied with great success to the fatty acids. Most of
their double decompositions (formation of salts, chlorides,
amides, esters, etc.) could be expressed by formulae of the
water-type, thus :

Formic acid, o|^HO Propionic acid,

Acetic acid, O-0 Butyric acid,

But "the acid radicals of the formula CnHgn-TO may be
viewed as composed of the carbonyl radical CO and an
alcohol radical CnH2n+1 :

CO(H) formyl CO(C2H5) propionyl

CO(CH3) acetyl CO(C3H7) butyryl."

xvn MOLECULAR ARCHITECTURE 431

" These formulae are justified by the reactions. We know,
for example, that acetyl-com pounds are resolved, in many
cases, into carbonic compounds and methyl compounds :
acetic acid can be transformed by heat into carbonic acid and
hydride of methyl (marsh gas) ; potassium acetate gives, by
the action of an electric battery, methyl * and carbonate of
potash etc." (Treatise, 1852, IV. 606).

Thus in order to express the decompositions represented by
the equations

C2H4O2 = CH4 + CO2 (decomposition by heat)

2C2H4O2 = (CH3)2 + 2CO2 + H2 (decomposition by electrolysis).

it was necessary to write acetic acid, not merely as
; 0{£H«0, but also as O {COW ,

where the ACETYL group, C2H3O, is shown as a conjugated
radical, composed of CARBONYL, CO and METHYL, CH3.

Other decompositions suggested that the acetyl group
should be represented as C2H3(O), i.e. as a compound of
a hydrocarbon radical C2H3 with oxygen. In the same way,
" even the alcohol-radicals may be considered as conjugated
radicals, of which the constituent radicals are the aldehyde
radical2 and hydrogen." Thus whilst acetyl= C2H3(O),

The theory of conjugated radicals was also • applied
effectively to represent nitration-products, obtained by the
action of nitric acid on organic compounds such as benzoic
acid (ibid. p. 664) ; thus

Benzoic acid = O /C;H5O

oft

Nitrobenzoic acid = {C,H4(NO2)O.

1 The decomposition of the potassium salt is shown by the equation
2C2H302K + 2H20 = (CH3)2 + 2KHC03 + H2. The product, which
Gerhardt called methyl, CH3, is really dimethyl or ethane, (CH3)2 or
C2H6.

2 i.e. C2H3, aldehyde being (C2H3)HO. This radical (acetyl
minus oxygen) was described by Liebig (see p. 405) as " acetyl."

432 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

The round brackets which Gerhardt used in formulating
his copulated compounds have been retained to the present
day as an integral part of the symbolism of structural
chemistry. The curly brackets of the theory of types, on the
other hand, have long since disappeared.

G. VALENCY.

Unequal combining-power of the elements indicated in
Gerhardt 's simple inorganic types, — The theory of types
as developed by Gerhardt carried with it the idea that :

(i) In compounds of the hydrogen type, a hydrogen-
atom could be linked to one radical, but only to one radical,
giving compounds such as

CH31

as

Methyl hydride
(marsh gas or
methane).

C,H,1
HJ

Ethyl hydride
(ethane).

C2HS0)

H;

Acetyl hydride
(aldehyde).

C7H50|
H/

Benzoyl hydride
(oil of bitter almonds
or benzaldehyde).

(2) In compounds of the hydrochloric acid type, a
chlorine-atom could be linked to one radical, but only to one
radical, giving compounds such as

C2

H5| C2H3Ch

C1J C1J

C1J C1J C1J ClJ

Methyl chloride. Ethyl chloride. Acetyl chloride. Benzoyl chloride.

(3). In compounds of the water type, however, two
radicals could be linked simultaneously to an oxygen-atom
as in

CH810
CH3/U

Methyl oxide
(Methyl ether).

clu)°

Ethyl oxide
(Ethyl ether).

C2H3OJ0
C2H3o/°

Acetyl oxide l
(Acetic anhydride).

cXo/°

Benzoyl oxide1
(Benzoic anhydride).

1 As a consequence of Berzelius's dualistic conceptions of molecular
structure these acid-anhydrides had been regarded for many years as
the real acids, of which ordinary acetic acid, benzoic acid, etc., were
hydrates. They were, however, purely hypothetical compounds, until
Gerhardt in 1853 (Ann. Chem. Pharm., 1853, 87, 57-84, 149-179)

xvii MOLECULAR ARCHITECTURE 433

(4) In compounds of the ammonia type, three radicals
could be linked together, as in

CH3|
CH°j

Trimethylamine. Trimethylphosphine. Dimethylaniline

Hofmann had also represented the compounds of the
amines and phosphines with HI and with CH3I by formulae
which showed five radicals associated with an atom of
nitrogen or of phosphorus (Phil. Trims., 1850, 357-398).

H
H

Tetramethylammonium iodide Phosphonium iodide

from N(CH3)3+CH3I. from PH3+HI.

Kekule (1857) on substitution-value or atomicity, —

These considerations, arising from the study of Gerhardt's
simple types, were expressed very clearly by Kekule, who
proposed in 1857 to describe the elements according to
their " substitution-value " or combining power as MON-
ATOMIC, DIATOMIC, and TRIATOMIC. In the first of his two
papers, "On Copulated Compounds and the Theories of
Polyatomic Radicals," he writes :

" The molecules of chemical compounds consist of aggre-
gations of atoms."

" The number of the atoms of other elements (or radicals)
combined with one atom (of an element. ... or of a radical)
is dependent on the basicity or substitution-value of the
constituents."

"From this point of view the elements fall into three
principal groups :

prepared them by the same method that Williamson had used for the
synthesis of the ethers, e.g.

C7H5O'C1 + C7H5OOK = KCl + (C7H5O)2O
C2H30-C1 + C2H3OOK = KCl + (C2H3O)2O.

F F

434 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

"(i) Monobasic or monatomic, I, e.g. H, Cl, Br, K;
" (2) Dibasic or diatomic, II, e.g. O, S ;
"(3) Tribasic or triatomic, III, e.g. N, P, As."
(Ann. Chem. Pharm. 1857, 104, 132-133).

A study of the multiple types and mixed types showed that
similar considerations could be applied to the radicals,
thus:

" A monatomic radical can never hold together two mole-
cules of the types.

" A diatomic radical can unite two molecules of the types
e.g.

SO"2,C12 SO/ H

HJO CO")N2

Sulphuryl chloride. Sulphuric acid. Carbamide (urea).

" A triatomic radical similarly unites three molecules of
the types, e.g.

, C3H6'",C13.

Phosphoric acid. Glycerine. Trichlorhydrin.

This "substitution-value" or combining power of ele-
ments and radicals is now described as VALENCY, elements
and radicals being described as UNIVALENT, BIVALENT, TER-
VALENT, and QUADRIVALENT (Lothar Meyer, Moderne
Theorien, 1864, 76), according as they combine with one,
two, three or four other univalent elements or radicals.

Williamson (1851), Frankland (1852), and Odling
(1855) on combining power or valency. — The concep-
tions thus clearly set out by Kekule were by no means
wholly new. Williamson in 1851 had represented the
conversion of "hydrate of potash" (potassium hydroxide)
into potassium carbonate by a scheme

xvn MOLECULAR ARCHITECTURE 435

which represented carbonic oxide as displacing two atoms
of hydrogen in a double-water-type ; in doing this he sug-
gested that

"One atom1 of carbonic oxide is here equivalent to two
atoms of hydrogen, and by replacing them, holds together
the two atoms l of hydrate in which they are contained, thus

necessarily forming a bibasic compound, ' x 2> cai>bonate

of potash" (Journ. Chtm. Soc., 1852, 4, 353; reprinted
from the Chemical Gazette, 1851).

Carbonic oxide was therefore, in Kekule's phraseology, a
dibasic or diatomic radical.

Frankland, in the following year, in the memorable
paper in which he described a " New Series of Organic
Bodies containing Metals" (Phil. Trans., 1852, 417-444),
had put forward the conception of a definite COMBINING
POWER for each element :

" When the formulae of inorganic chemical compounds
are considered, even a superficial observer is struck with
the general symmetry of their construction ; the compounds
of nitrogen, phosphorus, antimony and arsenic especially
exhibit the tendency of these elements to form compounds
containing 3 or 5 equivalents, of other elements, and it is
in these proportions that their affinities are best satisfied ;
thus in the ternal group we have NO3,2 NH8, NI8, NSa,2
P02,2 PH3, PCI,, Sb03,2 SbHw SbCl3J As03,2 AsH3, AsCla,
etc. ; and in the five-atom group NO5,2 NH4O,2 NH4I,
PO5,2 PH4I, etc. Without offering any hypothesis regarding
the cause of this symmetrical grouping of atoms, it is
sufficiently evident, from the examples just given, that such
a tendency or law prevails, and that, no matter what the
character of the uniting atoms may be, the combining
power of the attracting element, if I may be allowed the
term, is always satisfied by the same number of these atoms
(foe. dt. p. 440).

1 i.e. molecule, molecules.

2 /.*. N203, N2S?, P20S, Sb,03. As203, N2O5, (NH4)3O, P,O5, etc.,
using modern atomic weights in place of Gmelin's equivalents.

F F 2

436 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

Odling in 1855 (Journ. Chem. Soc., 1855, 7, 3) had pre-
sented the same idea in a graphical form :

" For the clear elucidation of the succeeding formulae, I
adopt in them, a simple plan of marking these different
substitution values, viz., by one or more dashes to the right
or left of the symbol, something after the fashion frequently
made use of in algebraical formulae : thus,

H', an atom of hydrogen, ....
Bi"', an atom of bismuth, .......

having a value represented by three atoms of hydrogen,
etc."

Water was thus represented as TJ/ rO", and alum as

K'CT

K'AL

if the atomic weight of aluminium be doubled, to conform
with modern usage.

Frankland and Odling were, however, greatly hampered
by their adherence to an old system of equivalents, from
which even Kekule himself did not wholly escape until
1867. The result was that Odling wrote tin and iron as
Sn' (stannous) Sn" (stannic) instead of Sn" and Sn""
Fe' (ferrous) Fe2"' (ferric) instead of Fe" and Fe'"
He thus regarded two atoms of aluminium and two atoms
of ferric iron as tervalent, as indicated by the symbols A12"',
Fe2'". In the same way, Kolbe and Frankland, in a paper
in which they foreshadowed the quadrivalency of carbon
(Ann. Chem. Pharm., 1857, 101, 257-265) were only able to
discuss the combining-power of a double atom C2 of atomic
weight 2x6 (see Japp's " Kekule Memorial Lecture, "Trans.
Chem. Soc.) 1898, 73, 97-138; p. 130). Their conceptions
of valency or combining-power could not therefore be put
forward in the same convincing way as in Kekule's papers,

xvn MOLECULAR ARCHITECTURE 437

where Gerhardt's newer atomic weights were employed. In
particular, it must be noticed that if A12 is tervalent, Al
must be at least quadrivalent, to provide for the linking
together of the two aluminium atoms ; so also, if C2 is
quadrivalent, the single carbon atoms must be at least
quinquevalent, etc.

Kekule on the " marsh gas type " (1857) and the
quadrivalency of carbon (1858). — The problem of molecular
structure was almost solved when Gerhardt's types (simple,
multiple and mixed) had developed into the conception of
valency. It was then seen clearly that oxygen and sulphur
might link together two radicals, whilst nitrogen and
phosphorus might unite three radicals. But the inner
structure of the radicals themselves remained obscure, until
Kekule detected the secret of their construction in the
quadrivalency of carbon ; the radicals, in fact, were held
together by their carbon atoms, which possessed the
remarkable power of uniting four other atoms or radicals.

In 1857 he had referred some ten different compounds of
carbon to the MARSH GAS TYPE (Ann. Chem. Pharm^ 1857,
101, 204), thus deriving them from a hydrocarbon containing
four atoms of hydrogen1 displaceable by other radicals.
In the following year, having accepted the atomic weights of
Gerhardt and of Cannizzaro, he was in a position to discuss
the combining-power of the single carbon atom in a marsh
gas of the formula, CH4, as follows :

" If one considers the simplest compounds of carbon
(marsh gas, methyl chloride, carbon chloride, chloroform,
carbonic acid, phosgene gas, carbon sulphide, prussic acid,
etc.) it appears that the quantity of carbon, which chemists
have recognised as the smallest possible, as the atom, always
binds four atoms of a monatomic, or two atoms of a diatomic
element ; that in general the sum of the chemical units of the
elements combined with an atom is equal to four. This

1 Kekule still wrote the formula of marsh gas as C2H4, taking the
atomic weight of carbon as 6.

438 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

leads to the view, that carbon is tetratomic (or tetrabasic) "
(Ann. Chem. Pharm., 1858, 106, 153).

" If one introduces carbon as a tetratomic radical into the
types, one obtains for several of its known compounds
relatively simple formulae " (ibid, footnote).

The QUADRIVALENCY of the single carbon atom was shown
in the following compounds in which it is combined with

(a) Four
tinivalent
radicals.
CH4
CC14
CHSC1
CHCls

(b) One bivalent
and two
univalent radicals.
COClo

(c) Two
bivalent
radicals.
C02
CS2

(d) One tervalent
and one univalent
radical.
CNH

When two carbon atoms were united, the valency of the
double carbon atom, C2, was not 2x4 = 8, but 6, since two
units of combining-power were used in holding the two
carbons together, as in :

Ethane . . C2H6 Acetonitrile . C2H3N'"

Ethyl chloride . Q>H5C1 Cyanogen . C2N2'"

Ethylene chloride C2H4C12 Aldehyde . C2H4O"

Carbon trichloride C2C16 Acetyl chloride C2H3C1O", etc.

(ibid., p. 154). In reference to such a group of compounds
Kekule suggests that

" If one compares together compounds, which contain the
same number of carbon atoms in the molecule and can be
converted into one another by simple metamorphoses
(e.g. alcohol, ethyl chloride, aldehyde, acetic acid, glycollic
acid, oxalic acid, etc.) one comes to the conclusion, that
they contain the carbon atoms arranged in the same way,
and that only the atoms attached to the carbon-skeleton
change (ibid., 155-156).

Structural formulae of organic compounds. Graphic
formulae. — Kekule's recognition of the quadrivalency of
carbon, and of the CARBON-SKELETON as the nucleus of all
organic compounds, provided a complete solution of the
^problem of molecular structure. In order to show the
structure of any compound it was only necessary to indicate

MOLECULAR ARCHITECTURE

439

the arrangement of the multivalent atoms forming the
skeleton, and the method of attachment of the other atoms
to this skeleton.

We also owe to Kekule a revival of the GRAPHIC
FORMULA, which Dalton had used to indicate the number
and the nature of the atoms (p. 295), but which Kekule
proposed to use in order to show also the way in which the
atoms were linked together. The univalent atoms were
represented by circles, the bivalent atoms by a dumb-bell
figure occupying the width of two circles, and so on. In
all the simpler compounds, the atoms were arranged in two
equal horizontal rows, and every point of contact between
the two rows indicated a linkage between two atoms. A
number of Kekule's graphic formulae (Lehrbuch, 1861, I.
1 60, 162, 164, footnotes) are shown below, together with

Simple Inorganic
Compounds.

Hydrogen chloride

Kekule's
graphic
formulae.

Modern Modern

structural graphic

formulae. formulae.

HC1

H-C1

HOH

H-O-H

NH3

H-<"

02

O=O

S02

-o-s-o-

or V7S<
C1-SO2-C1 C1-O-S-O-C1

or ^S^
HO-SO2-OH H-0-O-S-O-OH

ON-O-OH O = N-O-OH
or NO2-OH or °*N-OH

1 In these cases the structure given by Kekule differs from that wtdch
is commonly accepted to-day and which is shown as the second of two

440 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

Kekulfs
graphic
formulae.

Derivatives of
Marsh Gas,

Marsh gas
Methyl chloride .
Carbonyl chloride
Carbonic anhydride

Prussic acid .

Derivatives of
Ethane.

Ethyl chloride
Ethyl alcohol
Acetic acid .
Acetamide

Modern
structural
formulas.

Modern
graphic
formulae.

CH4

HTT
^|~»xJCT

HX NTT
n

CH3-C1

HVH

C1-CO-C1

0=cccl

C02

o=c=o

H-CN

H-CHN

H H

CH3-CH2-C1 H-C-C-C1

H A

TT TT

CH3-CH2-OH K-C-C-OH

CH3-CO-OH

«-ll

OH

CH3-CO-NH2 H-C-C-NH,

modern structural formulae, in which dots are used to show
the linkages between the radicals, and modern graphic
formulae, in which the linkages are shown in detail by a
series of straight lines.

Kekule's formula for benzene (1865). — The subsequent
history of organic chemistry has consisted largely of a
search for, and demonstration of structural formulae for
compounds of greater and greater complexity. At first
these compounds showed only " open-chains " of carbon
atoms, but Kekule himself in 1865 (Bull. Soc. C/iim., 1865,

alternative formula. It will be seen that Kekule regarded sulphur as
bivalent in all its compound and nitrogen as tervalent ; modern formulae
often show sulphur as quadrivalent or sexavalent and nitrogen as
quinquevalent.

xvii MOLECULAR ARCHITECTURE 441

3, 98-110; Ann. Chem. Pharm., 1866, 137, 129) recognised
the existence of a " closed chain " or " ring " of carbon
atoms in the formula of BENZENE ; similar ring-systems are
characteristic of nearly all aromatic compounds. Kekule's
graphic formula for benzene was of the same type as those
shown above, but may be translated into modern symbols

/~»TT /"*T1J\

as HO^-TT r-TT^CH. Only three valencies are needed

to form the ring of carbon-atoms and to hold the six hydro-

fTT ("'TT

gen atoms, as shown in the symbol HC/O^rr ^rr^CH ;

much controversy has therefore arisen as to the way in
which the fourth valency is disposed of, but Kekule's
formula is universally recognised as the opening of a new
chapter in organic chemistry and was made the subject of a
public celebration in 1890.

Position of the atoms in space. — Kekule represented
the four hydrogen atoms of marsh-gas as lying in a row
by the side of the carbon atom. More frequently they
have been represented as arranged round the carbon atom
in such a way as to occupy the four corners of a square.
Van t'HofFs study of the isomerism of carbon compounds,
and the crystallographic work of Barlow and Pope, and of
Bragg, have shown that they are probably arranged at the
corners of an enveloping tetrahedron. This idea, one of
the most fertile in modern chemistry, cannot, however, be
discussed in the present volume.

Fixed or variable valency. — Is the valency of an atom
fixed or variable ? Kekule held strongly to the view that
the valency or " atomicity is a fundamental property of the
atom, which must be constant and invariable like the weight
of the atom itself" (Comptes rendus, 1864, 58, 511). In
order to account for the existence of compounds such as
ammonium chloride and phosphorus pentachloride in which
the valency of nitrogen and phosphorus appears to be in-
creased, he assumed the existence of a class of MOLECULAR

442 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

COMPOUNDS, which " do not form vapours, but are decom-
posed by the action of heat, reproducing the molecules
which gave birth to them " (ibid. 513). Thus he wrote

NH3 + HC1 = NH3,HC1 not NH4C1
PC13 + C12 =PC13,C12 ^PC15

Frankland, on the other hand, in his first note on the
combining-power of the elements (1852) had already
adopted the idea of variable valency by ascribing to
nitrogen and phosphorus a tendency to unite with three or
five equivalents of other elements. This idea, in one form
or another, is now adopted very widely. Odling's attempt
to distinguish between " artiads " of even valency and " peris-
sads " of uneven valency (Phil. Mag., 1864, 27, 115-119)
broke down when it was found to lead to purely fictitious
formulae, such as N2O.2 for nitric oxide, NO, and to unjusti-
fied contrasts such as those shown by the formulae
FeCl2,Fe2Clfi (for FeCl3) and KQO4,K2Mn;JO8 (for KMnO4).

SUMMARY AND SUPPLEMENT.
A. THE RISE OF ORGANIC CHEMISTRY.

The rapid growth of inorganic chemistry between 1766 and
1816 is paralleled by the rise and development of organic
chemistry in the fifty years from 1815 (when Berzelius first
assigned formulas to organic compounds) to 1865 (when Kekule
put forward his structural formula for benzene). The recogni-
tion of organic substances as compounds of carbon and hydrogen
with oxygen, and sometimes nitrogen and phosphorus, is due to
Lavoisier (1784-1789), who also made the first organic analyses.
He estimated the carbon and hydrogen in charcoal, olive-oil,
wax and spirit of wine (1784) and showed (1789) that in vinous
fermentation there is a balance between the carbon, hydrogen,
oxygen, and nitrogen in the materials used (sugar, water, yeast)
and in the products obtained (alcohol, carbonic anhydride, etc.).
In doing this he enunciated the principle of the CONSERVATION
OF MASS in chemical changes and explained the significance of

xvn MOLECULAR ARCHITECTURE 443

a chemical equation. Other early analyses were made by Gay-
Lussac and Thenard (1810), who discovered the existence of
carbohydrates, by Berthollet (1809), by de Saussure (1807 and
1814), who analysed alcohol and ether, by Berzelius (1815),
who first expressed his organic analyses by means of formulas,
and finally by Liebig (1831), to whom belongs the credit of
perfecting the methods of organic analysis.

B. THE STRUCTURE OF SALTS.

Mayow, in 1674, showed that salts contained an acid and
a base, which might be set free again by the action of a stronger
acid or base. Lavoisier regarded salts as compounds of two
oxides, namely, the acid oxide of a non-metal and the basic oxide
of a metal. Berzelius (1819) adopted this dualistic theory, but
recognised that binary compounds could be formed which
contained no oxygen, and (at a later date) that salts might be
formed by uniting two hydrides or two sulphides. He suggested
that all compounds were held together by the electrical attrac-
tion between oppositely charged components ( Berzelius' dualistic
or electro-chemical theory].

C. THE THEORY OF RADICALS.

Berzelius's electro-chemical theory could not readily be applied
to organic compounds, which were not amenable to electrolysis.
In order to bring them into his scheme Berzelius (1819) adopted
Lavoisier's view that inorganic compounds were derived from
simple elementary radicals, but organic compounds from
compound radicals, containing two or more different elements.
The most important radicals were :

Cyanogen . . . CN or Cy Gay-Lussac, 1815.

[Ammonium . . NH4 or Am Ampere, 1816.

-j Ammonia . . NH3

(Amide . . . NH2 or Ad Berzelius, 1832

(Ethyl . . . C2II5orEt Liebig, 1834^

•! Aetherin (ethylene) C2H4 or Ae Dumas and Boullay, 1827.

( " Acetyl " now C2H3O) C2H3 Liebig, 1839.

/Methylene . . CH2 Dumas and Peligot, 1834.

\Methyl . . . CH3 or Me Dumas and Peligot, 1834.

/•Benzoyl . . . C7H5O or Bz Wohler and Liebig, 1832.

\Pikramyl . . C7H6 or Pk Berzelius, 1843.

444 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

After the discovery of the benzoyl-radical in 1832, Liebig
became the champion of the theory of radicals, whilst Berzelius
was chiefly concerned that the dissection of organic compounds
should be carried out in such a way as to segregate completely
the electro-positive and the electro-negative elements. Dumas,
who had much to do with the development of the theory of
radicals, soon broke away from Berzelius's interpretation of that
theory and became the chief exponent of the rival theory of
"substitution."

D. THE THEORY OF SUBSTITUTION.

Dumas, in 1834, showed that by a process of substitution or
metalepsy the hydrogen of organic compounds could be replaced
by an equivalent quantity of chlorine or oxygen, e.g.

Prussicacid. . HCN+ C12= HC1 + C1CN (Gay- Lussac, 1815)

Ethylene chloride C2H4Cl2 + 4Cl2 = 4HCl + CoCl6 (Faraday, 1821)

Oil of bitter C7H6O + C12= HC1 +C\II5OC1

almonds (Wohler and Liebig, 1832).

In some cases, hydrogen was removed without replacement, e.g.
Oxalic acid . H2C2O4 + O gave H2O +2CO2.

In such cases, it was assumed that hydrogen was present in the
form of water : thus in the preparation of chloral by the action
of chlorine on alcohol,

Ethylene. . C2H4 +3C12 gave C2HC13 +3HC1

Water . . H2O + C12 gave O +2HC1

therefore

Alcohol . C2H6O + 4C12 gave C2IIC13

E. ORGANIC TYPES, NUCLEI AND RESIDUES.

Laurent, a pupil of Dumas, asserted that substitution was
not merely an empirical rule, but, in defiance of all Berzelius''
electro- chemical conceptions, that chlorine took the place
and played the part of hydrogen in organic compounds.
Dumas, in 1839, adopted this view by suggesting that substi-
tution could take place without any change of type ; this idea
was cleverly ridiculed in an anonymous letter written by Wohler
in 1840. Dumas distinguished between chemical types, in which

xvii MOLECULAR ARCHITECTURE 445

the laws of substitution were obeyed strictly, and mechanical
types, in which a part of the hydrogen (existing separately from
the rest in the form of water) could be removed without replace-
ment by oxygen or chlorine. Laurent, in 1837, found an
explanation of the same facts in the theory that an organic
compound consists of a nucleus of atoms geometrically arranged,
e.g. in the form of a prism ; hydrogen atoms in the prism
could only be replaced according to the laws of substitution,
but other atoms, forming pyramids on the ends of the prism,
could be added, or removed without replacement. Gerhardt,
in 1839, regarded substitution as a particular case of the
copulation of two residues or radicals. Thus when chlorine
acted on oil of bitter almonds, the former lost a chlorine and
the latter a hydrogen atom in the form of hydrogen chloride and
the product was formed by the union of the two residues,

C7H6O!H +"C1JC1 = H£l + C7H5O 'Cl.

In this way the idea of substitution could be extended to include
the products obtained by the action of sulphuric acid and nitric
acid on benzene, of acetic acid on alcohol, of ammonia on
benzoyl chloride, etc. Gerhardt regarded his residues, not as
real substances, but as expressions of the changes which a
compound could undergo ; he therefore distinguished between
the radicals H and Cl and the gases H2 and C12. He also halved
the formulae of nearly all organic compounds and used simple
formulas based on Avogadro's hypothesis.

F. SIMPLE INORGANIC TYPES.

The discovery by Wurtz, in 1849, of the primary amines,
methylamine, CH3'NH2, and ethylamine, C2H6'NH2, and by
Hofmann, in 1850, of the secondary and tertiary amines, led
to the recognition of ammonia as a simple inorganic type^
from which derivatives could be obtained by replacing one or
more of the three hydrogen atoms by organic radicals. The
experiments of Williamson, in 1852, on etherification, showed
that alcohol and ether could be derived from the water type by
replacing one or both of the two hydrogen atoms by ethyl.
Gerhardt, in 1856, added the hydrogen type and the hydrochloric
acid type. The marsh gas type, CH4, in which four hydrogen

446 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

atoms can be replaced by radicals, was introduced by Kekule
in 1857.

These simple inorganic types showed only one stage in the
construction of the molecule, namely, the linking together of
radicals either directly (in pairs), or through an atom of oxygen,
nitrogen or carbon, whereby two, three or four radicals could
be held together. A further subdivision of the molecule was
effected in the multiple types of Williamson (1852-1855),
and Odling (1855), the mixed types of Odling (1855) and
Kekule (1857), and finally the conjugated radicals of Gerhardt

G. VALENCY.

Gerhardt's simple inorganic types indicated that the different
elements possessed unequal combining-power ; thus hydrogen and
chlorine could only be linked to one radical, oxygen to two,
and nitrogen to three. Kekule therefore suggested, in 1857,
that the elements should be described, according to their
substitution-values, as monatomtc, diatomic, triatomic. This
idea could also be supplied to radicals, the sulphuryl and
carbonyl radicals being diatomic, as in sulphuryl chloride,
SO2C12, sulphuric acid, SO2(OH)2, and carbamide, CO(NH2)2,
whilst triatomic radicals were present in phosphoric acid,
PO(OH)3, glycerine, C3H5{OH)3, and trichlorhydrin, C3H5C13.
The diatomic character of the carbonyl radical in potassium
carbonate, CO(OK)2, had already been indicated by Williamson
in 1851. Frankland, in 1852, had directed attention to the fact
that nitrogen, phosphorus, arsenic and antimony always com-
bined with three or five equivalents of other elements, and had
suggested that each element had a definite combining-power,
Odling, in 1855, had gone further and had represented the
combining powers or substitution values of the elements by
symbols such as H', Cl', O", S", N'", P'". Frankland and
Odling thus set forth clearly the idea now known as valency,
but they were hampered by an obsolete system of atomic
weights, and in the case of many elements could only indicate
that the double-atom A12 was tervalent, C2 was quadrivalent,
etc. The quadrivalency of the single carbon atom was first
postulated by Kekule in 1858 ; he had already, in 1857, intro-
duced the marsh gas type, containing four hydrogen atoms,

xvn MOLECULAR ARCHITECTURE 447

displaceable by radicals, but was not then convinced that
marsh gas contained only one atom of carbon. Kekule's dis-
covery of the carbon-skeleton as the basis of all the hydrocarbon
radicals provided a final solution of the problem of the molecular
structure of organic compounds ; in the fatty-compounds the
carbon-skeleton is composed of open chains of carbon atoms, but
in benzene and aromatic compounds derived from it Kekule
recognised, in 1865, the existence of a closed chain or ring of
carbon atoms.

CHAPTER XVIII

THE CLASSIFICATION OF THE ELEMENTS

A. METALS AND NON-METALS

A " sceptical chemist's " views on the elements (Boyle

1661). — Until the close of the alchemistic period, it was
universally agreed that all matter was composed of a very
small number of " principles " or " elements," e.g. the four
elements, earth, air, fire and water of Aristotle and the
Peripatetic School, or the three principles, mercury, sulphur
and salt, of Albertus Magnus (1205-1280) and the alchem-
ists. The modern conception of elements is due to Boyle,
who sets out his views in the first pages of " The Sceptical
Chymist " * as follows :

" I perceive that divers of my Friends have thought it
very strange to hear me speak so irresolvedly, as I have
been wont to do, concerning those things which some take
to be the Elements, and others to be the Principles of all
mixt Bodies. But I blush not to acknowledge that I much
less scruple to confess that I doubt, when I do so, than to
profess that I know what I do not " (Sceptical Chymist,
1661, 1-2).

In discussing " the number of Elements or Principles " it
was agreed to use "elements and principles as terms equi-
valent : and to understand, both by the one and the other,
those primitive and simple bodies of which the mixed ones

1 The Sceptical Chymist: or Chymico-physical Doubts and Paradoxes
touching the Experiments whereby vulgar Spagyrists are wont to
endeavour to evince their Salt, Sulphur and Merctiry, to be the True
Principles of Things. London, 1661. Compare Works,l72.$,\\\. 261.

448

CH. xvin THE CLASSIFICATION OF THE ELEMENTS 449

are said to be composed, and into which they are ultimately
resolved" (ibid. p. 16). But Boyle insisted that "a man
may rationally enough retain some doubts concerning the
very number of those material ingredients of mixed bodies,
which some would have us call elements and others prin-
ciples" (ibid. p. 10), and urged that the problem must be
settled rather by experiments than by abstract reasoning.

Although Boyle was able to combat very effectively the
idea that the number of elements was limited to three or
four, more than a century elapsed before it was possible to
compile a reasonable list of the chemical elements, or to
make any serious attempt to classify them. But as soon
as the discovery of oxygen had been followed by the proof
of the composite character of water, Lavoisier was able to
set out, in 1789, a list of some thirty elements, which was
increased to fifty in the hands of Berzelius (1819), and now
amounts to more than eighty. These elements, with their
atomic weights, are set out in Table F, p. 491.

Lavoisier's classification of the elements (1789),— In
his Elementary Treatise, Lavoisier classified the elements as
follows (Works, I. 135) :

Simple substances
belonging to the
three kingdoms,
which may be re-
garded as the ele- 1 TT
ments of bodies. [Hydr°gen

I Sulphur
Phosphorus
Carbon
Muriatic radical
Fluoric radical 3
Boric radical3
( Lime

' Simple substances, I ^a|naesia

2 Simple sub-
stances,
metallic,
oxidisable,
and
salifiable.

1 Silica

'Antimony

Silver

Arsenic

Bismuth

Cobalt

Copper

Tin

Iron

Manganese

Mercury

Molybdenum

Nickel

Gold

Platinum

Lead

Tungsten
\Zinc

1 Lavoisier considered it as " probable that the four salifiable earths
set out above contain oxygen " ( Works, I. 126), but he was not able to

G G

450 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

The dominant position which Lavoisier assigned to
oxygen is seen in the text preceding the table :

'• The acidifiable substances, combining with oxygen and
being converted into acids, acquire a great tendency to
combination ; they become capable of uniting with earthy
and metallic substances, and it is by this union that neutral
salts are produced.

"This manner of viewing the acids does not permit me
to regard them as salts, although they have some of their
principal properties, such as solubility in water, etc. . . .
For the same reason, I shall no longer place the alkalis nor
the earthy substances, such as lime, magnesia, etc., in the
class of salts, and I shall designate by this name only
compounds formed by the union of an oxidised element
and a base" ( Works, I. 115-116).

Lavoisier's broad classification of the elements into three
groups, (i) me.tals, (2) inflammable elements, yielding acids
by combustion, and (3) oxygen, the essential constituent of
acids, bases and salts, is curiously similar to the alchemistic
system in which the three elements were (i) mercury, the
typical metal, (2) sulphur, the typical combustible substance,
and (3) salt.

Oxygen and chlorine as "supporters of combustion
(Davy, 1812).— Davy, in his Elements of Chemical Philo-
sophy (1812), followed Lavoisier's classification but made
certain important modifications. His three groups were :

(i) Elements that support combustion, — oxygen, and
chlorine, which Davy had just shown to be an element.

prove this experimentally. The alkalis were omitted because Lavoisier
regarded them as compounds, in view of Berthollet's proof of the com-
posite nature of ammonia. Silica is not mentioned in the text accom-
panying the table, but seems to have been classified with the earths,
rather than with the boric radical, because it was insoluble in water.

2 Arranged in alphabetical order in French. In the text they are
described, by an obvious misprint, as "acidifiable" instead of "salifi-
able."

3 Three unknown elements, assumed to be present as oxides in
muriatic, " fluoric," and boric acids.

xvin THE CLASSIFICATION OF THE ELEMENTS 451

(2) Inflammable or aridiferous element 's, not metallic, —
hydrogen, nitrogen, sulphur, phosphorus, carbon, and boron
(prepared in a crude state from boric anhydride in 1807),

(3) Metals, 38 in number, as contrasted with Lavoisier's
17, including sodium and potassium, barium, strontium,
calcium and magnesium; also "aluminum" and "silicum,"
which Davy had prepared in an impure condition from
alumina and silica ( Works, IV. 165-346).

Elements classified as metals and non-metals. — In
Lavoisier's system oxygen occupied a unique position as
a supporter of combustion for metals and non-metals
alike. Davy was fully justified in putting chlorine in the
same class, but Gay-Lussac showed in 1814 (p. 246) that
chlorine could not stand alone, since sulphur, chlorine and
iodine agreed together in their power of uniting both with
hydrogen to form hydracids and with oxygen to form oxy-
acids. Berzelius in 1825 grouped together under the name
of halogens (p. 247) the three elements chlorine, iodine and
fluorine (to which bromine was added in 1826), which
united directly with metals to form binary haloid-salts ; on
the other hand, he grouped together oxygen, sulphur and
selenium, all of which could unite with metals and with
non-metals to form ternary amphi-salts (p. 248).

After undergoing so many modifications, the attempt to
classify the non-metals as " combustibles " and " supporters
of combustion" broke down. Later writers were content
to distinguish broadly between METALS and NON-METALS,
doubtful elements being described sometimes as METALLOIDS.
There is not much difficulty in classifying the elements
themselves as metals and non-metals, but confusion and
difficulty arise when attempts are made to effect an identical
classification by studying the basic or acid qualities of the
oxides, or the properties of the chlorides and other com-
pounds of the elements.

G G 2

452 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

B. NUMERICAL RELATIONSHIPS AND THE PERIODIC
LAW

Dobereiner's "triads" (1829).— The existence of small
groups or " families " of elements became evident at a very
early date. The metals of the alkalis and of the alkaline
earths formed two of these natural groups. Amongst the non-
metals, the group of halogens and the elements of the oxygen
group were enumerated by Berzelius as typical constituents
of " haloid " and of " amphi " salts respectively.

Dobereiner, in 1829, directed attention to the fact that a
determination by Berzelius of the atomic weight of bromine
agreed with a prediction of his own "that the atomic weight
of bromine would probably be the arithmetic mean of the
atomic weights of chlorine and iodine " ; thus

351470+ 1 26-470 m 8o.9?0

where Berzelius found 78*383. So also in the case of sul-
phur, selenium and tellurium, ^ — — — - — — = 80741,

where the actual atomic weight of selenium was 79*263.

In the case of the metals, similar relationships were
detected in the groups

calcium, strontium, barium
lithium, sodium, potassium

as well as in certain groups where the atomic weights differed

but little, as in

nickel, copper, zinc,
platinum, iridium, osmium, etc.

(Pogg. Ann. der Physik., 1829, 15, 301-307.) These groups
of three elements, with atomic weights in arithmetical
progression, became known as Dobereiner's TRIADS.

Dumas's "natural series" and "common differences "
(1859). The accuracy of the relationship discovered by

xvin THE CLASSIFICATION OF THE ELEMENTS 453

Dobereiner was tested by Dumas in a memoir " On the
Equivalents of the Elements " (Ann. Chim. Phys., 1859, 55,
129-210), in which he sought to establish an analogy
between the "natural families" of elements and the
" natural series " of organic radicals. Dumas concluded
that :

" In a family of three elements, the equivalent of the
intermediate element may be equal to the mean of the equi-
valents of the two extreme elements ; but the contrary may
also occur in regard to the most closely related elements "
(loc. tit. p. 163).

Thus P ^ J5^} Mean = 81*25, whereas Br = 8o'o.

When dealing with the equivalents of organic radicals, on
the other hand, the averages are exact, and may, moreover,
be extended throughout a homologous series in the form of
arithmetical progression, as represented by the simple
formula a + nd^ e.g.

H = i, CH3 =1 + 14, C2H5 =1 + 2x14, C3H7 =1 + 3x14, etc.

Dumas suggested that something similar might be recog-
nised in the equivalents of the following families of
elements :

F =19 * 6 '» 8

CI =19+ 16-5= 35-5 S = 8+ 8= 16

Br =19+ 2x16-5+ 28 = 80 Se — 8+ 4x8= 40

I =19+ 2x16-5 + 2x28+19=127 Te = 8+ 7x8= 64

N =14 1Mg=i2
P =14+17= 31 Ca=i2 + 8= 20

As=i4+i7+ 44=75 Sr =12+ 4x8= 44

Sb =14+ 17 + 2x44=119 Ba=i2+ 7x8= 68

Bi =14+17 + 4x44 = 209 Pb =24+10x8=104 etc.

To Dumas belongs also the credit of recognising the
existence of numerical relationships between the equivalents

1 The equivalents of the oxygen and magnesium families require to
be doubled in order to give the modern atomic weights.

454 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

of different families of elements, e.g. the fluorine and nitro-
gen families, and the magnesium and oxygen families, thus :

Fluorine
Chlorine
Bromine

19

'Nitrogen
Phosphorus
Arsenic

4 J Common
* > difference

Iodine

127

Antimony

122 J 5

1 Magnesium

12

1 Oxygen

8 ')

Calcium

20

Sulphur

1 6 J Common

Strontium
Barium
Lead

4375
68-5

I03-5

Selenium
Tellurium
Osmium

3975 /-difference
64*5 1 4
99'5 J

These differences were precisely similar to those observed
in homologous series of organic radicals, e.g.

Ammonium 18 Methyl 15

mmonum 1 ety 15

Methylammonium 32 Ethyl 29 ]

Ethylammonium 46 Propyl 43

Propylammonium 60 etc. Butyl 57 etc. J 3

Dumas concluded that

" Since the radicals of mineral chemistry show amongst
themselves the same general relations as the radicals of
organic chemistry, there is certainly room to bring the two
chemistries more closely together than is the case at the
present day" (Ann. Chim. Phys., 1859, 55, 202).

Newlands's "relations between equivalents" (1864),—

Dumas was much restricted by using the old equivalents in
place of the newer atomic weights which were adopted im-
mediately afterwards in consequence of the work of Deville
and of Cannizzaro (pp. 514 and 343). If he had doubled
the equivalents of the oxygen and magnesium families, he
would have recognised that his scheme of " common differ-
ences " could be applied generally to all the elements. Thus
the four families which he considered specially would have
come together as in columns IV, V, VI and VIII of Table A.

1 The equivalents of the oxygen and magnesium families require to
be doubled in order to give the modern atomic weights.

xvm THE CLASSIFICATION OF THE ELEMENTS 455

TABLE A
NEWLANDS'S FIRST TABLE OF THE ELEMENTS1

I.

II.

III.

IV.

V.

Li 7

B ii

C 12

N 14

O 16

|( Mg 24

—

Si 28

P 31

S 32

•c \ Zn 65
H [Cdii2

I

Sn 118

As 75
Sb 122

Se 79'5
Te 129

—

Au 196

—

Bi 210

Os 199

VI.

VII.

VIII.

IX.

X.

F 19

Na 23

Mg 24

—

—

-o('Cl 35'5

K 39

Ca 40

Mo 96

Pd 106-5

•g ^ Br 80

Rb 85

Sr 87-5

v 137

—

H (-1 127

Cs 133

Ba 137

W 184

Pt 197

Tl 203

Pb 207

—

This table, which shows all the triads in a single scheme,
was put forward by Newlands in 1864 in a paper on " Rela-
tions between Equivalents " (Chem. News, July 30, 1864, 10,
59 ; reprinted in a pamphlet On the Discovery of the Periodic
Law, London, 1884, p. 8). He had already described a
number of relations between the equivalents in continuation
of the work of Dumas ; but this new table was a direct
result of the adoption of Cannizzaro's system of atomic
weights. Newlands directed attention to the following
features in the table :

(1) There is a constant difference of sixteen to seventeen
units between the values for the initial member of each group
and the first term of the corresponding triad.

(2) This difference is equal to the value assigned to
oxygen ; simple multiples of this difference may be recog-
nised in other parts of the table.

(3) " Silicon and tin stand to each other as the extremities
of a triad," the "central term or mean of the triad. ... is

at present wanting; thus ?L^±±*LLL8, = 73 '» (Periodic

1 The triads are arranged horizontally in the original table. The
following elements would now be assigned to different families : Li, Au,
Os, Tl, Pb, V.

456 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

Law, p. 8). This missing element ( Mendel eeffs eka-silicon)
was discovered in 1886 by Winkler, who called it Germanium.
(4) " Palladium and platinum appear to be the extremities
of a triad, the mean of which is unknown," and remains
unknown at the present day.

Newlands's " Law of Octaves " (1865).— To Newlands
belongs also the credit of publishing, in 1864, the first table
in which the elements were arranged in the order of their
atomic weights, and of assigning to the elements thus arranged
a series of consecutive atomic numbers (Periodic Law, pp. 7
and n). When this arrangement was made, it was seen
that:

"The numbers of analogous elements generally differ
either by seven or by some multiple of seven ; in other words,
members of the same group stand to each other in the same
relation as the extremities of one or more octaves in music.
Thus, in the nitrogen group, between nitrogen and phos-
phorus there are 7 elements ; between phosphorus and
arsenic 14; between arsenic and antimony, 14; and lastly
between antimony and bismuth, 14 also" (Chem. News,
Aug. 1 8, 1865; Periodic Law, p. 14).

This peculiar relationship Newlands proposed to term
provisionally the LAW OF OCTAVES. It was described at a
meeting of the Chemical Society on March i, 1866, and
was criticised "on the score of its having been assumed
that no elements remain to be discovered." One Fellow,
whose misdirected wit has provided a perennial warning to
rash critics, " humorously inquired of Mr. Newlands whether
he had ever examined the elements according to the order
of their initial letters ? " Newlands replied that " he had
tried several other schemes " and had found that " no rela-
tion could be worked out of the atomic weights under any
other system than that of Cannizzaro " (Periodic Law, p. 19) ;
but the paper was rejected and failed to secure attention
until similar relationships were put forward five or six years
later by Mendeleeff and by Lothar Meyer.

xvni THE CLASSIFICATION OF THE ELEMENTS 457

It is not necessary to consider in detail the table which
Newlands used to illustrate the Law of Octaves. In spite
of the larger number of elements, the families were reduced
from ten to seven, in order to maintain the idea of octaves.

Table

B.

Newlands's " Elements arranged in Octaves."

H

i F

8j

Cl

15 ! Co & Ni 22

Br

29 j Pd 36 i Te

43 j Pt &

Ir 50

i.i

2

Na

9 !

K

16 i Cu 23

Rb

30 i Ag 37

Cs

44' Os

51

G

3

Mg

xo :

Ca

17 i Y 24

Sr

3i

Cd 38

Ba& V

45 Hg

52

Bo

4

Al

IX

Ti

18 j Zn 25

Zr

32

Sn 39

Ta

46,T1

53

C

5

Si

12

Cr

19 i In 26

Ce&La

33

U 4°

W

47 Pb

54

N

6

P

13

Mn

20 As 27

Di& Mo

34

Sb 41

Nb

48 ! Bi

55

0

7

s

H

Fe

21 | Se 28 Ro & Ru

35

I 42

Au

49iTh

56

The members of the triads, recurring at intervals of fourteen
elements, were separated by unrelated elements occupying
the intermediate octaves, thus disguising to some extent the
relations indicated so clearly in the first table.

Mendeleeff on "periodicity" (1869).— This fault was
remedied in a short note "On the Relationships of the
Properties to the Atomic Weights of the Elements," sub-
mitted to the Russian Chemical Society in March, 1869, by
D. Mendeleeff. In Mendeleeff s table the number of
families was increased to nineteen, and, in spite of many
gaps and a large number of interrogation-marks, all
Table C. MendeleefFs First Table of the Elements.

H=i

Be = 9,4

C=I2

0=i6

Na = 23

Si = 28
8-32

Cu = 63,4

" = 68'

?=7o

As = 75

86=79,4

Br = 8o

Sr = 87^6
La = 94
Th = ii8?

Zr= 90
Nb = 94
Mo= 96
Rh-r 104,4
Ru= 104,4
Pd=io6,6
Ag=io8

Cd=II2

Ur=ii6
Sn = n8

Sb=I22

Te=i28?

J = I27

Cs=i33
Ba=i37

Ta=i82
W=i86
Pt= 197,4
Ir-i98
Os=i99

Au=i97?

458 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

the old relationships were shown clearly. Moreover, new
determinations of atomic weights permitted the elements
Be 9*4, V 51, Ta 182, to be placed in their correct positions
in the table.

Mendeleeff's first table was accompanied by the following
statement of the LAW OF PERIODICITY : —

" i. The elements, if arranged according to their atomic
weights, exhibit an evident periodicity of properties.

" 2. Elements which are similar as regards their chemical
properties have atomic weights which are either of nearly
the same value (Pt, Ir, Os), or which increase regularly (K,
Rb, Cs).

" 3. The arrangement of the elements ... in the order
of their atomic weights, corresponds to their so-called
valencies as well as, to some extent, to their distinctive
chemical properties, e.g. Li, Be, B, C, N, O, F.

" 4. The elements which are most widely diffused have
small atomic weights.

" 5. The magnitude of the atomic weight determines the
character of the element, just as the magnitude of the
molecule determines the character of a compound.

" 6. We must expect the discovery of many yet unknown
elements, — for example, elements analogous to aluminium
and silicon, whose atomic weights would be between 65 and

75-

" 7. The atomic weight of an element may sometimes be
amended by a knowledge of those of the contiguous ele-
ment. Thus the atomic weight of tellurium must lie
between 123 and 126, and cannot be 128.

" 8. Certain characteristic properties of the elements can
be foretold from their atomic weights " (German trans.,
Zeitschr. f. Chemie, 1869, 5, 405-406, from Russ. Chem.
Ges., 1, 60. English translation in Mendeleeffs " Faraday
Lecture," Trans. Chem. Soc., 1889, 55, 634-656).

Lothar Meyer on periodicity (1869). — A few months
after the publication of Mendeleeffs brief note, a paper
appeared on " The Nature of the Chemical Elements as a
Function of their Atomic Weights" (Liebig's Ann. Chem.

>cvm THE CLASSIFICATION OF THE ELEMENTS 459

Pharm., 1870, 7, Suppl., 354-364), in which Mendele'efFs
suggestions were developed in a remarkable way by Lothar
Meyer. Meyer defined the law of periodicity as follows :

"The properties of the elements are largely periodic
unctions of the atomic weight. Identical or similar pro-
oerties recur, if the atomic weight is increased by a definite
amount, which is at first 16, then about 46, and finally 88
:o 92 units " (loc. cit. p. 358).

Thus there is a difference of 16 units between the alkali-
metals lithium and sodium, and again between sodium and
potassium. These elements resemble one another closely,
but differ widely from all the elements of intermediate
itomic weight. " Only the valency rises and falls regularly
.ind equally in the two intervals," thus

Valency i 2 3 4 3 2 i

j,j . Li Be B C N O F
Elements

^ ^ A1 gi p s Q

'loc cit. p. 358).

Meyer's paper includes a complete classification of the
elements, which shows all the essential features of the
:amiliar table which Mendeleeff produced two years later,
ind in some points is in even closer agreement with the
nodern table on p. 462.

The paper also contains a curve showing the variations of
;ttomic volume as the atomic weight increases (see below,
:}. 472); this curve remains, even to-day, one of the best
illustrations that can be given of the principle of periodicity.

Mendeleeff on "The Periodic Law" (1871).— In his
complete paper on " The Periodic Law of the Chemical
Elements," Ann. Chem. Pharm., 1872, 8, Suppl.,
[33-229), Mendeleeff, following Lothar Meyer, adopted a
compromise (compare Table D,1 p. 462), between Newlands's

1 This table differs from Mendeleeff 's table as follows : (i) An addi-
•ional column O has been added ; (2) Series 2 and 3 are placed in the

460 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

system, in which only 7 families were recognised, and his
own, in which there were 19.

Nearly all the elements were classified, according to the
general scheme of Newlands's octaves, in seven GROUPS,
numbered I to VII, arranged in seven vertical columns
under the seven TYPICAL ELEMENTS1 (loc. cit. p. 152) :

Li Be B C N O F

When the elements, in the order of their atomic weights,
had been allotted to their proper families they formed not
fewer than thirteen horizontal SERIES or SHORT PERIODS (loc.
tit. p. 145), of which the first was occupied by hydrogen
only, and the second by the seven typical elements referred
to above.

Mendeleeff, like Newlands, now found that the chief
triads appeared in alternate short periods or octaves, and
therefore distinguished between the EVEN SERIES and the

ODD SERIES (IOC. tit. p. 145), e.g.

Fourth Series :

K

Ca

—

Ti

V

Cr

Mn

Fifth Series :

Cu

Zn

—

—

As

&

Br

Sixth Series :

Rb

Sr

—

Zr

Nb

Mo

—

Seventh Series :

Ag

Cd

In

Sn

Sb

Te

/

This difficulty was overcome (again following Lothar Meyer)
by placing the members of the even series on the left hand,
and the members of the odd series on the right hand side
of the vertical columns ; in this way the number of columns
was virtually increased to fourteen and each group could be

centre of the columns instead of left and right ; (3) The four series
8, 9, 10, II have been compressed into two lines by clustering 15 rare
earth elements into a single space in the table ; (4) The elements Sc,
Ga, Ge, predicted by Mendeleeff have been inserted, and other missing
elements are indicated by numbers of a consecutive series (see below,
p. 491).

1 The phrase was derived from Gerhardt's "simple inorganic types'3
(p. 425), the chief of which may be recognised in the series

— — — CH4 NH3 OH2 FH (compare C1H).

xvm THE CLASSIFICATION OF THE ELEMENTS 461

subdivided into two distinct families. Unfortunately, the
distinction between the odd and even series is not a sharp
one, since the alkali-metals are found in series two, three,
four, six, and the halogens in series two, three, five, seven.
This difficulty may be diminished by placing the elements
of series two and three either on the right, or on the left
according to the relation which they show to the elements
of the later series ; or the elements of series two and three
may all be placed in the centre of the column as in Table D,
p. 462.

Mendeleeff also recognised that whilst there were " sharp
differences between the last members of the odd series
(halogens), and the first members of the even series (metals
of the alkalis)," yet " the last members of the even series
resemble in many respects. . .' . the first members of the
odd series" (loc. cit. pp. 145, 146). Thus there was a
sharp break between Cl and K or Br and Rb, but a gradual
transition from Cr and Mn to Cu and Zn. Moreover, even
this interval was bridged over by a group of three metals,
which Mendeleeff placed together in an additional Group
VIII. The even and odd series were thus linked together
by a remarkably smooth TRANSITION SERIES of elements
(loc. cit. p. 146) :

Cr = 52, Mn = 55, Fe = 56, Co = 59, Ni = 59, Cu = 63, Zn = 65.
Mendele'eff therefore proposed to describe the triad of
Group VIII with the octave or short period on either
side as a LONG PERIOD (loc. cit. p. 146) of 7+3 + 7 = 17
elements. Similar long periods were recognised on either side
of the triads Ru= 104, Rh= 104, Pd= 1 06 and Os= 193 (?),
Ir=i95(?), Pt=i97.

A periodic classification of the elements. — A modern
classification of the elements which embodies all the essential
features of the tables of Lothar Meyer and Mendeleeff is set
out in Table D.

462 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

TABLE D
PERIODIC CLASSIFICATION OF THE ELEMENTS!

uj

Groups.

Full -|

Periods, c/3 O. I. II.

III. IV. V. VI. VII. VIII.

I.

I

H

II.

2

He

Li

Be

B

C

N

O

F

III.

3

Ne

Na

Mg

Al

Si

P

S

Cl

TV

f 4

Ar

K

Ca

Sc

Ti

V

Cr

Mn

Ye Co Ni

A v .

I 5

Cu

Zn

Ga

Ge

As

Se

Br

v

/ 6

Kr

Rb

Sr

Y

Zr

Nb

Mo

43 iRu Rh Pd

v •

I 7

Ag

Cd

In

Sn

Sb

Te

I

VI.

f8,'

Xe

Cs

Ba

15 RARE
EARTHS

72

Ta

W

75 Os Ir Pt

in

Au

Hg

Tl

Pb

Bi

84

85!

VII. 12

Nt

87

Ra

|89

Th

91 U

In this table the number of columns has been increased
to nine, in order to include the newly-discovered RARE
GASES, shown in Group O.

He <3'99, Ne 20*2, Ar 39*88, Kr 82*92, Xe i3o'2, Nt 222-4.

The discovery of these six gases not only adds an additional
column to the table, but justifies its subdivision into seven
chief periods. These periods are as follows :

(1) The "hydrogen-period," which may perhaps contain
one or two other light gases. It may be extended, if
desired, to include helium, in which case neon would be
transferred to the end of Period II, and so on down the
series.

(2) Two short periods of eight elements. — Each period
starts with one of the rare gases, and both periods are similar
and complete. In Table E, the atomic weights are given
in round numbers and are seen to differ on the average by
1 6 units ; if exact atomic weights are taken the average
difference is 16*06, agreeing somewhat closely with the
atomic weight of oxygen. The average differences between
Period III and the initial portion of Period IV is i8j units.

1 See footnote, p. 459.

xvni THE CLASSIFICATION OF THE ELEMENTS 463

pq*?^ - H" I I

M|a» H<M

<U ON°0 r« *- I II

c/3 t- \s H N I
<! t*» \t»

M

H

PH O vOXi

Q C.

a

co M M

ffi

r^ "^ o\

CJ ro S

00

a §

o 9o -^ co

<N °O WO

CJ

r
L)

C 10 Ox r"

-0 K

u M

O\
< 0 °0

Os

O °O

PH ON

|_

O CT>

I |

* ;T

°o

00

10

oo

ro K.
00 ^

^ M i

Jz; ON°O ^H oo I
N5, I Ml &

ON O SH.LiJV3 1 1

00 C^ aava 1

PQ

tn ^
U to

O ^
CO Os

464 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

(3) Two long periods of eighteen elements. —These two
periods are again precisely similar to one another, with the
single exception of a missing element, of atomic weight
about 100, which is required to fill a gap below manganese,
and between molybdenum and ruthenium. The average
difference between the atomic weights of the long Periods
IV and V is nearly 46, a difference which persists between
Periods V and VI, until the elements of the rare earths
are reached, when the difference between analogous elements
jumps suddenly to 89, but finally drops back to 54 when
the rare earths are passed. This is shown in Table E,
which carries the differences forward to the end of the list
of elements.

(4) A long period of some thirty-two elements, inclu-
ding the elements of the rare earths. — Mendeleeff assumed
that the two long periods, including series 4-5 and 6-7,
would be followed by three more long periods of identical
type, including series 8-9, 10-11, and 12-13 (?). He there-
fore left spaces (loc. cit. p. 151), for a group of transition-
elements of atomic weight about 150, between periods
8 and 9, and for another group of atomic weight about
250, at the end of period 12, to correspond with the group
Os, Ir, Pt, of atomic weight about 193 between series 10
and n. He was also obliged to assume that there was an
element missing in each of the following families :

Cu Zn As

Ag Cd In Sn Sb

Au Hg Tl Pb Bi

Lothar Meyer, on the other hand (Ann. Chem. Pharm.,
1870, 7, Suppl., 356), suppressed these gaps and assumed
that the differences between analogous elements increased
abruptly from 16 to 46, and then again from 46 to 90 on
passing from Ba — Sr to Ta — Nb (see quotation, p. 459).
None of Mendeleeffs hypothetical transition-elements

xvin THE CLASSIFICATION OF THE ELEMENTS 465

have been found nor have the gaps between silver and
gold, cadmium and mercury, tin and lead, or antimony and
bismuth been filled by the discovery of additional elements
belonging to any of these families. It is therefore probable
that these gaps do not in fact exist, and that Lothar Meyer
was right in his suggestion that the differences in a family
of elements may be 16 or 46 or 90, as in the two series

Li = 7 Na=23 ' K - 39 Rb = 85j €3=133
Differences 16 16 46% 4J\

N=i4 P = 3i As = 75 Sb-i2o Bi=2o8
Differences 17 44 45 88

There is, however, a gap of about 40 units at the point
where the differences undergo this sudden increase. This
gap, which extends from barium (Ba -Sr=i37~88 = 49) to
tantalum (Ta-Nb= 181 - 93= 88) is filled by a series of
RARE EARTH ELEMENTS, derived mainly from monazite sand.
In addition to scandium and yttrium, two elements of much
smaller atomic weight, fourteen elements of known atomic
weight have been separated (see Table F, p. 492) as follows :

La Ce Pr Nd Sa Eu Gd Tb Ds Ho Er Tm Yb Lu

i39'o 140-25 140-6 144-3 i5<>'4 I52'c 157-3 159-2 162*5 163-5; l67'7 ^S'S 172*0 174-0

These elements exhibit a remarkable resemblance to
one another; their compounds are so similar that their
separation has occupied more than a century, and even now
may not be quite complete. They differ far less than
MendeleefFs transition-elements and cannot possibly be
scattered over the eight families of the short periods, or the
eighteen families of the long periods of the conventional
classification. With unmistakable clearness these elements
all claim a place in the same family as scandium and yttrium
and must be placed with these elements in Group III of
Tables D and E.1 The regular periodic change in the

1 Armstrong, Encycl. Brit., Ninth Edition, 1902, XXVI, 712;
compare Biltz, Ber., 1902, 35, 562-568.

H H

466 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

properties of the elements is evidently checked at this point
and only goes forward again after some 40 units have been
added to the atomic weight.

There is reason to think (see later, p. 492) that there
may be sixteen elements between barium and tantalum, of
which fourteen are now known ; if this be correct, the total
number of elements in this period would be thirty-two.

(5) A period composed of radioactive elements. — The
last period is a mere fragment, but includes four well-defined
elements of known atomic weight, occupying alternate
columns in the first octave of Period VII. and correspond-
ing closely in their properties with the earlier members of
these four families. These four elements are

Niton Nt 222 Thorium Th 234

Radium Ra 226 Uranium Ur 238

They differ from the lighter elements in that they are all
RADIOACTIVE, i.e. they constantly liberate energy, which
appears to be derived from- the DISINTEGRATION of the atom.
In the absence of any known synthetic process, a radio-
active element must either decay very slowly, or must be
reproduced continuously from some element of greater
atomic weight. Uranium and thorium, the two elements
of greatest atomic weight, show very slight radioactivity and
are credited with a very slow rate of decay. All other radio-
active substances appear to be derived, directly or indirectly,
from these two PARENT ELEMENTS.

Some of these radioactive elements, or, more briefly,
RADIO-ELEMENTS,1 which enjoy a relatively long life, are
found in appreciable quantities in the ores from which the
parent-elements are derived, and can be separated by ordi-
nary chemical methods. Thus the Bohemian pitchblende
of Joachimsthal (crude U3O8) has given radium, polonium
and actinium, in the uranium series, whilst thorium-

1 Soddy, The Chemistry of the Radio- Elements, Parts I and II,
London, 1914.

xviii THE CLASSIFICATION OF THE ELEMENTS 467

minerals have given mesothorium and radiothorium. Radium
has been prepared in sufficient quantity for an ordinary
determination of atomic weight, and the atomic weight of
niton, a gaseous emanation produced during the decay of
radium, has been determined by special modifications
of ordinary methods (Gray and Ramsay, Proc. Roy. Soc.,
1911, A, 84, 536-550). The chemical properties of other
radio-elements can be inferred from the readiness or other-
wise with which they are carried down with, or can be
separated from, typical elements of smaller atomic weight.
On these grounds polonium has been placed in the same
family as tellurium, in a vacant space immediately after
bismuth. In the same way, actinium is assigned to the
same family as lanthanum, the first of the rare-earth
elements, and probably occupies the vacant place between
radium and thorium. The vacant place between thorium
and uranium is perhaps occupied by one of the disintegra-
tion-products of uranium (UrX2) with an average life of
only i '65 minutes ; but no radioactive element, however
transient, has been found to occupy the vacant places
below caesium and iodine in the families of the alkali-
metals and the halogens respectively.

The total number of radio-elements that have been sched-
uled is about 40. These are all crowded into the narrow
range from thallium (204) to uranium (238), covering only
12 places in the periodic classification of the elements. It
is already occupied by three ordinary elements (thallium,
lead and bismuth) in addition to the four radio-elements
of known atomic weight (niton, radium, thorium and
uranium). As none of the radio-elements are allied to the
alkali-metals or the halogens there are really only 10 places,
of which 7 are clearly occupied, leaving only 3 vacancies
for the 36 remaining radio-elements. This crowding is
explained by the existence of ISOTOPIC ELEMENTS, having
similar (or perhaps identical) chemical properties (and

H H 2

468 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

perhaps identical spectra), but differing in atomic weight
by a very small number of units. The story of the rare-
earth elements is, in fact, assumed to be repeated at almost
every stage of the progress from thallium to uranium. The
40 radio-elements are distributed as follows :

Groups. o. I. II. III. IV. V. VI. VII.

Period VI. Tl(3) Pb(9) Bi(4) -(7)1 -(o)

Period VII. Nt(3) - (o) Ra(4) -(2)2 Th(5) -(i) Ur(2)

The ultimate products of disintegration appear to be
helium and lead, but the existence is indicated of 5 iso-
topic forms of lead, devoid of radioactivity, in addition to
4 radioactive forms. The theory of isotopic elements has
received confirmation from the fact that samples of " lead "
derived from different sources have been found to show
marked differences of atomic weight (p. 303)

C. ILLUSTRATIONS AND APPLICATIONS OF THE
PERIODIC LAW.

Periodicity of valency. — The ebb and flow of valency,
which was described by Meyer in 1869 (p. 459), was
discussed more fully in 1871 by Mendeleeff. He suggested
(foe. cit. p. 141) that whilst the valency in the hydrides
varied in the order indicated by Meyer (p. 459), the
valency in the oxides was identical with the group-number,
and varied in the order 12345678. The typical
hydrides and oxides may therefore be tabulated as follows : —

Group I. II. III. IV. V. VI. VI I. VIII.

f i 2 34321

Hydride^ LiH — BH3 CH4 NH3 OH2 FH

[NaH [CaH24] — SiH4 PH3 SH2 C1H

1 Including polonium. 2 Including actinium.

3 The hydrides of the earlier groups were not known to Mendeleeff.

4 MgH2 is not known.

xvm THE CLASSIFICATION OF THE ELEMENTS 469

12345678
Li.20 BeO B203 CO2 N2O5

( includes

Na20 MgO A1203 SiO2 P2O5 SO3

The "typical" hydrides and oxides at the head of Men-
dele'effs tables were selected so as to fit into a regular
series. Thus in Group I. the higher oxides of sodium and
potassium, Na2O2 and K2O4, were rejected as " peroxides "
(which did not form salts) ; but the typical salt-forming
oxides, cupric oxide, CuO, and auric oxide, Au2O3, were
also set aside in favour of the less stable oxides Cu2O
and Au2O. The regular ebb and flow of valency with
increasing atomic weight is, however, one of the most
striking features of the periodic classification, and is set
out in a very effective way in a chart attached to the
Faraday Lecture which Mendeleeff delivered before the
Chemical Society in 1889 (Trans. Chem.. Soc , 1889, 55,
facing p. 656), where all the salt-forming oxides are
tabulated in a series of well-defined waves. These
regular changes of valency afford the sole justification for
crowding the elements into eight groups, to which a new
Group o is now added to include the inert gases of the
helium family, which appear to be devoid of all combining-
power. Moreover, no excuse, but those arising from identity
of valency in selected compounds, can be given for
including in the same column elements so diverse as sulphur
and chromium, chlorine and manganese, or potassium and
copper.

Periodicity of atomic volumes (L. Meyer, 1869). — The
most remarkable feature of Lothar Meyer's paper is a
diagram, which shows the ebb and flow of the ATOMIC
VOLUME1 of the solid elements as the atomic weight
increases.

1 The atomic volume = At. wt. x specific volume or At. wt.-r- density.

470 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

11 The curve ... is divided by five maxima1 into six sec-
tions, which have somewhat the form of a series of suspended
chains, of which the second and third, and likewise the
fourth and fifth, are similar to one another and occupy nearly
equal portions of the horizontal axis.

" If one considers now the placing of the elements on the
curve, one finds in corresponding places on similar portions
of the curve elements with similar properties.

" All easily melted, volatile and gaseous elements are on
ascending branches of the curve : those fusible with difficulty
... on descending branches. . . near the minimum," where
also are found the few elements that do not obey the law of
Dulong and Petit, and so forth (Ann. Chem. Pharm., 1870,
7, Suppl., 359-363).

The chief feature of the atomic volume curve (Fig. 52,
pp. 472 — 473), as Lothar Meyer points out, is its subdivision
into a series of six catenary curves, which correspond with the
first six periods of the modern table of the elements shown on
p. 462. In the region covered by the two short catenaries
II and III, Newlands's Law of Octaves may be applied
with little modification, but the long catenaries IV and V
(which form an equally close pair), indicate clearly that the
interval between analogous elements is now much greater
than an octave ; the atomic volume curve thus makes the
distinction between the short and long periods compulsory
rather than optional.

The subsequent course of the curve is less certain.
Lothar Meyer represented the whole of the elements from
caesium 133 to bismuth 208, as lying upon a single big
catenary VI, corresponding with the big period VI of 32
elements shown in Tables 2, 3 and 4. Later writers, fol-
lowing Mendeleeff, have assumed that there are two
long periods here, separated by a maximum occupied by
an unknown alkali-metal. But there is no experimental

1 Corresponding, in Lothar Meyer's curve, with the five alkali-
metals.

xviii THE CLASSIFICATION OF THE ELEMENTS 471

evidence to support this view ; on the contrary, the few
measurements that have been made of the atomic volumes
of rare-earth elements suggest that these would fall on a
smooth curve running down gently from barium to tantalum
and tungsten. It is remarkable that Lothar Meyer, with
the help of his atomic volume curve should have antici-
pated at so early a date the modern classification of the
elements into five complete periods and two fragments.

Later measurements have added a number of additional
points to the curve. The caesium peak has been realised
experimentally, and the discovery of niton and radium
has rendered probable the existence of yet another peak
beyond the region covered by Lothar Meyer's curve. The
most striking additions to the curve are points corresponding
with 10 elements which are gaseous at ordinary tempera-
tures.1 Some risk attaches to the inclusion of liquefied gases
in a comparison which is primarily concerned with solid
elements, but the additional points are by no means devoid of
interest. Thus the inert gases of Group o are found on the
ascending portions of the curves, just below the alkali metals,
with the solitary exception of helium, which rises above and
dominates the point representing lithium. If the atomic
volume were the only periodic property, it would be con-
venient to place the inert gases at the ends of Periods I to
VI, instead of at the beginning of Periods II to VII, but
the point is not important and there are definite advantages
in placing the gases of zero valency at the beginning of the
table. Liquid oxygen and fluorine fall into position on
the second catenary, but liquid nitrogen produces a small
secondary peak on the curve.

Periodicity of atomic heat (Dewar, 1913).— According
to the Law of Dulong and Petit, the ATOMIC HEAT is a
constant quantity which does not vary with the atomic
weight of the element. No better illustration could be

1 Chlorine was the only substance studied by Meyer in the liquid
state.

472 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

given of the occasional absence of periodicity than the
specific heat, which can be plotted against the atomic weight
in the form of a rectangular hyperbola, or the atomic heat,
which gives a straight line parallel to the axis of atomic
weight. But this simple law ceases to hold at low tempera-

Atomic Heat

10 20 30 40 5O 60 70

FIG.

9O 1OO

52. DEWAR'S ATOMIC

tures. Thus the average values and average errors for
twenty-four elements from Mg to Bi (p. 365) are as follows :

Temperature. Atomic heat.

100° to 20°C. 6'2±o'2

20° to -i88°C 5-2 ±0-5

- 195° to - 253°C, 2-2 ± ro

Dewar, to whom the measurements in the lowest range of

XVIII

THE CLASSIFICATION OF THE ELEMENTS 473

temperatures are due, points out that " when plotted in terms
of their atomic weights, they reveal definitely a periodic
variation resembling generally the well-known Lothar Meyer
atomic volume curve for the solid state " (froc. Roy. Soc.,
1913, A., 89, 167). The striking character of this resem-

Atomic Volume

VI

VII d

,U-.

Q40

RAX£~EARTH'E-r*;

HE

24O

140 150 160 170

'URVE OK ATOMIC VOLUMES.

180 19O 2OO 21O

22O 23O

blance is well shown in Fig. 52, above. The similarity
of the two curves suggests that, at low temperatures,
equal volumes of different elements have the same
heat capacity, instead of equal numbers of atoms, as re-
quired at higher temperatures by the law of Dulong and
Petit.

Correction of atomic weights, (a) Valencies corrected
or confirmed. — When the equivalent of a metal has been

474 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP

determined there is often some doubt as to its valency and
therefore as to its atomic weight. If the metal can be
separated, its specific heat will give the information required.
If not, the isomorphism of its compounds is the most trust-
worthy guide ; but there are so many exceptions to the law of
isomorphism (e.g. Ag2S and PbS are isomorphous) that
further confirmation is often needed. It is just here that
the periodic classification finds its most important practical
application. The following examples show also the advan-
tages of the periodic system over the earlier system of
classification into families.

i. Beryllium. Equivalent 4'55- Atomic weight
4-55x2 = 9*1. The compounds of beryllium resemble
those of aluminium, e.g. the hydroxide is soluble in caustic
alkalis as well as in acids, the carbonate decomposes spon-
taneously at ordinary temperatures, the metal is prepared by
electrolysing the fusible double fluoride and dissolves in
alkalis with liberation of hydrogen. Beryllium was therefore
assumed, like aluminium, to be tervalent, with atomic weight
3 x 4-55 = 13*65. * But there is no vacant place here in the
periodic classification, and Newlands in his first complete
table of the elements (p. 457) was obliged to place beryllium
(with valency 2 and atomic weight 9) above magnesium, in
a position which he had assigned a few months before (from
a consideration of the individual family relationships) to
lithium (p. 455). Mendeleeff (Ann. Chem. Pharm., 1872,
8, SuppL, p. 1 66) was able to justify the position thus
assigned to beryllium by showing that

(a) there was a steady gradation in the series Li, Be, B ;

(b) the differences between beryllium and magnesium
were of the same character as in the corresponding
members of the two adjacent families,

Li

Na

Be

Mg

B

Al

1 Equivalent 6'8, if O = 8.

xviii THE CLASSIFICATION OF THE ELEMENTS 475

(c) the resemblance between beryllium and aluminium
was similar to that which Newlands had detected between
lithium and magnesium and that which is found between
boron and silicon, thus :

Li Be B

Mg A. Si

The correctness of the smaller atomic weight was proved
finally by determinations of the vapour-density of beryllium
chloride, which gave the molecular weight as 817 (Be = 9,
C12= 71) between 686° and 812° (Nilson and Pettersson,
Comptes rendus, 1884, 98, 988-990).

2. Indium, Equivalent 38-27. Atomic weight 3 x 38-27 =
114-8. This element, discovered by spectrum analysis1 in
the Freiburg zinc ores, by Reich and Richter in 1863, was
regarded by Newlands as divalent, but was placed correctly
by Lothar Meyer, with the help of his atomic volume curve.
Mendele'eff (Joe. cit. p. 178) confirmed the higher atomic
weight for the element by measuring its specific heat, which
he found to be 0*055 (agreeing closely with a number 0*057
given by Bunsen), whence the approximate atomic weight
equals 6*3 -"-0-055 = 115.

3. Uranium. Equivalent 59-6. Atomic weight 59*6 x 6
= 238*5. This element was regarded by Newlands as
divalent, at. wt. 60 x 2 = 120. Lothar Meyer showed that it
could not be placed on the atomic volume curve at this
point, and suggested Ur = 60 x 3 = 180. Mendeleeff (loc. tit.
pp. 178-184), after tracing its analogies with chromium,
molybdenum and tungsten, assigned to it the atomic weight
60x4 = 240. Its recently-discovered radioactivity has fully
justified the view that uranium should be placed at the end

1 Compare thallium, discovered in pyrites by Crookes (1861) and
gallium (p. 477), discovered in zinc blende from the Pyrenees by
Boisbaudran (1875).

476 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

of the table as the element having the greatest atomic weight
that is known.

Correction of atomic weights, (b) Equivalents
corrected. — Certain elements, which were found to be mis-
placed, when arranged strictly in the order of their atomic
weights, could be restored to their own families by assuming
smaller errors, due to the use of inexact equivalents. Thus
gold (196*2) was placed after osmium, iridium and platinum
(198-6, 1967, 1 96*7) ; later measurements have proved the
correctness of this sequence, thus :

Os Ir Pt Au

190-9 193-1 195-2 197-2

On the other hand, three cases are known in which the
atomic weights refuse to conform to the order suggested by
the periodic classification. These are

A K

39-88 39-10

Co Ni

58-97 58-68

Te I

127*5 126-92

The significance of these deviations is discussed below
(p. 494).

Prediction of missing elements. — Mendele'eff was so con-
vinced of the validity of the periodic law that, unlike
Newlands and Lothar Meyer, he altered atomic weights that
did not conform to his system and made detailed predictions
of the properties of unknown elements, which could be
verified or disproved by experiment. Missing elements,
were named by prefixing the Sanskrit numerals eka-^ dwi-,
tri- to the element in the odd or even series below which
vacancies were found. Detailed predictions were made of
the properties of

ekaboron = 44, ekaaluminium l = 68, ekasilicon = 72.

1 Mendeleeff gave a more detailed prediction of the properties of
ekaaluminium (Comptes rendus, 1875, 81, 970-971) after the discovery
of gallium had been announced by Lecoq de Boisbaudran.

THE CLASSIFICATION OF THE ELEMENTS 477

These were vindicated in a striking way by the discovery of
;he three patriotically-named elements

scandium, gallium, germanium,

(Nilson, 1879) (Boisbaudran, 1875) (Winkler, 1886)

Some of the chief predictions, and their verification, are set
out below. In the case of scandium or ekaboron, the pro-
perties of the metal and its anhydrous chloride were pre-
dicted by Mendeleeff, but are still unknown. The narrative
shows the most important points of agreement between
MendeleefFs predictions and the observed properties of the
three elements.

Ekaborop. Scandium.

Atomic Weight Eb = 44 Sc = 44- 1

Formula of Oxide Eb2O3 Sc2O3

Density of Oxide 3*5 3'8

The oxide is a weak base, intermediate between
alumina and magnesia. It forms a gelatinous hydrate,
carbonate and phosphate. The sulphate is soluble (but
less so than aluminium sulphate). The chloride is more
easily hydrolysed than MgCl2, liberating HC1. The salts
are colourless.

Ekaaluminium. Gallium.

Atomic Weight 68 69-9

Density 5-9 5-93

Melting-point { ^^ } 3°'i°C.

Formula of Oxide Ea2O8 Ga2O3

Density of Oxide 5*5

The metal is not volatile ; it is attacked slowly by air
and water, by acids and by alkalis. The oxide being
more basic than A12O3 and less basic than MgO, it is
precipitated by barium carbonate. The gelatinous
hydroxide is soluble in acids and alkalis. The sulphide
is precipitated by sulphuretted hydrogen and is not

478 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP,

soluble in ammonium sulphide. The metal forms a
soluble trichloride, a soluble sulphate and an alum.

Ekasilicon. Germanium.1

Atomic Weight £5 = 72 Ge=72'5

Density Es, 5-5 Ge, 5-469

Density of Oxide EsO2, 4*7 GeO2, 4'7°3

Density of Chloride EsCl4, 1-9 GeCl4, 1-887

B.P. of Chloride <ico° 86°C.

Density of Ethide Es(C2H5)4, 0-96 Ge(C2H5)4, <i

B.P. of Ethide 160° 160°

The metal, prepared by reducing the oxide, or by
the action of sodium on the double fluoride, is not
easily fusible, but burns to a very refractory white
dioxide. The oxide is feebly basic ; on adding acid to
its alkaline solutions, it may remain in solution as a
soluble hydrate or be precipitated as an insoluble
metahydrate. The disulphide is insoluble in water, but
dissolves in ammonium sulphide. The fluoride, like the
chloride, is a volatile liquid.

In the range covered by the rare-earth elements Men-
dele'efFs predictions were less successful, thus Ekacaesium,
EC =175, and Ekaniobium, £11=146, are still unknown;
but Mendele'eff was struck by " the absence ... of almost
a whole long period (beginning from Ce=i4o)" and
concluded that this " can scarcely be regarded as accidental
and probably has its origin in the nature of the elements "
(Ann. Chem. Pharm.^ 1872, 8, Suppl., 205). On the other
hand his Ekamanganese, Em = 100, and Trimanganese,
Tm = 190, are still to be looked for, whilst his Dwicsesium,
Dc=22o, and Ekatantalum, £1 = 235, correspond with two
of the three gaps at the commencement of the radioactive
Period VII.

Secondary relationships in the periodic system. — The
classification of the elements by vertical columns into
groups and families has come to be regarded as the

1 For a more detailed comparison, see Freund, Study of Chemical
Composition, pp. 480-481.

xvm THE CLASSIFICATION OF THE ELEMENTS 479

dominant property of the periodic system, and has been
made the basis for the descriptive portion of many text-
books. The undue emphasis of this property, which is
mainly an expression of valency-relationships, has created
an artificial atmosphere, in which the close similarity
between manganese and iron or between the compounds
of copper and mercury, is overlooked in favour of a strained
analogy between manganese and chlorine, or between
cuprous chloride and common salt. In the same way the
two combustible elements, sulphur and phosphorus, are
only rarely mentioned together, since they have the mis-
fortune to differ in valency.

For this fault Mendeleeff cannot be blamed, since he
insisted quite as much on the analogy between adjacent
elements in the same horizontal series as on that between
alternate elements in the same vertical group and used both
of these "atomic analogies" in predicting the properties of
missing elements. He was also careful to point out the
relationships which are observed between certain elements
lying on the same diagonal line (see above, p. 475.)-

Some of these diagonal analogies are strong enough to
overcome the limitations imposed by unequal valencies, as
in the case of boron and silicon, their fluorides and double
fluorides, BF3, SiF4, KBF4, K2SiF6, and the glass-forming
borates and silicates. In other cases valencies are developed
in defiance of the simpler rules of periodicity, as when
boron becomes quadrivalent in the hydrides B2H6 and B4H10,
as if in protest against its separation from the non-metals
carbon and silicon. Many similar cases might be quoted
of unexpected resemblance between elements lying
diagonally on lines sloping down from left to right in the
table.

The importance of the horizontal relationships is seen
most clearly in Mendele'efFs group of transition-elements.
Metallurgists will recognise in the series
V, Cr, Mn, Fe, Co, Ni,

48o HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

almost the whole of the elements, apart from carbon, silicon,
and tungsten, which are used in the manufacture of modern
special-steels. Equally remarkable is the series of
isomorphous bivalent sulphates studied by Mitscherlich
(P- 372),

MnSO4, FeSO4, CoSO4, NiSO4, CuSO4, ZnSO4,

which run right across a series of columns labelled VII,
VIII, I, and II, and headed by the symbols R2O7, RO4
R2O and RO. The horizontal relationship reaches its
climax in the case of the rare-earth elements which run
consecutively, with identical valency and almost identical
properties, over a range of 40 units of atomic weight.

If injustice is not to be done to the memory of Mendeleeff,
it is important that these secondary relationships should be
recognised and emphasised.

D. ATOMIC WEIGHTS AND ATOMIC NUMBERS.

Front's hypothesis (1815). — Almost immediately after the
first atomic weights had been determined, attempts were
made to discover numerical relationships between them.
The most important of these, now known as PROUT'S
HYPOTHESIS, was put forward anonymously in a paper " On
the Relation between the Specific Gravities of Bodies in
their Gaseous State and the Weights of their Atoms"
(Annals of Philosophy, 1815,6,321-330). It is generally
expressed by a statement that

" The atomic weights of the elements are exact multiples oj
the atomic weight of hydrogen"

The view first put forward by Prout was :

" That all the elementary numbers, hydrogen being con-
sidered as i, are divisible by 4, except carbon, azote, and
barytium *, and these are divisible by 2 " (he. cit. p. 330).

1 i.e. barium.

xvin THE CLASSIFICATION OF THE ELEMENTS 481

This view was supported by tables showing the actual
densities of various gases relatively to hydrogen, and the
hypothetical densities of many solid elements and com-
pounds ; from these the " specific gravities " or equivalents
of the elements were deduced as a series of integral
numbers, thus :

H C N P O S Ca Na Fe Zn Cl K Ba I
i 6 14 14 16 16 20 24 28 32 36 40 70 124

Hypothetical values for 23 other elements were all shown
as multiples of 4.

Prout on the specific gravity of hydrogen (1815).—
Whilst the idea that all the atomic weights are even
multiples of the atomic weight of hydrogen is obviously
untrue, credit must be given to Prout for determining for
the first time the correct density of hydrogen, by an indirect
method, as follows :

" The specific gravity of hydrogen, on account of its
great levity, and the obstinacy with which it retains water,
has always been considered as the most difficult to take of
any other gas. ... It occurred to me that its specific
gravity might be much more accurately obtained by calcu-
lation from the specific gravity of a denser compound into
which it entered in a known proportion. Ammoniacal gas
appeared to be the best suited to my purpose, as its specific
gravity had been taken with great care by Sir H. Davy, and
the chance of error had been much diminished from the
slight difference between its specific gravity and that of
steam. . . . The specific gravity of ammonia, according
to Sir H. Davy, is '590164, atmospheric air being 1*000.
We shall consider it as '5902. . . . Now ammonia consists
of three volumes of hydrogen and one volume of azote
condensed to two volumes. Hence the specific gravity of
hydrogen will be found to be '0694,1 atmospheric air being
i -oooo. It will be also observed that the specific gravity of

1 " Let * = sp. gr. of hydrogen, then 3* +'9722 =.59O2.
Hence *=ri8o4-'9722=-o694."

I I

482 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

oxygen as obtained above is just 16 times that of hydrogen
as now ascertained, and the specific gravity of azote just
14 times"1 (loc. cit. p. 322).

Hydrogen as "protyl" (Prout 1816).— This correct
determination of the density of hydrogen 2 was essential for
the development of the stimulating and suggestive hypothesis
of integral 3 atomic weights, which led Prout to conclude
that :

"If the views we have ventured to advance be correct, we
may almost consider the Trpwrrj vXrj of the ancients to be
realised in hydrogen ; an opinion, by the by, not altogether
new. If we consider this to be the case. . . the specific
gravities, or absolute weights of all bodies in the gaseous
state, must be multiples of the specific gravity or absolute
weight of the first matter (TT/OWTT? vXy), because all bodies in
a gaseous state which unite with one another unite with
reference to their volume" (Annals of Philosophy, 1816, 7,
US)-

Atomic weights in relation to oxygen. — Two years
before Prout's paper appeared, Thomas Thomson pointed
out (Annals of Philosophy, 1813, 2, 114) that if oxygen be
taken as unity "there are eight atoms of simple bodies
whose weights are denoted by whole numbers ; namely

Oxygen i Sulphur 2 Potassium 5 Arsenic 6
Copper 8 Tungsten 8 Uranium 12 Mercury 25."

The first of these ratios was confirmed by Dumas, but most
of the others are obviously incorrect. Dumas concluded
that :

" If oxygen be represented by 8, sulphur must be repre-
sented by 1 6. There exists then between these equivalents
the simple ratio i : 2, of which organic chemistry presents
so many examples" (loc. cit. p. 148).

1 " i 'II in -f- -0694=16 and -9722^-0694-14."

2 Thomson in 1813 using an earlier value for the density of hydrogen,
had only been able to find one integral multiple.

3 No longer even numbers in the second (1816) paper.

xvni THE CLASSIFICATION OF THE ELEMENTS 483

The average of five experiments actually gave the ratio 8 :
i6'oi, agreeing closely with the modern ratio oxygen : sulphur
= i6:32'o7. Dumas (still using equivalents rather than
atomic weights) also gave

Co : Ni : Sn= 29-5 : 29-5 : 59 = i : i : 2 and

N : Fe : Cd = 14 : 28 : 56 = i : 2 : 4 ;

these numbers may be compared with Crookes' suggestion
(Phil. Trans., 1908, A. 209, 44) that

B : Sc : Y : Yb= u'o : 44*1 : 89*0 : 173*0 = i : 4 :8 : 16
No special importance is now attached to the fact that
the atomic weight tables do actually show some very close
multiples of the atomic weight of oxygen, or that this
number, as Newlands pointed out in 1864 (The Periodic
Law, p. 6) is even more common amongst the differences ;
thus, when O= 16,

8 = 32-07 Ti = 48-i Br = 79-92 Mo = 96-0
8-0=16-07 K-Na=i6'io Na- Li= 16-06, etc.

It is, however, remarkable that integral atomic weights are
much more common when O=i6 and H=i'oo8 than
when H=i and 0=15-88.

Experiments to test Prout's hypothesis.— Thomson, who
gave to Prout's paper the place of honour in his annual
review of the progress of Chemistry for 1815 (Thomson's
Annals of Philosophy, 1816, 7, 17) confirmed Prout's hypo-
thesis by a series of experiments carried out between 1819
and 1825 (An Attempt to Establish the First Principles of
Chemistry by Experiment, London, 1825). But he appears
to have had an unconscious bias in favour of those experi-
ments which gave " correct " or integral values for the atomic
weights. In the words of Berzelius :

" He reduces all the numbers found by his predecessors to
the nearest multiple of the atomic weight of hydrogen,
calculates therefrom the atomic weights of their compounds,
and precipitates them in weighed quantities corresponding
with the corrected atomic weights, when they always

I I 2

484 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

decomposed one another exactly." Berzelius, whose own
numbers had been criticised adversely by Thomson,1 added
that " This investigation belongs to that very small class,
from which science can derive no advantage whatever . . .
and the greatest consideration which contemporaries
can show to the author is to treat this work as if it
had never appeared" (Jahresbericht, 1827, 6, 77).

But Prout's hypothesis was not to be got rid of so easily.
Dumas and Stas (p. 150) by bringing down the atomic
weight of carbon from 12*25 to ii'97 (H=i) obtained
striking evidence that Berzelius's deviations from integral
ratios might be due largely to unsuspected experimental
errors. So, also, Dumas's experiments on the composition
of water (p. 127) which brought down the atomic weight of
oxygen from 16-03 to I5'96, left it still very close to an
integral ratio. Many years later, after completing his experi-
ments on the equivalents of the elements, Dumas concluded
" that the equivalents of the elements are often integral
multiples of the equivalent of hydrogen taken as unity,"
but that in the case of chlorine2 and certain other elements
" the unit with which they must be compared is only 0*5 of
the equivalent of hydrogen " (Ann. Chim. Phys., 1859, 55,
141). Stas, on other hand, who for many years devoted all
his leisure to the elucidation of this problem, began his
experiments with " an almost absolute confidence in the
exactness of the law of Prout," but when they were com-
pleted had " arrived at the complete conviction, the entire
certainty, so far as certainty is possible on such a subject,
that the law of Prout ... is only an illusion, a pure
hypothesis definitely contradicted by experiment " ( Works>

1 "Berzelius's numbers are in general very near approximations to
the truth ; though I am persuaded that in very few instances he has
actually reached it" (First Principles, I. xvii.). For a fuller dis-
cussion see Mallet's "Stas Memorial Lecture" (Trans. Chem. Soc.,
1893, 63, 1-56), and Freund, Chemical Composition, chap. xix.

'2 Following Pelouze, Actes Soc. Helv. Sci. Nat., 1843, 64.

xvm THE CLASSIFICATION OF THE ELEMENTS 485

Prout's hypothesis and the law of probabilities, —

Although it is universally admitted that the atomic weights
do not conform to Prout's hypothesis, and cannot be
calculated or corrected by means of that hypothesis, the
frequent approximation of the atomic weights towards
integral numbers shows clearly that, as Stas admitted in
1887, "there must be something in it" (Mallet, loc. at.
p. 35). Marignac, in reviewing Stas's paper, pointed out
that the average difference between Stas's nine atomic
weights and those required by Prout's hypothesis (as modified
by Pelouze and by himself ) was only 0*056 (Geneva Archives,
1860, 9, 105). He concluded that the law of Prout, like
the laws of Boyle and of Charles, had been proved to be
inexact, but that such laws had still a practical value in
supplying useful approximations for everyday use, and were
theoretically important as affording an ideal standard or rule,
the deviations from which demanded very careful study.

The same idea has been expressed in a mathematical
form by Strutt in a paper " On the Tendency of the Atomic
weights to approximate to Whole Numbers " (Phil. Mag.,
1901, [vi], 1, 311-314). Taking a series of eight atomic
weights given by Richards to three decimal places, he
showed that the sum of the differences from integral
numbers was 0*809, three-fourths of this difference being
due to the two elements chlorine and potassium. The
probability of this total deviation is o-ooii59, or about i
chance in 1000, so that "the atomic weights tend to
approximate to whole numbers far more closely than can
reasonably be accounted for by any accidental coincidence."

Newlands on atomic numbers (1864 to 1878).— Prout's
hypothesis is perhaps responsible for the introduction of the
idea of ATOMIC NUMBERS, i.e. of representing the elements
by a series of integral numbers. The publication by New-
lands, in 1864, of a table showing the elements in the order
of their atomic weights (Periodic Law, p. 7) was followed

486 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

within a month by the addition to the table of a series of
consecutive numbers (p. 457), and by the discovery of the
Law of Octaves as applied to these integers.

In these early atomic numbers no space was left for
undiscovered elements and the correctness of the numbers
could only be checked by the vague indications of the Law
of Octaves. Newlands attempted to overcome this difficulty
by tabulating the ratios of the atomic weights to the atomic
numbers and showing that the ratios increased from 2-5 to
2 7 5 to 3 to 4 approximately (Periodic Law, p. 15). But it
was not until Mendeleeff's prediction of missing elements
had been justified by the discovery of gallium in 1875
that Newlands attempted a more open spacing of the
atomic numbers. Two tables were published in 1878
(Periodic Law, facing p. 32) to illustrate the new system of
atomic numbers. In the first the atomic weights were
divided by 2*3, so that Na= 10 ; the elements were arranged
in sixteen octaves, separate columns being assigned to
hydrogen, to the groups of transition elements, and at the
end of the table to the two elements thorium and uranium,
whilst two octaves were left vacant in the region now
occupied by the rare-earth elements. In this scheme of
atomic numbers

the number in brackets being the atomic numbers, and the
numbers in italics the reducd atomic weights ; the reduced
atomic weights and the atomic numbers usually agreed within
a few units and there were forty vacant spaces. In the
second table the atomic numbers were divided by 2-37, so
that 0=15; hydrogen, and the octaves beginning with
sodium and lithium, were followed by nine columns of ten
elements, the transition elements being placed below the
octaves in the longer columns. In this scheme

= 0-422,

xvin THE CLASSIFICATION OF THE ELEMENTS 487

the agreement between the reduced atomic weights and
the atomic numbers was closer and the number of vacant
spaces was reduced to thirty-seven.

Experimental determination of atomic numbers. —
Atomic numbers have acquired great importance in recent
years, owing to the discovery that several properties of the
atom, which are not directly related to the irregularly-distri-
buted atomic weights, are related in a very simple way to
integral ATOMIC NUMBERS closely analogous with those put
forward by Nevvlands in 1878, but showing seven places less
between hydrogen and gold. Amongst these properties
are

(1) The scattering of a-particles by gases.

(2) The absorption of X-rays by different elements.

(3) The high-frequency spectra of the elements.

Of these methods, the last is the simplest to explain, and
appears to give the most exact results : it may therefore be
described as a type of the new methods of determining
atomic numbers experimentally.

Moseley (1914) on the high-frequency spectra of the
elements. — When X-rays fall upon a crystal, a diffraction-
spectrum is produced, in just the same way as when ordinary
light falls upon a ruled diffraction-grating. The two phe-
nomena are very similar to one another, but in the case of
X-rays the linear dimensions are about 10,000 times smaller.
Thus while the yellow light of sodium consists of two radia-
tions of wave length 5890 and 5896 x io~8 cm., the radiation
from a rhodium target in an X-ray bulb gives radiations of
wave-length 0*54 and o'6i x io~8 cm., the latter in its turn
being a doublet of wave-lengths 0*614 and 0-619 x io~8cm.
(Bragg, Royal Institution Lecture, June 5, 1914). The
nature of the X-ray spectrum depends on the nature of the
elements composing the target (Moseley, Phil. Mag., Dec.
1913, 26, 1024-1034; April, 1914, 27, 703-713).. The

488 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

spectra (Fig. 53) are extremely simple, consisting, in the case
of fourteen elements from aluminium to zinc, of two
principal lines, decreasing in wave-length and increasing in
frequency as the atomic weight increases. The increase of

FIG. 53. HIGH-FREQUENCY SPECTRA.

frequency is perfectly regular, the square-root of the frequency
increasing by equal amounts between consecutive elements ;
thus by selecting a suitable constant the square-roots of the
wave-lengths can be plotted out against a series of integral
numbers (Fig. 54). The exactness of the relationship is

xviii THE CLASSIFICATION OF THE ELEMENTS 489

5 4

WAVE LENGTH X 108Cms.

2 1-5 1 -9 -8

£3

I

benes

^=h

-KVSeries

X*-

10 12 14 16 18 20

SQUARE ROOT OF FREQUENCY X 10'8

22

FIG. 54. -HIGH FREQUENCY SPECTRA OF THE ELEMENTS.

490 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

shown in the following table (Moseley, Phil Mag., Decv
1913, 26, 1028) :

Wave-lengths.

Atomic

Atomic

Element.

Xio-8. Ratio.

Q. Number. Weight.

Ca

3-094

3-368 1-089

19-00

20

40-07

Sc

—

— —

—

21

44-1

Ti

2-524

2-758 1-093

20-99

22

48-I

V

2-297

2-519

•097

2I-96

23

51-0

Cr

2-093

2-30I

•100

22-98

24

52-0

Mn

I-9I8

2-III

•101

23-99

25

54-93

Fe

1765

1-946

•103

24-99

26

55^4

Co

1-629

1798

-104

26 'oo

27

58"97

Ni

1-506

1-662

•104

27-04

28

58-68

Cu

1-402

i '549

•105

28-01

29

63-57

Zn

1-306

i '445

•106

29-OI

30

65-37

In this table Q = j -y— - [ , where v is the frequency of the
I v )

line of longer wave-length and v0 is a constant derived from
the study of ordinary series spectra. It will be seen that Q
gives a series of integral numbers one unit less than the
atomic numbers reckoned from H i, He 2, Li 3, etc., that it
leaves an integer vacant for scandium, and finally that the
Q-numbers show precisely that exact relationship which Prout
postulated incorrectly in the case of the atomic weights. It
is further of interest to notice that the value of Q is one
unit greater for nickel than for cobalt, although nickel has
the smaller atomic weight, thus justifying Mendeleeff's
arrangement of the transition-elements in the order Cr, Mn,
Fe, Co, Ni, Cu, Zn.

This " K series " of spectra is continued up to Ag(47) =
107-88. But already at Zr(4o) = 90'6, an "L series" of
spectra begins ; this is more complex, containing some five
chief lines, but the same laws apply to these lines as to the two
lines of the K series, and the atomic numbers can thus be
carried forward as far as Au(79)= 197*2.

A complete list of atomic numbers and atomic weights is
shown in Table F, p. 491. The atomic numbers deter-
mined experimentally by Moseley in 1913 and 1914 are

xvin THE CLASSIFICATION OF THE ELEMENTS 491

shown in heavy type. It should be noticed that CUiy) and
K(i9), both determined experimentally, leave a gap for
Ar(i8), although argon has a greater atomic weight than
potassium ; corresponding data for tellurium and iodine are
not given. In the case of the rare earths, all the numbers
from 57 to 72 were observed, either with pure prepara-
tions or with mixed fractions, with the exception of 61
and 72, which appear to be entirely wanting in these
materials.

TABLE F.— ATOMIC NUMBERS AND ATOMIC WEIGHTS.

I i Hydrogen

H 1-008

II 2

Helium

He

3 '99

III 10

Neon

Ne

20-2

3

Lithium

Li

6-94

ii

Sodium

Na

23-00

4

Beryllium1

Be

9-1

12

Magnesium

Mg

24-32

5

Boron

B

iro

13

Aluminium

Al

27-I

6

Carbon

C

12-00

14

Silicon

Si

28-3

7

Nitrogen

N

14-01

15

Phosphorus

P

3I-04

8

Oxygen

0

16

16

Sulphur

S

32-07

9

Fluorine

F

19-0

17

Chlorine

Cl

35'46

IV 18

Argon

Ar

39-88

V36

Krypton

Kr

82-92

19

Potassium

K

39-10

37

Rubidium

Rb

85^5

20

Calcium

Ca

40-07

38

Strontium

Sr

87-63

21

Scandium

Sc

44-1

39

Yttrium

Yt

89*0

22

Titanium

Ti

48-!

40

Zirconium

Zr

90-6

23

Vanadium

V

51-0

41

Niobium2

Nb

93 '5

24

Chromium

Cr

52-0

42

Molybdenum

Mo

96-0

25

Manganese

Mn

54 '93

43

—

—

—

26

Iron

Fe

55-84

44

Ruthenium

Ru

101-7

27

Cobalt

Co

58-97

45

Rhodium

Rh

102-9

28

Nickel

Ni

58-68

46

Palladium

Pd

106-7

29

Copper

Cu

63'57

47

Silver

Ag

107-88

30

Zinc

Zn

65-37

48

Cadmium

Cd

112-40

3i

Gallium

Ga

69-9

49

Indium

In

114-8

32

Germanium

Ge

72-5

50

Tin

Sn

119-0

33

Arsenic

As

74-96

51

Antimony

Sb

120-2

34

Selenium

Se

79'2

52

Tellurium

Te

127-5

35

Bromine

Br

79-92

53

Iodine

I

I26-92

1 Or Glucinum Gl.

2 Or Columbium Cb.

492 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.
TABLE F.— ATOMIC NUMBERS AND ATOMIC WEIGHTS — cont.

VI 54 Xenon

Xe 130*2

72 -

55 Caesium

Cs 132-81

73 Tantalum

Ta 181-5

56 Barium

Ba 137-37

74 Tungsten

W 184-0

rf) /57 Lanthanum

La 139*0

75 —

—

H

58 Cerium

Ce 140-25

76 Osmium

Os 190*9

'/>

59 Praseodymium

Pr 140-6

77 Iridium

Ir 193-1

*^2-

60 Neodymium

Nd 144-3

78 Platinum

Pt 195-2

W

61

— —

79 Gold

Au 197-2

J

62 Samarium

Sa 150-4

80 Mercury

Hg 200-6

M

63 Europium

Eu 152-0

8 1 Thallium

Tl 204-0

3

64 Gadolinium

Gd i57'3

82 Lead

Pb 207-10

H

65 Terbium

Tb 159-2

83 Bismuth

Bi 208-0

<5

66 Dysprosium

Ds 162-5

84

— —

w

67 Holmium

Ho 163 -5

85 -

—

w

68 Erbium

Er 1677

VII 86 Niton

Nt 222-4

v

69 Thulium

Tm 168-5

88 Radium

Ra 226-4

<

70 Ytterbium

Yb 172-0

90 Thorium

Th 232-4

• \71 Lutecium

Lu 174-0

92 Uranium

Ur 238-5

Atomic weights and atomic numbers. Total number
of missing elements. — The discovery of a real system of
integral atomic numbers has supplied just that element of
mathematical exactness which was lacking in the relation-
ships put forward by Prout, by Dobereiner, by Dumas, and
others. The agreement between these atomic numbers and
the consecutive numbering of the elements is such that
every confidence may be felt in an enumeration of the
elements on this new basis. Thus between Ba(56) and
Ta(73) there are sixteen places, of which fourteen are filled
by rare-earth elements of known atomic weight ; Mendele'eff
was probably wrong when he assumed nineteen l vacancies in
this interval. Apart from one gap at 61 in the rare earth
series, and one gap at 72 between the rare-earth elements
and tantalum, the new atomic numbers admit only two
vacancies between aluminium and gold, namely, Mendele'eff s
Ekamanganese(43) = 99 (?) between Mo(42) and Ru(44)
Trimanganese (75) = i87(?) between W(74) and 05(76).

1 The discovery of the gases of the helium family would have
increased Mendeleeff's estimate to twenty.

xvin THE CLASSIFICATION OF THE ELEMENTS 493

There are, however, two vacant places, 84 and 85, between
Bi(83) and Nt(86), and three others, 87, 89, 91, between
Nt(86), Ra(88), Thfeo), 11(92), in a region in which the
atomic numbers have not yet been determined by experi-
ment. The total number of vacancies between hydrogen
and uranium is thus increased to nine.1

The existence of two light gases, "coronium" and
" nebulium," between hydrogen and helium has been sug-
gested by Rydberg (see Phil. Mag., July, 1914, 28, 139) on
the following grounds. The two short periods, from Li to
Ne and Na to Ar, contain 2 x 8 = 42 elements. The two
long periods, from K to Kr and Rb to Xe contain 2 x 18 = 62
elements. These should be followed by two very long periods
containing 2 x 32 = 82 elements, of which 27 out of 32 are
known in the range from Cs to Nt, and 3 beyond. On
the other hand, the two short periods should be preceded by
periods containing 2 x 2 = 22 elements, so that helium would
be the fourth 2 element instead of the second.

The existence of a series of integral atomic numbers
proves " 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" (Moseley
Phil. Mag., 1913, 26, 1031). The atomic weights are
determined mainly by the atomic numbers, but there is
evidently some secondary factor at work which may modify
the atomic weight by at least a unit. Thus in the two short
periods the atomic weights are just twice the atomic numbers
in the case of He, C, N, O, Ne, S ; but they are one unit

1 Not counting the hypothetical gases " coronium " and " nebulium"
of atomic weight perhaps less than hydrogen.

2 On these grounds Rydberg suggests that the atomic numbers should
be converted into ORDINALS by adding two units to all the elements
out hydrogen (Phil. Mag., July, 1914, 28, 144). This would not affect
Moseley's data, but measurements of the scattering of X-rays by gases
.ndicate consecutive numbers for hydrogen and helium (Phil. Mag., Oct.
1914, 28, 631); this difficulty could be overcome by placing the two
hypothetical gases before hydrogen instead of after.

494 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

less for hydrogen, and almost exactly a unit greater for Li,
Be, B, F, Na, Al, P. The nature of this secondary factor
is not yet known, but if it exists, there is no reason why
the atomic weights should not occasionally overlap, as in
the case of Ar and K, Co and N, Te and I ; it need
not therefore be assumed that the atomic weights now
given for these elements are incorrect, as Mendeleeff
asserted.

SUMMARY AND SUPPLEMENT.
A. METALS AND NON-METALS.

The idea that all bodies were composed of four "elements"
or three " principles " was opposed by Boyle, who urged that
the number of elements could only be determined by experi-
ment (Sceptical Chymist, 1661). Lavoisier, in 1789, gave a list
of 30 elements ; Berzelius, in 1819, described 50 ; the number
now known exceeds 80.

Lavoisier classified the elements in three chief groups : (a)
Oxygen, with light and heat, azote and hydrogen, (b) Metals,
forming basic oxides, and (c) Non-metals, forming acid oxides.
Davy classified chlorine with oxygen as a " supporter of com-
bustion," and also described the elements as " positive " or
" negative " according as they were attracted to the negative or
to the positive pole in electrolysis. Berzelius classed fluorine
and iodine with chlorine, and sulphur and selenium with oxygen.
Finally the oxygen-group ceased to be distinguished from the
other non-metals, and the elements were divided into two groups,
as metals and non-metals.

B. NUMERICAL RELATIONSHIPS AND THE PERIODIC LAW.

Dobereiner, in 1829, showed that the atomic weight .of
bromine was approximately the mean of the atomic weights
of chlorine and iodine. Other triads were S, Se, Te ; Ca, Sr,
Ba ; Li, Na, K ; also Ni, Cu, Zn and Pt, Ir, Os, where the
atomic weights differed very little. Dumas, in 1859, discovered
regular increments of atomic weights in families of 4 or 5

xvin THE CLASSIFICATION OF THE ELEMENTS 495

elements, and showed that these might be identical in different
families ; thus there was a common difference between the
equivalents in the families F, Cl, Br, I and N, P, As, Sb, as well
as between those of Mg, Ca, Sr, Ba, Pb and O, S, Se, Te, Os.

Newlands, in 1864, using Cannizzaro's atomic weights,
arranged Dobereiner's triads and Dumas's families in a single
table. In 1865, after tabulating the elements in the order of the
new atomic weights, he found that similar elements occurred,
as in music, at intervals of 7 or 14 places. This law of octaves
failed to attract attention until similar schemes were put
forward by Mendele"eff in 1869 and in 1871 and by Lothar
Meyer in 1869 to illustrate the law of periodicity, which states
that " The properties of the elements are periodic functions
of their atomic weight? In Mendeleeff's classification, New-
lands's octaves appear again as 12 series or short periods, the
first containing hydrogen only, whilst the second contains the
7 typical elements, Li, Be, B, C, N, O, F, of Groups I to VII.
A Group VIII was added to take the triads of transition-elements,
which bridge the gap between the even series and the odd series,
4 to 5, 6 to 7, and 10 to n ; each of these triads, with the
octaves on either side, constituted a long period of 7 + 3 + 7 = 17
elements. A " Group o " is now added to take the rare-gases
of the helium series. The main periods, as now recognised,
are : —

Period I. ( = series i) Hydrogen.

Period II and III. ( = series 2 and 3) Eight elements each.

Periods IV and V. ( = series 4 to 7) Eighteen elements each.

Period VI. ( = series 8 to n) Thirty-two elements.

Period VII. (= series 12 etc.) Radioactive elements.

The sixth period is extended to 32 elements by a cluster of
some 15 rare-earth elements, crowded together in the boron-
group. The end of the sixth period and the fragmentary begin-
ning of the seventh period contain 7 elements of known atomic
weight from lead to uranium, but the whole space is crowded
with some 40 transient radio-elements, four of these occupying
well-defined places at the beginning of the seventh period ;
others are clustered together in small swarms of isotopic elements
having similar chemical properties and nearly equal atomic
weights.

496 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

C. ILLUSTRATIONS AND APPLICATIONS OF THE PERIODIC

LAW.

The periodic law was illustrated by L. Meyer (1869) in refer-
ence to the valency and the atomic volumes of the elements ;
the former property justifies the classification into octaves or
short periods, the latter property proves the existence of five
main periods and two fragments. Dewar (1913) has shown
that the atomic heats at low tempero-tures exhibit the same
periodic variations as the atomic volumes. Other periodic
properties are melting-point, compressibility, heat of formation
of oxides and chlorides, etc.

The periodic system has been used to correct certain atomic
weights, in which there was doubt as to the equivalent, as in the
case vigold, or the valency, as in the cases of beryllium, indium
and uranium ; but the atomic weights of argon and potassium,
cobalt and nickel, tellurium and iodine do not conform to the
order suggested by the periodic classification. Mendeleeff also
used the periodic classification to predict the properties of
missing elements, of which three have since been discovered,
thus

Rka-boron Rka-aluminium Eka-silicon

= Scandium, 44*1 = Gallium, 69*9 = Germanium, 72*5

(Nilson, 1879) (Lecoq de Boisbaudran, 1875) (Winkler, 1886)

His predictions in the range covered by the rare-earth elements
were incorrect, but his eka-manganese= 100 and tri-manganese
= 190 are still regarded as vacant places in the periodic classifi-
cation.

D. ATOMIC WEIGHTS AND ATOMIC NUMBERS.

Prout, in 1815-1816, suggested that the atomic weights of the
elements were integral multiples of the atomic weight of hydro-
gen. " Prout's hypothesis " has not been confirmed by experi-
ment, but Strutt has shown that the tendency of the atomic
weights to approximate to whole numbers is greater than can
be accounted for by the law of probability in the ratio
of 1000 : i.

Newlands, in 1864, attached to the elements a series of
consecutive atomic numbers, and thereby discovered the law

xvni THE CLASSIFICATION OF THE ELEMENTS 497

of octaves. In this arrangement hydrogen was i and gold was
49, but in 1878 a more open system was used, in which gold
was 86. Moseley in 1913-1914 showed that integral atomic
numbers could be determined experimentally by studying the
high-frequency spectra emitted by different elements when used
as targets in an X-ray bulb. In this system, if aluminium be
taken as 13, then gold is 79, and there are only four vacant
places between these two elements after filling in all elements
of known atomic weight. The total number of missing ele-
ments between hydrogen and uranium appears to be nine,
namely, numbers 43, 61, 72, 75, 84, 85, 87, 89, and 91. The
atomic number for nickel was found to be one unit greater
than for cobalt, in spite of its smaller atomic weight, thus
justifying MendeleefTs view that nickel should be placed
between cobalt and copper ; in the same way the atomic
numbers for chlorine and for potassium were found to be
separated by two units, leaving a place between them for
argon, the atomic weight of which is greater than that of
potassium. Atomic numbers may also be deduced from the
absorption of X-rays by solid elements and the scattering of
X-rays by gases. These atomic numbers satisfy all the require-
ments of Prout's hypothesis, but the irregularities shown by the
atomic weights are still unexplained.

X K

CHAPTER XIX

BALANCED ACTIONS

Mayow (1674) on the displacement of nitric acid by
oil of vitriol. — Like many other important theories, the idea
of reversible or balanced actions had its origin in the study of
acids, bases, and salts. Mayow (A.C.R. XVII, 161 ; quoted
on p. 22) showed in 1674 that, when oil of vitriol was added
to nitre, the -nitric acid could be distilled out under a heat
no greater than is required for the rectification of the free
acid. This easy distillation was explained by supposing
that the nitric acid had been " expelled from the society of
the [alkali] by the more fixed vitriolic acid."

Similarly, it was noticed that the volatile alkali could be
displaced from any of its salts by the fixed alkalis, the
reason being given " that the acid ... is capable of entering
into closer union with any fixed salt than it is with a volatile
salt."

Baume (1760) on the decomposition of vitriolated tartar
by nitric acid. — Whilst nitric acid is expelled from nitre by
sulphuric acid with the aid of gentle heat, the opposite
change may take place when the mixture is kept cold. This
fact was discovered in 1760 by Baume, who described the
action as follows :

" I took four gros of vitriolated tartar,1 reduced to a fine
powder, I mixed it with three gros of very pure fuming
spirit of nitre ; ... the mixture became pasty, I diluted it
1 Potassium sulphate.

CH. xix BALANCED ACTIONS 499

with a sufficient quantity of water to dissolve the saline
mass, and I set the liquid to crystallise : the salt which I
obtained proved to be very pure nitre, crystallised partly in
needles and partly in small cubic crystals ; I set it to drain
on grey paper . . . ; the nitre thus drained, showed just the
same phenomena as nitre prepared by the combination of
nitric acid with fixed alkali"1 (Mem. Math. Phys'.^ 1774,
VI. 231-236; paper read before the Paris Academy of
Sciences, Dec. 23, 1760).

Thus under one set of conditions, sulphuric acid may
expel nitric acid from its salts, but under other conditions
nitric acid expels sulphuric acid.

Reversible actions. — This reversal of the action may be
shown by writing two equations, thus :

May ow found that:

2KNO3 + H2SO4 -> K2SO4 + 2HNO8

Nitre + SUSdC gave *^~ + nitric acid

Bourne found that :

K2SO4 + 2HNO3 -> 2KNO8 + H2SO4

These two equations may then be combined into a single
equation in which arrows pointing in opposite directions
show the direction taken by the action under different
conditions, thus :

Hot

2KNO3 + H2SO4 = K2SO4 + 2HNO3

Cold

Bergman's tables of affinity (1775). — Mayow's idea that
a " closer union " existed between fixed than between volatile
acids and alkalis was extended to series of acids and bases
in the " tables of affinity " drawn up by Geoffrey in 1718, by
Stahl in 1720, and finally by Bergman in 1775 (A Disserta-
tion on Elective Attractions, tr. T. Beddoes, London, 1785).
In these tables one acid or base was selected and the
different bases or acids were arranged under it in the order
1 Potassium carbonate.

K K 2

500 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

of their decreasing affinity. But Bergman was obliged to
recognise that the order of affinity was not quite constant,
and in particular that different results were often obtained
when the affinities were compared : (i) in the dry way, by
heating the substances together (as Mayow had done) and
discovering which combination remained fixed in the product
and (2) in the moist way, by mixing the substances in solu-
tion (as Baume had done) and finding which base secured
possession of the acid.

Whilst, therefore, there was a fixed order of affinity of
acids for bases and vice versa, this order might be altered
considerably by changing the conditions of the experi-
ment.

Thus, in the case of sulphuric acid (using modern names)
the affinities of the different bases were as follows :

SULPHURIC ACID

In the Moist Way In the Dry Way

Baryta Potash

Potash Soda

Soda Baryta

Lime Lime

Magnesia Magnesia

Ammonia Metallic calces

Alumina Ammonia

Metallic calces Alumina

From this table it is seen that ammonia, added " in the
moist way " to a solution of a metallic salt, precipitates the
calx of the metal and secures possession of the acid with
which it was combined. On the other hand, when a
metallic calx is heated with an ammonium salt " in the dry
way," the volatile alkali is expelled, whilst the acid remains
behind in combination with the calx.

This reversal of the action can again be shown by writing
two equations, thus :

xix BALANCED ACTIONS 501

In the moist way :
2NH3 + PbQ2 + 2H2O -> 2NH4C1 + Pb(OH)2

Ammonia +

lead . • ammonium lead

chloride ' chloride hydroxide

In the dry way :
2NH4C1 + Pb(OH)2 -> 2NH3 + PbCl2 + 2H2O

Ammonium , lead • lead

chloride ->' hydroxide §lves ammonia + ch,oride + water

Combining these equations, we may write :

Moist

2NH3 + PbCl2 + 2H2O ^± 2NH4Cl + Pb(OH)2.

Dry

Berthollet (1799) investigates the laws of chemical
affinity, — Bergman had recognised the influence of tem-
perature on the order of affinity, and had shown that this
order was altered considerably when substances were ignited
together instead of being merely mixed in solution or
in presence of water. To Berthollet belongs the credit
of directing attention to the influence on chemical action
of the quantities of the interacting substances, an influence
which is now described as MASS-ACTION.

In his Researches into the Laws of Chemical Affinity
(read in Cairo in 1799; published 1801 ; tr. M. Farrell,
London, 1804) he criticises Bergman's method in the
following terms :

" He prescribes then, for determining the elective affinity
of two substances, to try if one of them can remove the
other from its combination with a third, and vice versa. He
takes it for granted, that that body which has removed
another from its combination, cannot, in like manner, be
expelled by that other, and that both experiments will
concur to prove that the first has a greater elective affinity
than the second. He adds at the same time, that it may be
necessary to employ six times as much of the decomposing
substance as would be necessary to saturate immediately
the substance with which it tends to combine."

" The doctrine of Bergman is founded entirely on the

502 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

supposition that elective affinity is an invariable force, and
of such a nature that a body which expels another from its
combination, cannot possibly be separated from the same
by the body which it eliminated " (loc. at. pp. 3-4).

" It is my purpose to prove in the following sheets, that
elective affinity, in general, does not act as a determinate
force, by which one body separates completely another from
a combination ; but that, in all the compositions and
decompositions produced by elective affinity, there takes
place a partition of the base, or subject of the combination,
between the two bodies whose actions are opposed ; and
that the proportions of this partition are determined, not
solely by the difference of energy in the affinities, but also
by the difference of the quantities of the bodies ; so that an
excess of quantity of the body whose affinity is the weaker,
compensates for the weakness of affinity " (loc. cit. pp. 4—5).

" I shall prove . . . that, in opposing the body A to the
combination BC, the combination AC can never take place ;
but that the body C will be divided between the bodies A
and B, proportionally to the affinity and quantity of each "
(loc. cit. p. 6).

Berthollet illustrates the law of mass-action. — Several

examples, of which two may be quoted, are given of the

influence of mass in compensating for weakness of
affinity.

" I have kept an equal quantity of potash, and of sulphate
of barytes, in a small quantity of boiling water. The potash
had been prepared by alcohol,1 and contained no carbonic
acid . . . The operation was performed in a retort, and . . .
was continued until the mixture was desiccated : the residue
was washed with alcohol, which dissolved the potash, and
after that with water, which also produced an alkaline
solution, the alkali of which I saturated with acetic acid ;
after which, by evaporation, the solutions yielded crystals
possessing all the characters and qualities of the sulphate of
potash. Whence it appears that the sulphate of barytes was

1 This dissolves caustic potash, but leaves the carbonate, sulphate,
&c., undissolved.

xix BALANCED ACTIONS 503

partially decomposed by the potash, and that the sulphuric
acid was divided between the two bases " (pp. 8-9).

" Equal weights of potash and of carbonate of lime, finely
pulverised, were boiled in a small quantity of water, which,
after being filtered and rendered transparent, effervesced
strongly with acids; and the residue, after evaporation,
having been treated with alcohol, in order to dissolve the
alkali, furnished a substance that had all the qualities or
characters of the carbonate of potash" (pp. 10-11).

Balanced actions. — All the cases quoted by Berthollet
are examples of BALANCED ACTIONS, i.e. reversible actions in
which chemical change proceeds in both directions at the same
time and under the same conditions. Such actions may be
represented by equations in which the opposing actions are
shown by two arrows, but without any signs to suggest that
different conditions are required to produce reversal. The
cases quoted above are shown thus :

Ba(OH)2 + K2S04 ^ 2KOH + BaS<V
Ca(OH)2 + K2CO3 ^ 2KOH + CaCO8.

In the first case, it had been supposed that baryta and
potassium sulphate interacted to form caustic potash and
barium sulphate by an interaction that was non-reversible
and complete. Berthollet showed that the action could be
reversed, and was, in fact, incomplete, the acid being shared
to some extent between the two bases. In the second case
it had been supposed that carbonate of potash was deprived
of all its carbonic acid and rendered completeiy caustic by
the action of lime ; Berthollet proved that the converse
change could be brought about, and that the action was
" balanced " and incomplete. He concludes :

"It is evident, from the preceding experiments, that the
bases which are supposed to form the strongest combinations
with the acids, may be separated from them by others, whose
affinities are supposed to be weaker, and that the acid

1 It is probable that carbonic anhydride was absorbed from the air, and
that the balanced action was really BaCO3 + K2SO4 ^± K2CO3 + BaSO4.

504 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

divides itself between the two bases. It also appears, that
acids may be partially separated from their bases by other
acids, whose affinities were supposed to be weaker; in
which case, the base is divided between the two acids "
(loc. tit. p. 1 1 ).

Mass-action. — Berthollet's method for detecting the
reversed action in the cases quoted above depended on
using a solution of potash which was concentrated by
evaporating finally to dryness. In this way the action of
mass, which was the chief subject of his research, was
made use of to the fullest extent. On this point he writes :

" If but a small quantity of the decomposing substance be
employed, the effect will not be perceptible ; but if, on the
contrary, a large quantity be employed, as for instance, if
I had treated the sulphate of barytes successively with
additional quantities of potash, and removed, by repeated
washing, the disengaged barytes, I should have ultimately
decomposed the sulphate of barytes almost entirely.1 The
greater then the relative quantity of the decomposing
substance, the greater will be the effect produced " (loc. tit.
p. 12).

Effect of precipitation on balanced actions. — In consider-
ing balanced actions in solution, Berthollet recognised that the
action of mass was seriously limited in the case of insoluble
substances, since " If an insoluble substance be opposed to
a combination, it is evident that only a very small quantity
of it can act " (loc. cit. p. 32). This is expressed in modern
terms by saying that the " active mass " of a solid is constant,
no further increase of activity being produced by increas-
ing the quantity of solid beyond the point at which it
suffices to saturate the liquid.

When one acid is being competed for by two bases or
one base by two acids, the result is influenced mainly by
the relative strengths of the competing bases or of the com-
peting acids. But when two acids are present in sufficient
1 In order to do this, carbonate of potash must be used.

xix BALANCED ACTIONS 505

quantity to neutralise the whole of two bases, or conversely,
the competition is less keen, and the course of the action
is determined mainly by the order in which the salts are
precipitated from solution. In their aqueous solutions,
all neutral salts are almost equally strong or equally stable ;
moreover, unlike the neutralisation of an acid by an alkali,
no marked liberation of heat occurs when two soluble salts
are formed by " double decomposition " from two others
(Hess's "Law of Thermoneutrality," Ann. de Chim., 1842,
[iii], 4, 222). The nature of the salts that separate on
evaporating the solution is therefore determined more by
their relative solubilities than by any factor depending on
chemical affinity.

Berthollet's experiments (1804) on the interaction of
neutral salts. — In his Essay on Chemical Statics (1803;
tr. B. Lambert, 1804), Berthollet made a very careful study of
the balanced interactions of pairs of neutral salts. Amongst
other cases he studied the following :

(a) One of the four salts is almost insoluble in water. — In
this case it is precipitated almost completely when the
alternative pair of soluble salts is mixed.

" If sulphate of potash and nitrate of lime are mixed, in
any proportions, the sulphate of lime which is formed
separates by the excess of its insolubility compared with
that of nitrate of potash " {Chemical Statics ', I. 70) :

K2SO4 + Ca(NO8)2 ^± CaSO4 + 2KNOS.

(precipitated)

Barium sulphate, which is even less soluble than calcium
sulphate, is precipitated so completely that it can be filtered
off and weighed in order to estimate either the amount of
barium or the quantity of sulphates present in a solution :

K2SO4 + Ba(NO3)2 -> BaSO4+ 2KNO3.

insoluble

(b) Two salts are almost equally sparingly soluble. — If
these two salts are on opposite sides of the equation, either

506 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

may crystallise out first according to the composition of the
solution.

" A mixture of nitrate of potash and muriate of lime
yields also a result, in which the influence of the proportions
is still more marked^ because the two least soluble salts
which can be formed, the nitrate of potash and the muriate
of potash, differ but little in this property : either of these
salts may also be obtained at the first crystallisation by a
little variation in the proportions of the nitrate of potash
and muriate of lime" (Chemical Statics, I. 70-71) :

2KNO3 + CaCl2 = 2KC1 + Ca(NO3)2.

If potassium nitrate is in excess, the potash will crystallise
out as nitrate, but an excess of calcium chloride will displace
the equilibrium towards the right-hand side of the equation
and cause the potash to separate as chloride.

(c) One salt is least soluble in the cold, but another in the hot,
solution.

" The solubility of salts varies by a difference of temper-
ature ; but this does not follow the same progression in all
of them. In some it acquires a considerable augmentation
by an elevation of heat ; in others it remains nearly the same.
This condition, which determines the separation of salts,
may therefore produce different effects according to the
thermometric state ; hence it happens, that salts whose
solubility is nearly equal at one degree of heat, may, never-
theless, be easily separated, by producing a great change in
the temperature, and, by making the effect of the propor-
tions, and that of the difference of solubility, predominate
alternately."

" Nitrate of potash and muriate of soda furnish us with
a striking example of this effect. Near the freezing-point,
nitrate of potash has much less solubility than muriate of
soda ; but it is considerably increased by heat, and that of
muriate of soda very little ; so that the solubility of the latter,
which was only about half that of the nitrate of potash,
comes to a degree at which it is equal, and finally at the
boiling point becomes nearly eight times less. By boiling
the mixture, therefore, the muriate of soda is made to

xix BALANCED ACTIONS 507

crystallise at a high temperature, and then, by cooling, the
nitrate of potash, is crystallised : the proportion of each salt is
diminished alternately, and by repeated crystallisations they
are both entirely separated" (Chemical Statics, I. 71-72).
Here the action

NaNO3 + KC1^KNO3 + NaCl

can be brought to completion in the direction of the upper
arrow by crystallising out, alternately, nitrate of potash at
the freezing-point and common salt at the boiling-point. If
the common salt were allowed to accumulate, it would, by
the law of mass-action, check the formation of potassium
nitrate from sodium nitrate and potassium chloride, with
the result that one of these salts (potassium chloride)
might crystallise out and carry the action backwards in
the direction of the lower arrow.

An action of this kind actually occurs when a solution
containing sulphate of potash and nitrate of soda is
evaporated. The sulphate of potash, which is the least
soluble of the four salts,

2NaNO3 + K2SO4^±2KNO3 + Na2SO4,
crystallises first.

" But when the proportion of the first shall be diminished
by the crystallisation, nitrate of potash will also be obtained,
because the water remaining at this period is incapable of
holding in solution the quantity of this salt which may be
formed, and because the sulphate of potash, on its part, is
rendered more soluble by the action of the other salt : this
result might have been obtained from the commencement
of the crystallisation by augmenting the proportion of nitrate
of soda" (Chemical Statics, I. 70).

Action of acids on insoluble salts. — When a strong acid
acts on an insoluble salt of a weak acid, the insolubility of
the salt greatly assists the weaker acid in its effort to retain
possession of the base.

"In fact, if oxalic acid is added to a salt with base of

508 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

lime, a precipitate of oxalate of lime is obtained, much less
abundant than if a solution of a neutral oxalate had been
used, because the action of the acid allows only part of the
oxalate of lime to be formed, but with an oxalate this
obstacle would not have existed" (Chemical Statics, I. 67).
Here the action

H2C2O4 + CaCl2^±CaC2O4 + 2HC1

(insoluble)

is reversible, since the insolubility of its calcium salt enables
the oxalic acid to rob the stronger muriatic acid of part of
its lime.

Interesting cases of the same sort are found amongst the
sulphides, e.g. :

CuS04 + H2S -> CuS + H2S04

(precipitate)

ZnSO4 + H2S <- ZnS + H2SO4.

(dissolves)

In the former case the insolubility of the sulphide causes it
to be precipitated, in opposition to the tendency, which is
dominant in the latter case, for the stronger acid to secure
possession of the base.

Influence of vaporisation on balanced actions. — In
reference to the influence of gases on balanced actions,
Berthollet writes :

" When a substance assumes the state of gas, on separating
from an intimate combination, it becomes elastic, and can
oppose no further resistance to the decomposing action :
whence it appears that substances of this nature do not
act by their mass. The decomposing substance can then
effect a complete decomposition ; and it will suffice to
employ just as much of it as would have been necessary to
form the same combination immediately, or at least a very
trifling excess."

" Thus carbonic acid may be disengaged from its combina-
tion by another substance, whose affinity for the base of the
carbonate might be less ; because that other substance can
act by its mass, and can therefore overcome the affinity of
the carbonic acid, by acting successively : but to expel the

xix BALANCED ACTIONS 509

entire of the carbonic acid, the decomposing substance must
be used in quantity somewhat greater than that necessary to
produce saturation."

" If concentrated sulphuric acid be poured on desiccated
muriate of soda, the affinity of the muriatic acid is diminished ;
and that acid assumes the gaseous state in consequence,
and acts no longer by its mass : but if an aqueous solution
of muriate of soda be employed, or a diluted acid, whether
the sulphuric, or any other, then the muriatic acid may be
retained in the water ; in which case it can act by its mass "
(Chemical Affinity, pp. 46-48).

Berthollet shows that a weak fixed acid may displace
a stronger volatile acid. — " It is to this effect of elasticity
that the decompositions produced by the most fixed acids
... are to be attributed . . . ; it is thus that sulphuric
acid, by means of heat, decomposes the muriates and the
nitrates with a fixed base."

" I distilled a mixture of oxalic acid and muriate of soda,
and the liquor which passed over contained much muriatic
acid" (Chemical Statics, I. 192-193).

In the latter case the weak oxalic acid displaced the
strong muriatic acid because of its greater fixedness :

H2C204+ 2NaCl H> Na2CA+ 2HC1.

(weak (strong

fixed acid) volatile acid)

Again, phosphoric acid is a very weak acid, since nearly
all its salts are decomposed by acetic acid even though
insoluble in water ; but sulphuric acid, which " expels
entirely the muriatic and nitric acids from their combina-
tions, with the assistance of a sufficient degree of heat . . .
is itself expelled from its combinations by phosphoric acid,
independently of affinities" (Chemical Affinity, p. 54).

"Therefore when, by the aid of heat, one body has
separated another from its combination, it must not be
inferred that its affinity is greater at an ordinary tempera-
ture" (Chemical Affinity, p. 55).

In fact, if mere fixedness were the real test of affinity,
ordinary sand must be regarded as one of the strongest

5io HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

acids, since in the manufacture of glass it expels the acid
from sodium sulphate and converts it into a glassy silicate.

Balanced actions in gases. — Whilst a balanced action in
a liquid is often disturbed by the vaporisation of one of the
products, this action may be checked by increasing the
pressure (and thereby the "active mass") of the gas
(Chemical Statics, I. 188).

An interesting example of a balanced action in which
the gas-concentrations are of dominant importance is shown
by the 'equation

Fe304 + 4H2 — 3Fe + 4H20,

The solid iron and its solid oxide have a constant activity,
but the action can be influenced strongly by varying the
proportions of hydrogen and steam. The strong tendency
for the hydrogen to reduce the oxide is shown by the upper
arrow; but when, as in Lavoisier's experiment (Ch. VII,
p. 1 1 8), large quantities of steam are passed over the solid the
slight back-action shown by the lower broken arrow becomes
important, and condensation of the steam immediately
reveals the small quantities of hydrogen that have been
produced.

SUMMARY AND SUPPLEMENT.

Mayow, in 1674, showed that nitric acid could be expelled
from nitre by oil of vitriol and suggested that this was due to
the fact that oil of vitriol was " more concordant " with the
alkali than nitric acid.

Baume, in 1760, showed that this action could be reversed
and that nitre could be precipitated by adding nitric acid to
vitriolated tartar (potassium sulphate). The reversal of the
action is shown in the equation

Cold

Bergman (1775), following Geoffroy (1718) and Stahl (1720),;
drew up tables of " Elective Attractions," showing the order of

xix BALANCED ACTIONS 511

affinity of a series of different acids for a given base and
conversely. He distinguished between the relative attractions
"in the moist way," i.e. in solution and "in the dry way," i.e.
by ignition. Thus, we have :

In the moist way :

2NH3 + PbCl2 + 2H2O -> 2NH4Cl + Pb(OH)2.
In the dry way :

2NH4Cl + Pb(OH)2 -> 2NH3 + PbCl2 + 2H2O.

Berthollet, in his "Researches into the Laws of. Chemical
Affinity" (1801), and in his " Chemical Statics" (1803), recognised
the influence of " mass-action," and suggested that " Every
substance which has a tendency to enter into combination,
acts in the ratio of its affinity and of its quantity." A weak
acid in large quantities may therefore compete with a stronger
acid in smaller quantities. The action of the weak acid " in the
wet way " is greatly helped if it forms an insoluble salt, e.g. :

H2C2O4 + 2CaCl2 ^ CaC2O4+.2HCl.

(insoluble)

or "in the dry way" if it is less volatile than the stronger
acid, e.g. :

H2C2O4 + 2NaCl ^± Na2C2O4+2HCl.

(volatile)

The influence of mass is seen best, however, in the balanced
actions between pairs of neutral salts : these differ so little in
stability that the interaction is controlled almost entirely by
mass-action and solubility. The following cases may be noticed :

(a) K2S04 + Ca(N03)2 ^± CaS04 + 2KNO3,

CaSO4 is almost insoluble and is always precipitated first.

(£) K2S04+2NaNO3 :^± Na2SO4 + 2KN03

K2SO4 separates first on evaporation, but as NaNO3 accumu-
lates in the solution the action passes from left to right and
KNO3 crystallises.

(c) 2KCl + Ca(NO3)2 ^± 2KN03 + CaCl2

KClor KNO3 may crystallise according to the proportions in
which the salts are mixed.

(d) NaNO3+KCl ^ KN03 + NaCl.

512 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

At o° KNO3 is least soluble, at 100° NaCl ; the action may
be driven across completely from L to R by crystallising out
these salts alternately at o° and at 100°.

In quantitative analysis these balanced actions are used to
secure complete precipitation of Ag or Cl as AgCl, Ba or SO4
as BaSO4, &c.

In qualitative analysts the balance of actions between
neutral salts may be used to secure a broad separation of those
metals which form

(1) Insoluble chlorides: Ag', Pb", Hg'

(2) Insoluble sulphides :

(a] Cu", Pb", Hg", Bi" (*) [Fe'"], [Cr'"], [Al'"]
(£) Sn", Sn"", Sb'", As'" (d) Zn", Mn", Ni", Co".

(3) Insoluble carbonates : (a) Ca", Sr", Ba" (V) Mg'

(4) Insoluble platinichlorides : K', NH4'.

In class (i), since hydrochloric acid is one of the strongest
acids, it may be used instead of a neutral chloride ; thus in the
action

the nitric acid formed as a second product is not strong enough
to redissolve the insoluble chloride or to prevent its precipitation.
In class (2), however, the use of sulphuretted hydrogen in place
of a neutral sulphide tends to liberate a strong acid ; this
prevents the separation of the insoluble sulphides of the metals
(2f) and (2rtT), but those of the metals (20) and (zb) are
precipitated even in presence of much acid, thus :

CuSO4+H2S -> CuS + H2SO4

(precipitated)

ZnSO4 + H2S <-ZnS + H2SO4.

(dissolves)

The further separation of the metals (20) and (ib] depends
on the fact that the latter dissolve in alkaline sulphides to form
soluble sulpho-salts, such as Na3AsS4, Na3SbS4, &c. The
separation of the metals (2?) and (2(f) depends on the fact that
the tervalent hydroxides, Fe(OH)3, Cr(OH)3, A1(OH)3, are ex-
tremely weak bases, which do not form carbonates or sulphides,
and are therefore precipitated in the free state when soluble
carbonates or sulphides are added ; in qualitative analysis they
are separated by the addition of a mixture of ammonia and

xix BALANCED ACTIONS 513

ammonium chloride which is not strong enough to precipitate
the divalent hydroxides of the metals (id), thus :

A1C13 + 3NH4OH -> Al(pH)s

(precipitated)

but ZnCl2 + 2NH4OH <- Zn(OH)2 + 2NH4Cl.

(dissolves)

In class (3) a similar method is used to separate magnesium
from calcium, strontium, and barium. Like chalk and barium
carbonate, magnesium carbonate is insoluble in water ; it is,
however, soluble in presence of ammonium salts, thus :

2Na2CO3 -* MgC03 + 2NaCl

(precipitated)1

but MgCl2 + 2(NH4)2CO3<- MgC03+2NH4Cl.

(dissolves)

In the former case, the tendency of the strong acid (HC1) to
combine with the stronger base (NaOH) drives the action in the
direction L to R, already indicated by the insolubility of
magnesium carbonate. In the latter case the magnesia is the
stronger base and the action from R to L is assisted further (i)
by the mass-action of an excess of ammonium chloride and (ii)
by the tendency of this substance to unite with magnesium salts
to form compounds, such as MgCl2, NH4C1,6H2O, and thereby
to weaken the mass-action of the magnesium salt.

1 Usually as a basic salt.

L L

CHAPTER XX

DISSOCIATION
A. DISSOCIATION AND ASSOCIATION

St. Claire Deville (1857) on dissociation. — The name
DISSOCIATION was introduced in 1 85 7 by H. St. Claire Deville l
(Comptes rendus, 1857, 45, 857) to describe the "spontane-
ous decomposition " of substances by heat, without the
intervention of chemical agencies. This definition (which
would include thermal decompositions such as the " coking "
of coal and the "charring" of sugar) was afterwards limited
(Comptes rendus, 1863, 56, 730) so as to include only
cases " in which the decomposition takes place partially,
and at a temperature inferior to that which corresponds to
the absolute destruction of the compound," i.e. to cases of
REVERSIBLE DECOMPOSITION, in which theproducts recombine
when the decomposing forces are removed. These reversible
decompositions are very difficult to detect because, after
heating and cooling, the original substance reappears un-
altered, bearing no trace of the changes which it has under-
gone.

1 St. Claire Deville's experiments were summarised in his Lessons on
Dissociation delivered to the Paris Academy of Sciences in 1864.
These were published as a pamphlet, which includes (almost verbatim)
the Comptes rendus papers of 1864 and 1865, together with the
diagrams, which were not reproduced in the Comptes rendus.

CH. xx DISSOCIATION 515

Grove's experiments on the dissociation of water

(1847). — This difficulty was first overcome by Grove, who
showed in 1847 (Bakerian Lecture, Phil. Trans., 1847, 137,
1-16) that the combination of hydrogen and oxygen to form
water was partially reversed at high temperatures, e.g. by heat-
ing a platinum wire in steam by an electric current, by pass-
ing sparks through steam between platinum points, and by
plunging the fused end of a platinum wire under cold water.
Under these conditions bubbles of gas were produced,
which could be collected and detonated.

In these experiments a small quantity of steam is decom-
posed into its elements as shown by the equation

2H2O -> 2H2 + O2.

As recombination is checked by dilution with a large
volume of steam, the mixed gases can be cooled to 100°
without recombining ; below this temperature the mixture
is quite stable, even when the steam is condensed, but
detonates in the ordinary way when ignited. The maximum
amount of gas collected was about 7^nr of the volume of
the steam.

The dissociation of water may be represented by the
scheme

in which the main course of the action is represented by a
full arrow, and the slight reversal by a dotted arrow.

Deville's experiments on the dissociation of water (1863)
and of carbonic anhydride. — Grove's experiment was
repeated on a larger scale in 1863 by Deville (Comptes
rendus, 1863, 56, 322), who poured i to 2 kilogrammes of
fused platinum into water and so secured an abundant
liberation of explosive gas.

Deville also proved the dissociation of water by passing a
stream of carbonic anhydride, saturated with water-vapour
at 90° to 95° C, through a porcelain tube packed with pieces

L L 2

516 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

of broken porcelain and heated to 1300° C. On absorbing
the carbonic anhydride by potash a continuous supply of
detonating gas was obtained; at the end of two hours, 25
to 30 cubic centimetres had been collected (loc. tit. p. 323).
Dissociation of carbonic anhydride. — When dry carbonic
anhydride was passed through the same apparatus, an
explosive mixture of carbonic oxide and oxygen was
collected at the rate of 20 to 30 cubic centimetres per hour.
The combustion of carbonic oxide is therefore a reversible
process as shown by the dotted arrow in the equation :

Deville's hot-cold tube (1864).— In most cases of
dissociation the extent of the decomposition increases very
rapidly with the temperature. If a substance be strongly
heated and then suddenly chilled it is sometimes possible to
fix the products of dissociation in the larger proportions in
which they exist at higher temperatures ; on the other
hand, if the cooling is gradual they recombine and are only
retained in the small proportions which are natural at the
lower temperature at which recombination ceases. Deville
secured this strong heating and rapid cooling in his "hot-
cold " tube shown in Fig. 55 (Lessons on Dissociation, p. 64).
The outer tube of porcelain was heated to whiteness ; the
inner tube of brass was cooled by a current of water. With
this apparatus he demonstrated the decomposition of car-
bonic oxide, of sulphurous anhydride, and of hydrogen
chloride :

2CO ^ C +CO.;

3so2 ^± s + 2s63

In the first case, carbon was deposited on the cold tube
and carbonic anhydride carried forward (Comptes rendus,
1864, 59, 875); in the second case, sulphide and sulphuric
anhydride were both deposited ; in the third case, chlorides

xx DISSOCIATION 517

of silver and mercury were formed on a cold tube of
amalgamated silver and several cubic centimetres of
hydrogen were carried forward (Comptes rendus, 1865, 60,

3177325)-

Similar changes were produced by sparking. Sulphurous
anhydride was condensed completely, SO2 -> S + 2SO3, by
sparking over sulphuric acid, which absorbed the sulphuric
anhydride as fast as it was formed. A mixture of sulphurous
anhydride with oxygen was absorbed still more readily :

-> 2SO3.

FIG. 55.— DEVII.LE'S "HOT-COLD" TUBE.

PP is a hot tube of porcelain. LL is a brass tube cooled with water. The gas to be
decomposed passes in at T and out at T'.

Hydrogen chloride, which had been shown by Davy fifty
years previously (Ch. XI, p. 224) to give calomel and one-
half its volume of hydrogen when sparked over mercruy,
was brought into the category of dissociable substances by
Deville, who confirmed Davy's earlier observation.

Dissociation of ammonia. — In the course of these experi-
ments Deville discovered that the decomposition of ammonia
into nitrogen and hydrogen by sparking as used by Berthollet
to establish its composition (p. 262) was a reversible change.
By using an acid to absorb the ammonia as fast as it was

5i8 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

formed, Deville was able to recombine the gases completely.
Deville wrote :

" If ammonia gas is sparked during several hours, until
its volume appears to be exactly doubled, no sensible
absorption is observed on introducing drops of water 1
into the eudiometer : the decomposition appears to be
complete. But if, instead, one passes in some bubbles
of gaseous hydrochloric acid, a slight cloud obviously
obscures the mixture of nitrogen and hydrogen into which
the ammonia has been transformed. This transformation
is then not absolute. This affords an easy explanation of
the following experiment. After having decomposed one
volume of ammonia as perfectly as possible by sparking,
giving two volumes of a mixture of nitrogen and hydrogen,
one volume of gaseous hydrochloric acid is introduced into
the eudiometer and the sparking is renewed during 8 to 10
hours. At the end of this time the upper part of the
apparatus is covered with a layer of sal-ammoniac and the
mercury has risen to the platinum wires" (Comptes rendus,
1865, 60, 324).

The decomposition and recomposition of ammonia may
be represented by the equations :

2NH3 — N2 + 3H2. N2 + sH2 + 2HC1 -> 2NH4C1.

The recombination is now usually demonstrated by
sparking over sulphuric acid, a method already used by
Deville in combining sulphur dioxide with oxygen.

Bineau (1838) investigates the vapour-densities of ammo-
nium and phosphonium salts. — During the half-century which
followed the enunciation of Avogadro's hypothesis, great
difficulty was experienced in accepting his view that equal
numbers of molecules, under similar physical conditions, always
occupied equal volumes. One difficulty arose from the fact
that one " equivalent "occupied a different volume in different
gases. Thus if i volume of oxygen was required to convert
a given weight of potassium into its oxide, the quantities of

1 In Lessons on Dissociation, " concentrated sulphuric acid " is used.

xx DISSOCIATION 519

gas required to convert the same weight of potassium into
chloride were

2 volumes of chlorine gas,
4 volumes of hydrogen chloride gas,
and 8 volumes of sal-ammoniac.

The first three observations can be explained by
Avogadro's hypothesis as expressed by the equations :
4K + O, (i vol.) =2K2O

4K + 2C1, (2 vols.) =4KC1
4K + 4HC1 (4 vols.) = 4KC1 + 2H2.

But the case of sal-ammoniac is quite anomalous, since
Avogadro's hypothesis would indicate that only four volumes
of the gas were required, as shown by the equation :
4K + 4NH4C1 (4 vols.) - 4KC1 + 4NH3 + 2H2.
A similar anomaly was observed in the vapour-densities of
the ammonium and phosphonium salts examined by Bineau
in 1838 (Ann. Chim. P/iys., 1838, 68, 416-441).
The following compounds were studied :

PH3 + HI = PH4I (phosphonium iodide)
PH3 + HBr = PH4Br (phosphonium bromide)
2NH3 + CO.j =(NH3)2CO2 (ammonium carbamate)1
NH3 + SH2 = NH4HS (ammonium hydrosulphide)
NH3 + HCN =NH4CN (ammonium cyanide)
NH8 + HC1 - NH4C1 (ammonium chloride)

In each case an easily crystallised compound was formed,
but a study of the vapour showed that no contraction in
volume had taken place. Thus Bineau found that
phosphonium iodide "is formed from equal volumes of
phosphoretted hydrogen and hydriodic acid united without
condensation, and that its density, in the state of vapour, is
the mean of the densities of its components " (foe. cit.
p. 430.)

1 The product "sal-volatile" is not a true carbonate, but the
ammonium salt of carbamic acid, NH2*CO*OH ; compare p. 430.

520 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

Again ammonium carbamate \loc. tit. p. 434) "when
heated, gives a volume of gas equal to the sum of the
volumes of the components of the salt." "The vapour of
the salt contains one-third of its volume of carbonic gas
and two-thirds of its volume of ammonia gas, The conden-
sation is consequently nil, and the density of the vapour is

o

the densities of the components (relatively to air) being
1*524 for carbonic anhydride and 0*591 for ammonia.

The analogous case of nitric oxide, which is formed from
its components without condensation, had been discussed by
Dalton in 1802, but was explained by Avogadro (Chap.
XV, P- 338) on tne assumption that the molecules of
nitrogen and oxygen were divisible into halves, the " atom "
being therefore the half and not the whole molecule :

N2 + O2=-2NO.

But if a similar method were applied to Bineau's observa-
tions it would be necessary to halve again the atomic weight
of nitrogen in ammonium chloride, ammonium cyanide, and
ammonium hydrosulphide, whilst in the case of the
carbamate it must be divided into 3 parts. This further
division into 2 and into 3 parts would reduce the atomic weight

to one-sixth of Avogadro's figure, namely,— = 2 '3, and the

6

formula of nitrogen gas would be N12. The carbon atom
would be divided into two parts in the cyanide and three
parts in the carbamate, thus reducing its atomic weight to

T *2

— = 2. It would also be necessary to divide the atomic

weight of oxygen by 3 and to reduce those of chlorine,
bromine, iodine and sulphur to one-half of the values
generally accepted for these elements.

Such alterations were far too drastic to be justified by the
behaviour of a single group of substances, and therefore

xx DISSOCIATION 521

served merely to throw discredit upon Avogadro's
hypothesis.

Pebal's proof of the dissociation of sal-ammoniac
(1862). — Deville's discovery of dissociation suggested a
simple explanation of these abnormal vapour-densities. Kopp,
following Cannizzaro (Nuovo Cimento, 1857, 6, 428),
pointed out (Liebig's Annakn, 1858, 105, 390) that the
vapour obtained by heating a dissociable substance might

FIG. 56. —PEBAL'S AI-PARATUS FOR PROVING THE DISSOCIATION OF
SAL-AMMONIAC.

be merely a mixture of the decomposition-products, occu-
pying a larger volume and having a much lower density
than the undecomposed vapour. This explanation had
been suggested tentatively by Bineau in the case of am-
monium carbamate, but was first verified in the case of sal-
ammoniac, by the experiments of Pebal (Liebig's Annakn,
1862, 123, 199-202).

522 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

Crystals of sal-ammoniac were heated in a glass tube (Fig.
56), 13 mm. wide, over a plug of asbestos 20 mm. in length ;
a current of hydrogen was maintained on each side of the
asbestos. After the sal-ammoniac had vaporised, the hydro-
gen A from that side of the plug was found to contain
free hydrochloric acid and turned blue litmus-paper red,
whilst the hydrogen B from the other side of the plug con-
tained free ammonia and turned red litmus-paper blue.
This separation of ammonia and hydrogen chloride depends
on the fact that the ammonia, being a lighter gas than
hydrogen chloride, diffuses more quickly through the
asbestos, rendering the gas on the other side alkaline and
leaving behind a gas containing an excess of acid. Such
a separation could not take place, however, if the vapour
consisted entirely of undecomposed sal-ammoniac, since
this would then diffuse as a whole and would not be resolved
into its constituents.

It should be noted that, as ammonia and hydrogen
chloride recombine immediately in the cold, the products
of dissociation cannot be fixed by the method of sudden
chilling, as used by Grove and by Deville.

Variation of vapour-density with temperature dis-
covered by Cahours (1844). — Another anomaly which
found a simple explanation in Deville's theory of dissociation
is the variation of vapour-density with temperature, dis-
covered by Cahours in the cases of acetic acid (Comptes
rendus, 1844, 19, 771-773) and phosphorus pentachloride
(Ann. Chim. Phys., 1847, \.Z\ 20, 369-378).

For acetic acid, Cahours found the following values :

Temperature... 145° 152° 219° 231° C.

Density 275 272 2-17 2*12 (relative to air)

By diluting the vapour with hydrogen Playfair and Wanklyn
(Journ. Chem. Soc., 1862, 15, 153) were able to measure
its density below the normal boiling-point (119°) and found

xx DISSOCIATION 523

that the value rose to 4/0 at 60° C. The values required by
Avogadro's hypothesis are: for C.2H4O2 2*073, for C4H8O4
4-146. The density at 60° C. approximates to that of an acid
of the formula C4H8O4; at 23o°C. it approximates to the
value for an acid C2H4O2, and is not changed by further
heating. Playfair and Wanklyn concluded (loc. tit. p. 143)
that "in the few cases where any great difference is found
between a vapour-density taken at a point near the point
of liquefaction, and one taken at a point far removed from
the point of liquefaction, there is a chemical and not a
merely physical change." The vapour " has sometimes one
and sometimes another chemical formula, according as it
is more or less heated." In the case of acetic acid the
dissociation is expressed by the scheme :

High temp.

C4H804 - 2C2H402,

Low temp.

the action proceeding from left to right with increasing
temperature and conversely.

The dissociation of phosphorus pentachloride. — For
phosphorus pentachloride vapour, Cahours found (Ann.
Chim. Phys., 1847, [3], 20, 373) :

Temp 182° 200° 250° 300° 336° C.

Density 5-078 4-851 3-991 3-654 3-656

By diluting with air, Wurtz was able to measure the
density of the vapour at 129° C. under a partial pressure of
175 mm., and found that the value increased to 6-4 (Comptes
rendus, 1873, 76, 603), thus approaching the value 7*217
required by the formula PC15. The lower values observed
at temperatures above 129° C. were obviously due to dissocia-
tion, as shown by the equation :

PC15 ^ PC13 + C12,
the dissociation being substantially complete above 30o°C»

524 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

If the low vapour-densities were actually due to dissocia-
tion, they should be raised by adding one of the products
since this would tend to accelerate their recombination.
Wurtz found (foe. at. p. 608) that when the pentachloride
was vaporised at 160 to iy60C. into an atmosphere of
phosphorus trichloride, it had the vapour density 7*226,
agreeing closely with the value of 7*217 calculated for the
formula PC15.

The dissociation of sulphur vapour. — The decrease of
vapour density with rising temperature is shown in a still
more marked degree by sulphur, the vapour-density of which
falls from 6*57 at 524°C. (Dumas, Ann. Chim. Phys., 1832,
50, 175) to 2*23 at 86o°C. and io4o°C. (Deville and Troost,
Ann. Chim. P/iys., 1860, 58, 287 and 298). The latter value
corresponds with the formula S2 ; the former approximates
to that required by the formula S6 ; later measurements at
lower temperatures and reduced pressures have given values
approximating to that required for the formula Sg.1 It is
probable that the vapour contains molecules of two types,
as shown by the equation :

but intermediate forms such as S6 and S4 are by no means
excluded.

The dissociation of nitrogen peroxide discovered by
Playfair and Wanklyn (1862).— Playfair and Wanklyn
discovered in 1862 (loc, cit. p. 156) that the vapour-density
of nitrogen peroxide (diluted with nitrogen) decreases with
rising temperature as follows :

Temp ............. 4*2° 11-3° 24-5° 97'5°C.

Density ........ 2-588 2*645 2'52° I783

1 Bleier and Kohn (Ber., 1900, 33, 51) found S7.8s at 193° C. and
2*1 mm. pressure. Biltz (Ber., 1901, 34, 2493) obtained an average
value 87-96 in ten readings at the normal boiling-point, under pressures
ranging from 200 to 540 mm.

xx DISSOCIATION 525

Deville and Troost found (Comptes rendus, 1867, 64,
237-243) for the pure vapour :

Temp. ... 267° 49-6° 60-2° 8o'6° ioo'i° 121-5° I35° l$4° C"
Density... 2*65 2-27 2*08 i'8o i'68 1-62 i'6o 1-58

The values calculated by Playfair and Wanklyn were
3*179 for N2O4, 1*589 for NO2. The percentage com-
position of the vapour is therefore :

At 267° C ............................ 80% 20%

At 49-6° .............................. 60 40

At 100-1° ............................. ii 89

the dissociation being complete from 1 54° C. upwards.

Playfair and Wanklyn concluded "that peroxide of
nitrogen exists in two states ; that there are two bodies having
the same percentage composition as peroxide of nitrogen,
but which are polymeric." The change of molecular state
was associated by them with " the wonderful deepening of
colour which peroxide of nitrogen undergoes on being
heated," and which had been noted by Priestley ninety years
before (p. 76).

The two changes are undoubtedly due to a common
cause, namely, the dissociation of a dense colourless
tetroxide, N2O4, into a deeply-coloured dioxide, NO2, of
lower density. The liquid, boiling at 22° C., which Gay-
Lussac prepared in 1816 by heating lead nitrate and cooling
the vapours by means of ice and salt (Ann. Chim. Phys.^
1816, 1, 405), is pale yellow; it probably consists of N2O4
with only a trace of NO2. The solid, melting at— 10° C.,
is a colourless ice (Deville and Troost, Comptes rendus^ 1867,
64, 238, footnote), and must be regarded as the pure
tetroxide, free from all traces of dioxide.

The dissociation of nitrogen peroxide,

N204— 2N02,

cannot be proved by chilling the brown dioxide, which loses

526 HISTORICAL INTRODUCTION TO CHEMISTRY CH/

its colour immediately when cooled ; nor can Pebal's method
of diffusion be applied to a substance which gives only one
product of decomposition. But the dissociation, indicated by
the changing vapour-densities and made visible by changes
of colour, is revealed in a striking way by the large
absorption of heat which accompanies it. To raise the
temperature of 92 grams of the gas through i° C. requires
(Berthelot and Ogier, Ann. Chim. Phys., 1883 [v], 30, 392)

Between 27 and 67° C 747 calories

67 ,, 103° 57-0 „

„- 103 „ 150° 27-0 „

,, 150 ,, 198° 9-1 „

As the heat-capacity of a gas usually increases with rising
temperature, the remarkable absorption of heat in the lower
ranges must be due to a chemical change : a direct com-
parison has shown that the heat-capacity is greatest just
when the degree of dissociation (as calculated from the
vapour-densities) is increasing most rapidly and that the two
changes run parallel throughout ; thus the percentage of
dissociation occurring over three consecutive ranges of tem-
perature were calculated to be as follows (Berthelot and
Ogier, loc. at. p. 398) :

From vapour density. From heat capacity.

27-70° C. 35 -6% 27-67° C. 40-0%

70-100° 23-6% 67-103° 26-3%

100-136° 9-5% 103-150° 13-2%

Deville (1867) does not accept the dissociation of sal-
ammoniac, acetic acid, phosphorus pentachloride, and
nitrogen peroxide. — It is remarkable that Deville found
himself unable to accept the view that sal-ammoniac vapour
contained free ammonia and free hydrogen chloride : these,
he held, were formed from the undecomposed vapour by
the process of diffusion. Further, he did not consider that
dissociation was responsible for the change of vapour-density
with temperature, and concluded that :

pC DISSOCIATION 527

" For all vapours condensible above zero, what is called
the vapour-density is a mathematical fiction, to which it is
impossible to give a physical meaning. Its interpretation
is only possible when the vapour, at a point sufficiently
above the boiling-point, obeys the law of Mariotte, and
possesses the constant coefficient of expansion, 0*00367, of
the perfect gases" (Comptes rendus, 1867, 64, 242).

The association of water. — In the case of water-vapour,
the vapour-density agrees almost down to the boiling-point
with that calculated for the formula H2O. But Bose (Zeit.
Elektrochem., 1908, 14, 271) has recently found evidence of
polymerisation in the neighbourhood of 100°, the increase of
density corresponding with the formation of about 9%
of H4O2. Liquid water at 100° appears to be polymerised
to such an extent that the average size of the molecules is
approximately that shown by the formula H4O2 (Guye,
Trans, Faraday Soc., 1910, 6, 78). The formation of
unstable complex molecules during the condensation of a
vapour is usually described as ASSOCIATION. The phenome-
non does not differ essentially from those cases of dissocia-
tion (S8 ^ 4S2, N2O4 ^ 2NO2), in which only a single
product is formed. The term is, however, usually applied
to cases in which the vapour has a normal density and the
polymerisation is, for the most part, confined to the liquid
and solid states.

The evidence of association in liquid water is particularly
strong. Rontgen pointed out in 1891 that the anomalous
contraction between o° and 4° might be explained by a
gradual dissociation of " ice-molecules " into denser " water-
molecules." Sutherland suggested that these might be
formulated as follows :

" steam-molecules " H2O = " hydrol "
" water-molecules " H4O2 - " dihydrol "
"ice-molecules" H6O3 = « trihydrol."

Following the analogy of nitrogen peroxide, it may be

528 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

supposed that steam above (say) 150° C. consists entirely
of hydrol, but that small quantities of dihydrol are formed
as the boiling-point is approached. Liquid water is obviously
a mixture consisting mainly of " water-molecules " ; but at
high temperatures " steam-molecules " are present in large
proportions, whilst " ice-molecules " become important as
the freezing-point is approached. Ordinary ice, like solid
nitrogen peroxide, is probably a homogeneous material,
consisting entirely of " ice-molecules," H6O3. But Tammann
by compressing to 2,500 atmospheres and cooling to — 22°C.
has frozen out a dense ice, which sinks instead of floating
in liquid air, but swells up and reverts to ordinary light ice
when warmed up to — 130° C. ; this dense ice is probably
composed of the denser " water-molecules," and may be
regarded as pure dihydrol, H4O2. There are, however, four
or five varieties of dense ice and two or more of light ice
known (Tammann, Annual Reports, 1910, 1912, 1913;
Bridgman, Proc. Amer. Acad., 1912, 47, 441).

B. THE CONDITIONS OF CHEMICAL CHANGE.

H, B. Baker (1894) on the influence of moisture on the
dissociation of sal-ammoniac. — In his earlier papers Deville
described dissociation as a " spontaneous decomposition "
by heat without the intervention of any chemical agency.
The later work of H. B. Baker and others has rendered it
doubtful whether dissociation can ever occur under such
conditions. In his experiments on the " Influence of
Moisture on Chemical Change "( Trans. Chem. Soc., 1894,
65, 611-624), H. B. Baker found that dry ammonium
chloride could be vaporised without decomposition into
the dried bulb of a Victor Meyer's vapour-density apparatus,
specially constructed of hard glass and heated to 350° C.
Six experiments gave an average vapour-density 27*8,
relatively to hydrogen, as compared with 2675 calculated

xx DISSOCIATION 529

for the formula NH4C1. A similar value, 28*8, was
obtained (Trans. Chem. Soc., 1898, 73, 426) by Dumas'
method, the dried salt being vaporised into a dried bulb
protected from the moisture of the atmosphere by a long
tube of phosphoric anhydride.

Baker also found that ammonia and hydrogen chloride,
after being dried during a week by contact with purified
phosphoric anhydride,1 did not contract on mixing and
gave no fumes of ammonium chloride. But " If a trace of
moisture be admitted to the mixture of dried gases, dense
white fumes are at once produced, and the mercury rushes
up into the tube" (Trans. Chem. Soc., 1894, 65, 615).

Carefully-prepared quicklime and ammonium chloride,
after drying during 1 7 days, gave no ammonia when mixed
and heated, the ammonium chloride subliming unchanged
from the lime. But when the rest of the ammonium
chloride was mixed with lime, after allowing moist air to
enter the tube, ammonia was evolved freely on heating the
mixture.

From such experiments it appears that moisture is
necessary both for the production of sal-ammoniac from
ammonia and hydrogen chloride and for its decomposition
by heat or by the action of quicklime.

Influence of moisture on the dissociation of calomel. —
In 1900 (Trans. Chem. Soc., 1900, 77, 646) Baker showed
that calomel or mercurous chloride, dried by sulphuric
acid and vaporised in a Victor Meyer apparatus containing
nitrogen and heated to 444° C, gave a vapour-density 118-4
(hydrogen = i) ; similar material, dried during three weeks
in presence of phosphoric anhydride, gave in five experi-
ments the average value 217-4, approaching the value
235 required for the formula Hg2Cl2. Calomel vapour
dried by phosphoric anhydride did not amalgamate gold

1 If the phosphoric anhydride is pure it does not absorb ammonia
that has been dried by potash.

M M

530 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

leaf, but the moist vapour at once attacked it owing to the
liberation of mercury in the dissociation :

Alexander Smith has recently shown (Zeit.physikal. Chem.,
19.11, 76, 713) that calomel, heated at 115° C. during 5!
months in a sealed bulb containing phosphoric anhydride,
became so thoroughly dried that it would not vaporise at
all at 352° C. (usual vapour pressure 347 mm.).

Baker's work on the oxides of nitrogen (1894-1912). —
Baker found in 1894 (loc. tit. p. 613) that oxygen and nitric
oxide if carefully prepared and dried by phosphoric anhydride
during 10 days did not combine when mixed ; no con-
traction took place over mercury and no darkening of the
gas could be seen. But " when a small quantity of water
was introduced, dense brown fumes of the peroxide were
immediately produced."

The dissociation of the gaseous peroxide,

did not appear to be affected by drying the gas ; but later
experiments showed that the boiling-point of the liquid was
raised from + 22° to over + 69° C. by drying during more
than a year (Trans. Chem. Soc., 1912, 101, 2341).

The trioxide, N2O3, or nitrous anhydride, forms blue
crystals melting at - 1 1 1° C., but dissociates completely when
vaporised at atmospheric temperatures :

N8O8.-» NO2 + NO or 2N2O3 -> N2O4 + 2NO.
The dried liquid, when evaporated, gave vapour-densities
ranging from 38*1 to 62*2 (hydrogen = i). As N2O3 requires
V.D. 38 only, the vapour was not merely undissociated, but
actually associated, consisting largely of complex molecules,
e.g. N4O6, analogous to those present in the vapour of phos-
phorous anhydride, ~P4O6(Tnms. Chem. Soc., 1907,9!, 1862).
After 3 years' drying the boiling-point of the liquid was
raised from - 2° to + 43° C. :

XX DISSOCIATION 531

" The vapour of the very dry trioxide was red, and on
cooling to +10° C. it condensed to a green liquid, which on
further cooling turned bright blue, showing that it was still
nitrogen trioxide. On allowing some nitrogen, dried by
passage through a long column of phosphoric oxide, to enter
the tube, the small amount of moisture it contained caused
rapid dissociation, and the resulting sudden increase blew
out the stopper of the tube."

Five years' drying gave a liquid which, instead of becoming
olive-green above - 20° C., was permanently blue at ordinary
temperatures and perhaps consisted of undecomposed N4O6
(Trans. Chem. Soc., 1912, 101, 2341).

Influence of moisture on combustion. — Baker's work on
dissociation formed part of an investigation initiated by
H. B. Dixon, on the influence of moisture on chemical
change and especially on combustion. The following
actions are checked, or altogether arrested, by careful
drying :

2CO + O2 -> 2CO2

C + O2 -> CO2

S + O2 ->«6O2

P4 +502->P4010

2H2 + O2 -> 2H2O

A dried mixture of carbonic oxide and oxygen was sparked
without exploding (H. B. Dixon, Phil. Trans., 1884, 175,
630). Both sulphur and phosphorus were distilled to and fro
in sealed tubes containing dried oxygen without taking fire,
whilst purified carbon was only partially burned by prolonged
heating in oxygen (H. B. Baker, Phil. Trans., 1888, 179,
571). Hydrogen and oxygen were found to be almost
incapable of exploding when pure and dry (Baker, Trans.
Chem. Sac., 1902, 81, 400).

The mixed gases, prepared by the electrolysis of purified
baryta, were passed through a column of phosphoric an-
hydride and collected in cleaned vacuous tubes of hard glass,

532 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

each containing a small tube filled with purified phosphoric
anhydride. The mixture was sealed in by a plug of fusible
metal and put aside to dry (Fig. 57).

" Comparative tubes were made at the same time from
the same length of tubing and treated in precisely the same
way, except that no phosphoric oxide was sealed up in them.

After 10 days' drying, two
such tubes were heated
side by side in the same
Bunsen burner flame. In
twelve experiments, the
wet tube exploded and
the dry tube did not. In
only one experiment has
a dry tube exploded, but
in this case the tube had
been carried for some
miles by hand, and most
probably some of the
phosphoric oxide had
been shaken into the part
of the tube which was
heated. In two experi-
ments, where only 2 days'
drying had been allowed,

PyPJl'P.. OXIDE
FUSIBLE METAL

water was slowly formed
in the dried tube, but
although visible moisture
was present, no explosive
took place, and a slow combination only

FIG. 57.— H. B. BAKER'S APPARATUS FOR
HEATING DRIED MIXTURES OF HYDRO-
GEN AND OXYGEN (i) by means of a
burner, (2) by a silver wire heated to fusion.

combination
occurred."

Silver was melted in the dried gas, and small sparks
were passed through it without producing an explosion :
but the melting of platinum and the passage of larger
sparks always detonated the mixture.

Conditions under which chemical change takes place.—
The experiments described in the preceding paragraphs
show that chemical changes frequently involve factors which

xx DISSOCIATION 533

are not shown in the ordinary equations. The metals potass-
ium, sodium, magnesium, zinc, cadmium and iron do not
dissolve in liquefied hydrogen chloride, nor does it act at all
on quicklime or marble (Gore, Proc. Roy. Soc.> 1865, 14, 204-
213). Dry chlorine does not act upon sodium either at the
ordinary temperature or when fused (Wanklyn, Chem. News,
1869, 20, 271), and has but little action on silver, zinc, magne-
sium and potassium (Cowper, Trans. Chem. Soc., 1883, 43,
153-155). Nitric acid, which has been freed from water and
from lower oxides of nitrogen, has no action upon purified
copper, silver, cadmium or mercury, nor upon commercial
magnesium at ordinary temperatures ; even when boiling it
does not attack iron or tin, nor dissolve Iceland spar or marble
(Veley, Phil. Tra/is.,i8^S, A. 191, 388). Copper does not
dissolve readily even in diluted nitric acid if freed from
lower oxides of nitrogen, but as soon as it begins to dissolve
these are produced in increasing quantities and the action
is rapidly accelerated (Veley, Proc. Roy. Soc.^ 1889, 46, 216-
222). Highly purified zinc, redistilled in a vacuum, is
"nearly unacted on by sulphuric or hydrochloric acid"
(Reynolds and Ramsay, Trans. Chem. Soc., 1887, 51, 857),
but it is rendered soluble by contact with the less active
metals, such as lead, copper and platinum. Metals do not
combine with dried chlorine, nor sodium and potassium
with dried oxygen ; dry lime does not combine with carbonic
anhydride (Veley, Trans. Chem. Soc., 1893, 63, 821), nor
does sulphuric anhydride combine with dry quicklime or
with the dry oxide of copper (Baker, Trans. Chem. Soc.,
1894, 65, 622). Pure iron does not rust, does not dissolve
in cold dilute sulphuric or nitric acid, and has no action
on aqueous copper sulphate or copper nitrate (Lambert,
Trans. Chem. Soc., 1912, 101, 2069).

In the case of metals dissolving in acids, the conditions
for dissolution are clearly identical with those for the pro-
duction of a battery, namely, two electrodes (e.g. zinc and

534 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

copper) and an electrolyte (e.g. dilute sulphuric acid). The
pure anhydrous acids do not conduct and do not therefore
provide a proper electrolyte. Pure zinc, on the other hand,
fails to dissolve in dilute acids because, in the absence of
impurities, such as lead, arsenic and carbon, there is only
one electrode provided for the passage of the current.

Armstrong (1885) regards chemical change as reversed
electrolysis. — H. E. Armstrong, in 1885 (B.A. Report, Aber-
deen, 945-964), applied these ideas boldly to chemical
changes of every type. In electrolysis, an electrolyte is
decomposed by energy supplied from an external source :
in a battery, decomposition of the electrolyte is itself a source
of energy. The decomposition of the electrolyte taking
place in the single fluid cell results in the conversion of

Zn + SO4|H2 + Cu
into ZnSO4 +H2 + Cu.

In the Daniell cell two electrolytes are decomposed in
series, the system

Zn + S04]H2 + S04|Cu + Cu
changing to ZnSO4 + H2SO4 + Cu + Cu.

In the Grove cell the change is from

Zn + SO4|H2 + ONO2H + Pt
to ZnSO4 + H2O + NO2H + Pt;

dilute sulphuric acid is electrolysed, but the concentrated
nitric acid in the cell is reduced, rather than electrolysed,
and may be regarded as constituting the second of the two
electrodes between which the electrolysis of the sulphuric
acid takes place.

Armstrong, adopting a hint given by Faraday, suggested,
in 1885 that all chemical changes are electrolytic in char-
acter and take place under the same conditions that prevail
in a cell, namely, by the co-operation of three distinct substan-
ces, at least one of which must be an electrolyte. In the union

xx DISSOCIATION 535

of hydrogen and oxygen these two substances would func-
tion as " depolarising " electrodes for a trace of water
acting as electrolyte :

2H2 + 2O|H2 + O2
giving 2H2O + 2H2O.

Seventeen years before Baker's experiments were
carried to a successful conclusion, Armstrong acting on this
belief, ventured to affirm, " that some day it would be
ascertained that a mixture of pure oxygen with pure hydro-
gen was not explosive" (Proc. Chem. Soc., 1885, 1, 39).
He added, however, that " Water not being an electrolyte, . . .
it was difficult to understand that the presence of water pure
and simple should be of influence in the case of a mixture
of oxygen and hydrogen " (loc. tit. p. 40) ; to be effective
the water must be sufficiently impure to render it con-
ducting.

The startling verification of these two daring predictions
by Baker's experiments in 1902 affords the strongest
evidence of the correctness of the views on which they were
based.

SUMMARY AND SUPPLEMENT.
A. DISSOCIATION AND ASSOCIATION.

Grove, in 1847, discovered that steam could be decomposed
by hot platinum, bubbles of detonating gas to the amount of
sj1^ being formed when the steam condensed. The decomposi-
tion actually amounts to

0-0078% at 1124° C. n8% at 1882° C.
0-0189% at 1207° 177% at 1984°
0-034% at 1288°

(Nernst, Wartenburg, Zeit. physikal. Chetn., 1906, 56, 533,
541). If the steam were cooled slowly, the hydrogen and
oxygen would recombine as shown by the lower arrow in the
equation

536 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP.

but by cooling quickly to a temperature below that at which the
gases recombine it is possible to fix the decomposition products
as they are formed at higher temperatures.

St. Claire Deville, in 1857, described the "spontaneous
decomposition " of substances by heat as dissociation, but in
1863 limited the term to reversible decompositions. By using a
" hot-cold tube " he was able in 1865 to prove the dissociation of :

(1) Water 2H2O '^± 2H2 + O2

(2) Carbon dioxide 2CO2 ^± 2CO + O2

(3) Carbonic oxide 2CO i± C + CO2

(4) Sulphur dioxide 3SO2 '4i± S + 2SO3

(5) Hydrogen chloride 2HC1 ^ C12 + H2 (the chlorine

being fixed in the form of metallic chlorides on the

cold tube).

He brought about similar decompositions by sparking, but
showed that sulphur dioxide and oxygen could be recombined
by sparking over sulphuric acid and hydrogen and nitrogen by
sparking in presence of hydrogen chloride :

2SO2 + O2 '^± 2SO3 -> (fixed as H2S2O7)
3H2+ N2 ^± 2NH3 -> (fixed as NH4C1).
Pebal, in 1862, separated by diffusion the products of dissocia-
tion of ammonium chloride :

High temp.

NH4C1 -^ NH3 + HC1,

Low temp.

He thus proved the correctness of the view suggested by
Cannizzaro (1857) and by Kopp (1858) that the abnormalities
detected by Bineau (1838) in the vapour densities of the
ammonium and phosphonium salts were due to the dissociation
of the salts. Deville, however, rejected all cases of dissociation
in which the products could not be fixed by suddenly chilling
the vapour.

Cahours,in 1844 and 1847, discovered that the vapour-densities
of acetic acid and of phosphorus pentachloride diminished as the
temperature rose. Playfair and Wanklyn, in 1862, observed a
similar phenomenon in nitrogen peroxide and attributed it to
dissociation :

C4H8O4 ^± 2C2H4O2

xx DISSOCIATION 537

Above 1 50° C. nitrogen peroxide consists entirely of the brown
gas NO2 ; at 100° C. there is I of N2O4 by weight, at 50° C. f, and
at 27° C. f. The liquid (b.p. 22° C.) consists almost entirely of
N2O4, which finally freezes out at - 10° C. as a colourless ice.
The dissociation of the colourless tetroxide is accompanied by an
absorption of heat, and produces a remarkable increase in the
heat-capacity of the gas.

Similar phenomena occur in the case of water. The vapour
at 100° C. consists of " steam-molecules," H2O, mixed with about
9% of " water-molecules," H4O2. The liquid is probably a
similar mixture, but consisting mainly of " water-molecules,"
H4O2, in which at lower temperatures " ice-molecules" (perhaps
H6O3) of lower density are formed in increasing proportions,
giving rise finally to an expansion of volume between 4° and o° C.
Ordinary light ice is evidently composed of the same "ice-
molecules" in the crystalline state, but the dense ice, which
freezes out under pressures above 2,300 atmospheres (Tammann ),
may be composed of the denser " water-molecules." There are
probably 4 or 5 varieties of dense ice, one of which melts at
78CC. under 20,000 atmospheres pressure (Bridgman) : it is
also possible that light ice exists in several forms (Tammann).

B. THE CONDITIONS OF CHEMICAL CHANGE

Wanklyn, in 1869, showed that dry chlorine did not act on
solid or on molten sodium ; this observation was confirmed, and
extended to other metals, by Cowper in 1883.

H. B. Dixon, in 1884, found that a dried mixture of carbonic
oxide and oxygen was not explosive.

H. B. Baker, in 1888, showed that sulphur and phosphorus
could be distilled and that purified carbon could be heated to
redness in dry oxygen without ignition. In 1894 he showed that
dry ammonium chloride could be vaporised without dissociating
and that it sublimed unchanged from quicklime ; conversely, dry
ammonia and dry hydrogen chloride did not combine when
mixed. Sulphur trioxide did not combine with quick-
lime or with cupric oxide. Dry nitric oxide did not combine
with oxygen. In 1900, he showed that dry calomel could be
vaporised, as Hg2Cl2, without decomposition and that the vapour
did not amalgamate gold-leaf : the later work of A. Smith has

538 HISTORICAL INTRODUCTION TO CHEMISTRY CH. xx

shown that further drying prevents the calomel from vaporising
at 352° C. In 1902, Baker showed that pure dry hydrogen and
oxygen did not inflame at a red heat, nor when a silver wire
was melted in the gas ; when dried less perfectly, a slow
combustion took place, but there was no explosion even when
visible moisture was present in the tube. In 1907, he showed
that nitrous anhydride, which dissociates completely at ordinary
temperatures (N2O3 -> NO2 + NO), can be vaporised as a
mixture of N2O3 and N4O6 if carefully dried. In 1912 he
showed that longer drying raised the boiling-point of the
anhydride from -2° to +43° C. and that the boiling-point of
nitrogen peroxide was raised from +22° to +69° C.

The experiments of Gore, Veley, Reynolds and Ramsay,
Lambert and others have shown that purified acids do not act
on commercial metals, and that purified metals are not acted on
by commercial acids. The conditions for interaction are the
same as in a battery, which consists of three essential parts,
namely, two electrodes and an electrolyte. Armstrong, in 1885,
suggested that all chemical action was reversed electrolysis^ a
view that is supported by the whole of Baker's work.

BIOGRAPHICAL INDEX

ALBERTUS MAGNUS (1205-1280). Mercury, sulphur and salt,

222, 448

AMPERE, ANDRE MARIE (1775-1836). Fluorine, 235

1816. The ammonium radical, 267
ARISTOTLE (384-322 B.C.). Earth, air, fire and water, 112,

222, 448

ARMSTRONG, H. E.

1885. Chemical change as reversed electrolysis, 534 ; hydro-
gen and oxygen not explosive, 535

1901. Periodic classification, 465

AVOGADRO, AMADEO, b. 1776, Turin; d. 1856, Turin. Pro-
fessor of Physics in the University of Turin

1811. Avogadro's hypothesis, 338; elementary molecules
divisible, 339; composition of water, 341, nitric
oxide, ammonia and muriatic acid, 342
BAKER, H. B.

1888. "Combustion in Dried Oxygen," sulphur and phos-
phorus distilled, and carbon heated in dry oxygen,

53i

1894. " Influence of Moisture on Chemical Change." Dis-
sociation of sal-ammoniac, 528, nitric oxide and
oxygen, 530

1900, Dissociation of calomel, 529

1902. Dried hydrogen and oxygen not combustible, 531
1907. Vapour-density of dry nitrous anhydride, 530
1912. Boiling-point of dried N2O3 and N2C»4, 530

BALARDJ ANTOINE JEROME (1802-1876)

1826. Discovery of bromine, 240
BARKER, T. V. Isomorphism, 382
BARLOW, W., AND POPE, W. J. Structure of crystals, 441

539

540 BIOGRAPHICAL INDEX

BAUME, ANTOINE (1728-1804)

1760. Vitriolated tartar (potassium sulphate) decomposed by

nitric acid, 498
BECHER, JOHN JOACHIM (1635-1682)

Theory of phlogiston, 31

1669. Preparation of ethylene, 387

BERGMAN, TORBERN OLAF (1735-1784). Professor of
Chemistry and Pharmacy at the University of Upsala.

1779-1792. Physical and Chemical Essays (Opuscula Physica
et Chemica, Six Volumes * ; the earlier essays,
tr. E. Cullen, Two Volumes, 1784). Amongst these
Essays are the following :

(a) 1774. "On the Aerial Acid." Fixed air an acid, 101 ;

present in aerated waters, 102 ; reddens litmus, 102 ;
dissolves iron and zinc, 103 ; dissolves chalk and
converts it into crystals of calcareous spar, 105 ;
analysis of bicarbonate, 109

(b) 1778. "Of the Analysis of Waters." Hardness produced by

acids or by salts with an earthy, metallic base, 107

(c) " Of the Fulminating Calx of Gold." Composition of

fulminating gold, 259

1775. A Dissertation on Elective Attractions, tr. T. Beddoes,
1785. Tables of affinity in the moist way and in the
dry way, 500

All quotations are taken from the contemporary English

translations
BERTHELOT, M., AND OGIER, J.

1883. Molecular-heat of nitrogen peroxide, 526

BERTHOLLET, CLAUDE Louis, b. 1748, Annecy, in Savoy ;
d. 1822, Arcueil, nr. Paris. After the revolution, Berthollet took
the place of Lavoisier as the recognised leader of French science.
He assisted Napoleon in organising the expedition to Egypt,
and read his paper on " Chemical Affinity " at Cairo. The
Socie'te d' Arcueil, which met at his house at Arcueil, near Paris,
included Berthollet (father and son), La Place, Biot, Gay-Lussac,
Humboldt, Thenard, De Candolle, and Decostils. Three volumes
of Memoirs, published in 1807, 1809, and 1817, contain some
of the most important papers of this period of chemistry

1 Vol. IV. (1787). " Meditationes cle Systemate Fossilum Naturale,"
translated from a Swedish paper dated 1784, appears to contain the
earliest use of the word "ammonia" (loc. «'/., p. 293) ; Kopp, how-
ever, gives the date 1782 for the first use of this word.

BIOGRAPHICAL INDEX 541

1785. "Analysis of the Volatile Alkali." Ammonia decom-
posed by chlorine, 261 ; by oxide of gold in aurum
fulminans, 261; by sparking, 262; nitrous oxide
from ammonium nitrate, 79, 260

178$. "On Dephlogisticated Marine Acid." Pyrolusite
decomposed by heat, 211 ; chlorine water, 213 ;
chlorine hydrate, 214 ; chlorine water decomposed
by light, 214; chlorine an oxide, 214 ; chlorine not
an acid, 219

1787. "On Prussic Acid." Composition of prussic acid, 243.

1787. System of Chemical Nomenclature. See under

LAVOISIER

1788. '' Observations on some Combinations of the Dephlo-

gisticated Marine Acid, or of Oxygenated Muriatic
Acid." Preparation and properties of chlorate of
potash, 227

1789. "Bleaching of Cloths and Yarn by Oxygenated

Muriatic Acid." Chlorine as a bleaching agent, 228
1796. " On Sulphuretted Hydrogen." Sulphuretted hydrogen

an acid, 176; phosphoretted hydrogen not an acid,

181
1799. Laws of Chemical Affinity. Mass action, 501 ; balanced

actions, 503 ; influence of precipitation, 504, and

vaporisation, 508
1803. Essay on Chemical Statics. Variable proportions, 299,

in oxides of lead and iron and in sulphides, 300
1809. Density and composition of carbonic oxide, 139, 143

BERZELIUS, JONS JACOB, b. 1779 ; d. 1848, Stockholm. Pro-
fessor of Medicine and Pharmacy at Stockholm

1803. "Galvanic Experiments," by Hisinger and Berzelius.
Salts decomposed by electrolysis, 273 ; ammonia
decomposed, 265

1808. " Electrochemical Researches on the Decomposition
of the Alkalis of the Earths," by Berzelius and
Pontin. Ammonium amalgam, 265

1811 "Research to Determine the Fixed and Simple

-12. Proportions in which the Constituents of Inorganic
Nature are Combined with One Another" (repr.
Ostwald's Klassiker der Exakten Wissenschaften,
No. 35). Law of multiple proportions, 306 ; law of
reciprocal proportions tested, 317. Composition of
sulphides of iron, 174, 307 ; oxides of lead, sulphur,
copper and iron, 307 ; sulphates and sulphites, 375 ;
potassium chlorate, 377 ; chalk, 378 ; lead nitrate,
379

542 BIOGRAPHICAL INDEX

1814. New System of Mineralogy (in Swedish ; French trans-

lation, 1819)

1815. " Experiments to Determine the Proportions in which

the Elements of Organic Nature are Combined."
Formulae of organic compounds, 391

1819. Essay on the Theory of Chemical Proportions and the

Chemical Influence of Electricity. List of fifty
elements, 449 ; chlorine an oxide, 222 ; nitrogen an
oxide, 265 ; atomic symbols, 295 ; measurement of
combining weights with oxygen as standard, 312 ;
ratio of oxygen in acid and base, 376—380 ; dualistic
or electro-chemical theory, 394, 397 ; compound
organic radicals, 398

1820. The Use of the Blowpipe i?i Chemical Analyst's and in

the Examination of Minerals, 1820 ; English trans-
lation J. G. Children, 1822

1820. "New Determinations of the Proportions of Water and
of the Density of some Elastic Fluids," by Berzelius
and Dulong. Composition of water, 124; com-
position of carbonic anhydride, 145

1823. Jahresbericht. Symbol Am, 268

1825. "Memoir on Sulpho-salts." Definition of a salt, 248 ;
haloid-salts and amphi-salts, 247 ; halogens, 247,

45i

1827. Jahresbericht. Atomic weights based on atomic heats
and on isomorphism, 363, 373 ; criticism of Thom-
son's integral atomic weights, 483

1832. On the benzoyl-radical, proi'n and orthrin, 400 ;
symbols Bz and Ad, 401 ; Ae (aetherin), 402

1838. No chlorine in organic radicals, 409

1839. Jahresbericht. Symbol Cy, 399

1840. Jahresbericht. Allotropy, JOT.

1843. " Picramyl " as the radical of oil of bitter almonds, 409.

From 1822 to 1848 Berzelius wrote a series of Annual
Reports on the Progress of Physical Science. The German
translations by Gmelin and by Wohler are familiar under the
name of Berzelius's Jahresberichte, and are quoted in several
places in the text. Another important publication is his Text-
book of Chemistry, First Edition, 3 Vols. (in Swedish), 1808-
1818 ; Fifth Edition, 5 Vols. (in German), 1843-18/18 ; quoted on
P- 398

BILTZ, HEINRICH

1901. Vapour-density of sulphur, 524

1902. Periodic classification, 405
BINEAU, AMAND (1812-1861)

BIOGRAPHICAL INDEX 543

1838. Vapour-densities of ammonium and phosphonium salts,
519

BLACK, JOSEPH, b. 1728 near Bordeaux; d. 1799, Edinburgh.

Professor of Anatomy and Chemistry at Glasgow ; then

Professor of Chemistry at Edinburgh

1755. On Magnesia Alba (A.C.R. i.). Use of quantitative
methods, 297. Magnesia from mother of nitre,
bittern, and Epsom salts, 19, 59 ; muriate of mag-
nesia, 20 ; acetate of magnesia, 2 1 ; magnesia alba
decomposed by heat, 59 ; calcination of chalk to
lime, 49 ; " fixed air," in chalk, 50 ; in air and water,
51 ; in mild alkalis, 53, and in sal volatile, 56, 94;
alkalis, mild and caustic, 54 ; discovery of bicarbon-
ates, 109 ; alumina from alum, 60

1757. Lectures on the Elements of Chemistry (published 1803).
Fixed air a product of respiration, fermentation and
combustion, 385

BLEIER, O. AND KOHN, L.

1900. Vapour-density of sulphur, 524
BOISBAUDRAN, LECOQ DE

1875. Discovery of gallium, 475, 476, 477
BOSE, J. C. Density of steam, 527

BOYLE, THE HON. ROBERT, b. 1627, Cork ; d. 1691, London.
Seventh son of the Earl of Cork, and one of the founders of the
Royal Society ; founder of Modern Chemistry

1660. The Spring of the Air. New air-pump, 321 ; elasticity

of air, 65, 320 ; its causes, 322 ; artificial gases, 65

1661. Sceptical Chymist. The elements, 448

1662. Defence against Linus. Boyle's Law, 324 ; air ex-

panded by heat, 325

1663. Experiments on Colours. Action of acids and alkalis

on vegetable tinctures, 24

1672. Propagation of Flame in Vacua Boyliano, 30

1673. New Experiments to make Fire and Flame Stable and

Ponderable. Calcination of copper, tin, iron, and
lead, 28 ; calcination of tin in sealed flashes, 29
1680. Producibleness of Chymical Principles. Definition of a
" salt," 3 ; vinegar upon nitre, 4 ; alkali from sea-salt,
60 ; properties of alkalis, 62

The original tracts were collected by Shaw (1725) into 3 vols.
(abridged), by Birch into 5 vols. (1744), and again into 6 vols.
(1772). Quotations are taken from the original tracts, or from
the abridged edition (1725).

544 BIOGRAPHICAL INDEX

BRAGG, W. H. Structure of crystals, 441, 487
BRAND

1669. (?) Preparation of phosphorus, 177
BRIDGMAN

1912. Dense ice, 528
BRUN

1630. Calcination of tin

BUNSEN, ROBERT WILHELM, b. 1811, Gottingen ; d. i
Heidelberg. Professor of Chemistry at Heidelberg

1837-40. "Organic Compounds of Arsenic" and " Investiga-
tions in the Cacodyl series " (Klassiker, No. 27)
(describes the compounds of the radical cacodyl,
As(CH3)2 or AsC2H6)

1842. " Use of Carbon in the Voltaic Battery." The Bunsen
battery

1854. Volumetric Methods

1855. "Photochemical Investigations" (with H. E. Roscoe)

" Preparation of Lithium"
1857. Gasometry (tr. Roscoe, 1857)
1860-61. "Chemical Analysis by Spectral Observations"

(with Kirchhoff). Discovery of caesium. The

spectroscope. Discovery of rubidium
1866. " Flame-reactions " (in book-form, 1880). The Bunsen

burner, 395 ; specific heat of indium, 475
1868. " Washing of Precipitates." The water filter-pump
1870. " Calorimetric Investigations." The ice-calorimeter

CAHOURS, AUGUSTS (1813-1891)

1847. Variation of vapour-density with temperature, 522
CANNIZZARO, STANISLAO, b. 1826, Palermo ; d. 1910, Rome.
Professor of Chemistry at Genoa, Palermo and Rome

1857. Abnormal vapour-densities explained by dissociation,

521

1858. " Sketch of a Course of Chemical Philosophy given in

the Royal University of Genoa" (A.C.R. XVI II).
Hydrogen as a standard of molecular weights, 343 ;
method of determining atomic weights, 347 ; atomic
weights of hydrogen, 347, oxygen, 348, chlorine,
nitrogen, 349, sulphur, carbon, 350, phosphorus,
arsenic, and mercury, 351 ; formulae of volatile
compounds, 352 ; molecular heats of compounds, 367

BIOGRAPHICAL INDEX 545

CAVENDISH, THE HON. HENRY, b. 1731, Nice; d. 1810,
London

1766. "Experiments on Factitious Air," 65; inflammable

air, 66 ; density of gases, 67, 94 ; weight of fixed air
in marble, sal-volatile, pearl-ashes, etc., 69, 109 ;
solubility of fixed air in water, fixed air confined
over mercury, 70 ; gas set free from muriatic acid by
heat, 8 1 ; equivalent weights of potash and of marble,
310
1772 (?). Fixed air from burning charcoal, 95

1767. "Experiments on Rathbone- Place Water," fixed air

renders chalk soluble in water, 104
1784. " Experiments on Air." (A.C.R., III.)
Part I. 178 1. Pure water formed by burning common air and
inflammable air, 114 ; water produced by exploding
inflammable air with oxygen, 115, contains nitric
acid, 1 88 ; combining volumes of hydrogen and
oxygen, 330 ; gypsum formed by oxidation of sulphite
of lime, 164; charcoal deflagrated with nitre, 187
Part II. 1784. Nitric acid produced by sparking air, 189;
nitrous acid also formed, 199 ; residue of air not
absorbed (argon), 192

1788. Reciprocal proportion in the neutralisation of acids, 310

1798. "Experiments to Determine the Density of the Earth"

CHAPTAL, JEAN ANTOINE CLAUDE (1756-1832). Calico-printer,

and Professor of Chemistry at Montpellier
1791. Nitrogen, 45, 191.
CHARLES, J. A. C. (1746-1823)

1787. Expansion of gases by heat, Charles's Law, 327
COURTOIS, BERNARD (1777-1838). Manufacturer of saltpetre

1811. Discovery of iodine, 235 ; metallic iodides, 238
COWPER, RICHARD. Metals heated in dry chlorine, 533
CROOKES, SIR WILLIAM

1861. Discovery of thallium, 475
1908. Atomic weights of B, Sc, Y, Yb, 483
CRUIKSHANK , WILLIAM (1745-1800), Chemist to the Royal
Artillery.

1801. Composition of carbonic oxide, 140 ; carbonic oxide
prepared by reducing metallic oxides with charcoal,
139, and by the action of iron on carbonates, 140
Combination of hydrogen and chlorine, 223

D ALTON, JOHN, £., 1766, nr. Cockermouth ; </., 1844, Man-
chester. For eight years teacher of Mathematics and Physics
at New College, Manchester.

N N

546 BIOGRAPHICAL INDEX

1802. The atmosphere composed of atoms, 317, but these do
not occupy equal volumes, 336

1808. New System of Chemical Philosophy. Part I. The
atomic theory 292 ; atomic symbols, 294 ; assumes
the law of fixed proportions, 297 ; law of multiple
proportions, 304 ; against the theory of variable pro-
portions, 305 ; hydrogen as a standard of combining
weights, 312; formula for water, 314; rejects the
law of equal volumes (Avogadro's hypothesis), 336

1 8 10. New System of 'Chemical Philosophy. Part 1 1. Rejects
Gay-Lussac's Law of Volumes, 337. Carbonic an-
hydride decomposed by sparking, 144 ; composition of
olefiant gas, 155 ; olefiant gas decomposed by spark-
ing., 156 ; properties of marsh gas, 156 ; composition
of marsh gas, 157 ; marsh gas decomposed by spark-
ing, 1 57 ; composition of coal gas, 1 58 ; composi-
tion of sulphurous anhydride, 168 ; combination of
sulphurous anhydride with oxygen, 169 ; prepara-
tion and composition of sulphuretted hydrogen, 172,
175 ; preparation of phosphorus, 177 ; phosphor-
etted hydrogen, 180 ; combination of nitric oxide and
oxygen, 201
DANIELL, JOHN FREDERIC, (1790-1845), Professor of Chemistry

at King's College, London, and inventor of the Daniell cell

1840. Salts as compounds of two radicals, 270

DAVY, Sir HUMPHRY, b., 1778, Truro ; d., 1829, Geneva.

Professor of Chemistry at the Royal Institution ; President of

the Royal Society.

1800. Researches, Chemical ajid Philosophical chiefly concern-
ing nitrous oxide, or dephlogisticated nitrotis air, and
its respiration. Preparation and properties of nitrous
oxide, 79 ; nitrous oxide as an anaesthetic, 80 ; com-
position of nitrous oxide, 193; composition of nitric
oxide, 194 ; its density, 337 ; combination of nitric
oxide with oxygen, 201, composition of nitrogen
peroxide, 195

1807. Bakerian Lecture, " On the Decomposition of the Fixed

Alkalies." Davy decomposes potash, 273 ; soda, 278 ;
properties of potassium, 275; sodium, 278; (1810),
composition of the alkalis, 283 ; suspects oxygen in
ammonia, 264 (1812), and hydrogen in sulphur and
phosphorus, 265

1808. Bakerian Lecture, "On the Decomposition of the

Earths. . . and on the Amalgam procured from Am-
monia." Composition of the earths, 280 ; properties

BIOGRAPHICAL INDEX 547

of barium, strontium, calcium, magnesium, 281 ; am-
monium-amalgam, 265 ; metallic ammonium, 267 ;
preparation of boron, 288

1808. Bakerian Lecture, " New Analytical Researches."
Action of potassium on hydrogen chloride, 224

1810. " Researches on the oxymuriatic acid." Charcoal heated
in chlorine, 221 ; chlorine an element, 220 ; composi-
tion of hydrogen chloride by volume, 223 ; action of
chlorine on sulphuretted hydrogen, 223 ; action of
metals on hydrogen chloride, 224 ; metallic chlorides

as binary compounds, 226
1810. Bakerian Lecture, "On

some of the Combinations of

oxymuriatic gas and oxygen." The name "chlorine"
proposed, 212, 221 ; chlorine an element, 222 : oxygen
displaced from lime by chlorine, 227

1812. Elements of Chemical Philosophy. Fluor Spar, 233 ;

classification of the elements, 450 ; oxygen and
chlorine as supporters of combustion, 450

1813. "Substances produced in different chemical processes

on Fluor Spar." 1814 "On the Fluoric Compounds."
The name " fluorine," 235 ; fluor spar a compound of
metal and fluorine, 235 ; hydrofluoric acid a com-
pound of hydrogen and fluorine, 235 ; chlorine an
element, 222

1814. "On a new substance which becomes a violet-coloured

gas by heat" (dated Paris, Dec. 10, 1813). " Further
Experiments and Observations on Iodine" (dated
Florence, March 23, 1814). The name iodine, 237 ;
iodine an element, 237 ; potassium iodide, 238

1815. "A solid compound of Iodine and Oxygen," 238

1816. "On the Analogies between the Undecompounded

Substances and on the Constitution of Acids." The
name "chloride," 218 ; the hydrogen theory of acids,
248

1 8 1 8. On the Safety Lamp, with some Researches on Flame.

All quotations are from The Collected Works of Sir Humphry

Davy, 9 Vols., 1840, or from the Alembic Club Reprints (No. 6,

Decomposition of the Alkalis and Alkaline Earths ; No. 9, The

Elementary Nature of Chlorine).

DESCARTES, REN£ (1596-1650). Kinetic theory of gases, 323
DESORMES AND CLEMENT

1801. Composition of carbonic oxide, 141 ; carbonic oxide

prepared by the action of charcoal on carbonates,

141 ; on carbonic anhydride, 142, and on steam, 142

1813. Iodine, 236; hydriodic acid, 239; metallic iodides,

228 ; iodine and ammonia. 239

N N 2

548 BIOGRAPHICAL INDEX

DEVILLE, STE.-CLAIRE (1818-1881)
1857. Dissociation, 514

1863. Dissociation of water, 515, and carbonic anhydride, 516

1864. Dissociation of carbonic oxide, sulphur dioxide, and

hydrogen chloride, 516

1865. Dissociation of ammonia, 263, 517

1867. Rejects, dissociation of NH4C1,(C2H4O2)2,PC16 and
N204, 526

DEVILLE AND TROOST

1860. Vapour densities of sulphur, 524 ; cadmium, 361 ;

aluminium chloride, bromide, and iodide, 361
1867. Vapour density of nitrogen peroxide, 525 ; solid

nitrogen peroxide, 525

DEWAR, SIR JAMES.

1905. Vapour-density of iron carbonyl, 361

1913. Atomic heats at low temperatures, 366, 471, 473

DIEMAN, TROOSTWYCK, BONDT, AND LAUWERENBURGH
1794. "Oily carburetted hydrogen" or olefiant gas, 155;
"Dutch liquid" or ethylene chloride, 155

DIXON, H. B.

1884. Explosion of dried carbonic oxide and oxygen, 531

DOBEREINER, JOHANN WOLFGANG (1780-1849)
1829. Triads, 452

DULONG AND PETIT

1819. Law of Atomic Heats, 362

DUMAS, JEAN BAPTISTS, b. 1800, Alais ; d. 1884, Cannes. Pro-
fessor of Chemistry at the Sorbonne

1826. Vapour-density of stannic chloride, 361

1827. "On the Formation of Sulphuric Ether," by Dumas

and Boullay. Sulphovinic acid, 402
1827. "On the Compound Ethers," by Dumas and Boullay.

The aetherin radical, 401
1832. Vapour-density of sulphur, 524
1834. "On a New Alcohol," by Dumas and Peligot. Methyl

alcohol, inethylene, and methyl, 403.
1834. "The Law of Substitutions," 406 ; removal of hydrogen

without replacement, 407

1837. Dumas and Liebig united, 405

1838. Vapour-density of chromyl chloride, 361

1839. Theory of types, 411 ; chlorine plays the part of

hydrogen, 411

1840. Chemical and mechanical types, 413

BIOGRAPHICAL INDEX 549

1841. "Researches on the true Atomic Weight of Carbon,"

by Dumas and Stas. Composition of carbonic,
anhydride, 148, 484

1842. "Researches on the Composition of Water," 125

1859. "On the Equivalents of the Elements." Natural
series and common differences, 452 ; simple multiples
amongst atomic weights, 482 ; on Prout's hypothesis,
484

— Hydrogenium as a metal, 249 ; acid anhydrides, 249,
250

FARADAY, MICHAEL, b. 1791, Newington ; d. 1867, Hampton
Court. Professor of Chemistry at the Royal Institution
1816. "Analysis of Native Caustic Lime of Tuscany."

1820. "On the Alloys of Steel."

1821. "On Two New Compounds of Chlorine and Carbon."

Chlorination of ethylene chloride, 407
1823. " On Fluid Chlorine."
1825. "On New Compounds of Carbon and Hydrogen and

on Certain Other Products Obtained during the

Decomposition of Oil by Heat." Butylene, 404
1830. "On the Manufacture of Glass for Optical Purposes."
1845. "Liquefaction and Solidification of Gases."
1831-55. Experimental Researches in Electricity (repr. 3 Vols.,

1839, 1844, 1855), includes :
(a) 1831. " Induction of Electric Currents."
(b} 1833. " Identity of Electricities of Different Sources."

(c) 1833-4. "Electro-chemical Decomposition."

(d) 1846, "On the Magnetisation of Light."

The 1834 paper (repr. Experimental Researches in Electricity,
Vol. I. pp. 195-258), includes the definitions of electrode,
anode, and cathode (p. 196), electrolytes (p. 197), ions, anions,
and cations (p. 198), and Faraday's two laws of electrolysis,
stated as follows : " That the chemical power of a current of
electricity is in direct proportion to the absolute quantity of
electricity which passes" (p. 241), and that the "electro-
chemical equivalents {of the ions} are the same as their ordinary
chemical equivalents?
FOURCROY, ANTOINE FRANCOIS DE (1755-1809)

1787. System of Chemical Nomenclature. See under

LAVOISIER
FISCHER

1802. Equivalent weights of acids and bases, 31 1
FRANKLAND, SIR EDWARD ^825-1899), Professor of Chemistry
at the Royal Institution and at the Normal School of Science
and Royal School of Mines

550 BIOGRAPHICAL INDEX

1852. "New Series of Organic Bodies containing Metals."
Combining-power of N, P, As, Se, 435 ; combining-
power variable, 442

1855. Vapour-density of zinc ethyl, 361

1857. (Frankland and Kolbe.) Quadrivalency of the double
carbon atom C2, 436

1864. Vapour-density of zinc methyl, 361
FRIEDEL, CHARLES

1888. Vapour-density of ferric chloride, 361

GAY-LUSSAC, Louis JOSEPH, b. 1778, Limousin ; d. 1850,
Paris. Professor of Chemistry at the Ecole Polytechnique and
of Physics at the Sorbonne ; Professor of Chemistry at the
Jardin des Plantes.

1802. "Dilatation of Gases and Vapours." Expansion by

heat, 327

1805. "Ratio of the Constituents of the Atmosphere," by
Gay-Lussac and Humboldt. Combining volumes of
hydrogen and oxygen, 330

1807. " Decomposition of Sulphates by Heat," 169
1809. " Memoir on the Combination of Gaseous Substances
with each other." The law of Volumes, 334 ; ex-
plained by the atomic theory, 335 ; multiple pro-
portions in gases, 306, 325 ; combination of ammonia
with acids, 331 ; composition of nitric oxide, 195,
nitrous oxide and nitrogen peroxide, 332 ; of am-
monia, sulphuric anhydride, and carbonic anhydride,

331
1811. Physico- Chemical Researches, by Gay-Lussac and

Thenard, includes :

(a) 1808. Preparation and properties of potassium, 275, and
sodium, 279 ; formation of peroxides by burning
potassium and sodium, 283 ; preparation of boron
and silicon, 288

(b} 1809. Pure hydrofluoric acid, 234 ; muriatic acid contains
hydrogen, 215 ; combination of hydrogen and
chlorine, 216 ; chlorine not decomposed by charcoal,
219 ; muriatic acid as a hydrate of the muriatic
radical, 222

(c] 1 8 10. Organic analysis, 390
1811. Anhydrous prussic acid, 244

1813 "On Iodine" (Klassiker, No. 4), with an appendix on

-14. "Acidity and Alkalinity." Iodine an element, 238 ;

hydriodic acid and the iodides, 239 ; nitrogen iodide,

239 5 hydracids, 246, 247 ; hydrochloric acid, 218 ;

hydrofluoric acid, 235 ; hydriodic acid, 239 : analogy

BIOGRAPHICAL INDEX 551

between chlorine, iodine, and sulphur, 226, 451 ;
chloric acid and chlorates, 228

1815. "On Prussic Acid." The cyanogen radical, 399;
hydrocyanic acid and the cyanides, 245 ; cyanogen,
245, and cyanogen chloride, 407 ; use of copper
oxide in the combustion of mercuric cyanide, 394

1815. "On the Analysis of Alcohol and of Sulphuric Ether."

Vapour-density of alcohol and ether, 388, 401

1816. "On the Combinations of Azote with Oxygen." Coin-

position of nitrogen peroxide, 195 ; nitric and nitrous
acids, 202 ; liquid nitrogen peroxide, 525

GEBER. An Arab chemist, said to have been born in
Mesopotamia at the end of the eighth century ; or
to have been born in Persia, lived at Seville and died
there A.D. 765. Greatest of the alchemists, 2 ; pre-
paration of oil of vitriol, 13 ; nitric acid ("dissolving-
water"), 14; aqua regia, 14; lunar caustic or silver
nitrate, 19 ; action of acids on metals, 16 ; prepara-
tion of caustic alkali, 54 ; sulphur dissolved by alkalis,
171

GENGEMBRE

1783. Discovery of phosphoretted hydrogen, 180, 181

1785. Properties of phosphorous acid, 179
GEOFFROY, £TIENNE FRANCOIS (1672-1731)

1718. Tables of affinity

GERHARDT, KARL FRIEDRICH, b. 1816, Strassburg ; d. 1856,
Strassburg. Professor of Chemistry at Montpellier' and then at
Strassburg

1839. On the copulation of residues, 417

1843. Halvesthe formukeof organic compounds, 41 9; corrected
atomic weights, 437

1853. Preparation of acetic anhydride, 432

1855. On conjugated radicals, 430

1856. Traite de Chimie Organique, 4 Vols., 1853-1856. Dis-

tinguishes the hydrogen and chlorine radicals from
hydrogen and chlorine gas, 420 ; four inorganic
types, 425, 432 ; primary, secondary, and tertiary
amines, 423

GLAUBER, JOHN RUDOLPH, b. 1603, Karlstadt ; d. 1668,

Amsterdam

1648-50. Philosophical Furnaces. Preparation of green
vitriol and blue vitriol, 10, 18 ; oil of vitriol, 13 ;
nitric acid, 14; spirit of salt, 14; aqua regia, 15 ;
properties of oil of vitriol, 15 ; action of oil of vitriol
on salt and on nitre, 21 ; preparation of muriates,

552 BIOGRAPHICAL INDEX

20, 21 ; preparation of spirit of hartshorn, 57 ; flint
and rock-crystal dissolved by alkalis, 171
1653 and 1658. Miraculum Mundi
1658. "Sal mirabile" or Glauber's salt, 18

Wealthy Storehouse of Treasures. Sulphur from oil of

vitriol, 163, 175

All quotations are from contemporary translations, or from
the collected Works published in London in 1789
GORE, GEORGE

1865. Properties of dry hydrogen chloride, 335
GRAY, R. W. Composition of nitric acid, 207
GRAY AND BURT. Volume of hydrogen in hydrogen chloride,

335

GRAY AND RAMSAY. Atomic weight of niton, 467
GROVE, SIR WILLIAM ROBERT (1811-1896)

1839. "On a new electric battery of great energy." The

Grove cell.
1847. Dissociation of water, 515

VON GUERICKE, OTTO (1602-1686)

1654. The Magdeburg hemispheres, 320
GUYE, A.

Measurements of atomic weights, 318 ; association of water,

527

GUYE AND PlNTZA. Combining- ratio of hydrogen and nitrogen,
335

GUYTON DE MORVEAU (1737-1816)

1787. System of Chemical Ncmeitclature. See under
LAVOISIER

HALES, STEPHEN, b. 1677, Kent ; d. 1761, Teddington. Doctor
of Theology and Vicar of Teddington

1727. Vegetable Statics. Nitrous air obtruded itself upon,
him, 71 ; apparatus for collecting and measuring
gases, 84

1733. Haemostatics

VAN HELMONT, JOHANN BAPTIST, b. 1577, Brussels; d. 1644,
Vilvorde. Lecturer in Medicine and in Chemistry at the
University of Louvain.

Willow tree grown from water, 112

gas sylvestre and gas pingue, 64, 94 ; fixed air and

water as products of combustion, 385
HENRY, WILLIAM (1774-1836)

1800. Muriatic acid gas decomposed by sparking over
mercury, 224

BIOGRAPHICAL INDEX 553

HERMANN, HANS RUDOLPH (1805-1879)

1830. Radicals of organic acids as oxides of hydrocarbons
HESS, GERMAIN HENRI (1802-1850)

1842. Law of thermoneutrality, 505
HIGGINS, WILLIAM

1799- On Bleaching. Potash an element, 257
VAN T'HOFF, JACOBUS HENDRICUS (1852-1911)

Isomerism of carbon compounds, 441
HOFFMANN.

1772. Sulphuretted hydrogen, 171

HOFMANN, AUGUST WILHELM, b. 1818, Giessen ; d. 1892,
Berlin. Professor of Chemistry at Berlin, and at the Royal
College of Chemistry, London

1850. Three types of nitrogen-bases, 423 ; tetramethyl-

ammonium iodide, phosphonium iodide, 433
HOMBERG, WILHELM, b. 1652, Batavia ; d. 1715, Paris

1702. Preparation of boric acid or " sedative salt," 288
HOOKE, ROBERT, b. 1635, Freshwater; d. 1703, London.
Assistant to Boyle, Curator of Experiments to the Royal
Society, Professor of Geometry at Gresham College, London

1665. Micrographia. Air compared with fluid saltpetre, 32

KEKULE, FRIEDRICH AUGUST, b. 1829, Darmstadt ; d. 1896,
Bonn. Professor of Chemistry in the University of Bonn
1857. "Constitution of Mercury Fulminate." The marsh-
gas type, 437

1857. "On Copulated Compounds and the Theory of Poly-

atomic Radicals." Multiple types, 427 ; mixed
types, 429 ; substitution value or atomicity, 433

1858. "Constitution and Metamorphoses of Chemical Com-

pounds and the Chemical Nature of Carbon"
(Klassiker, No. 145). Carbon tetratomic, 438 ; the
carbon skeleton, 438

1864. "On the Atomicity of the Elements." Fixed valency,

441 ; molecular compounds, 441

1865. "On the Constitution of Aromatic Substances.

Structural formula for benzene, 440

1869. Apparatus for combustion of organic compounds, 395 ;
1866-1887. Lehrbuch der Organischen Chemie, 4 Vols.,

Graphic formulae, 439

KOPP, HERMANN, b. 1817, Hanau ; d. 1892, Heidelberg.
Professor of Physics and Chemistry in the University of
Giessen

1843-1847. Geschichte der Chemie, 4 Vols.

554 BIOGRAPHICAL INDEX

1 858. Abnormal vapour densities explained by dissociation, 521
1865. Molecular heats of compounds, 367

KUNKEL, JOHANN (1638-1703).

Preparation of phosphorus, 177
LADENBURG, ALBERT (1842-1911)

1872. Vapour-density of tin triethyl and tetramethyl, 361
LANE, TIMOTHY, (1734-1807)

1769. Fixed air dissolves iron, producing chalybeate water,

108

LAMBERT, BERTRAM. Properties of pure iron, 533
LASSONE, JOSEPH MARIE FRANCOIS, (1717-1788)

1776. Discovery of carbonic oxide, 137
LAURENT, AUGUSTE, &, 1807, Langres ; d.> 1853, Paris

1837. On nuclei, 415

1846. Hydrogen and chlorine as compounds, 421

1854. Chemical Method (issued after his death by Biot, tr.
Odling, 1854). Acids as hydrogen-salts, 249

LAVOISIER, ANTOINE LAURENT, £., 1743, d., 1794, Paris, a

victim of the French Revolution

1772. Sulphur and phosphorus gain in weight when burnt,

1 68, 178

1772. "Destruction of the Diamond by Fire," 99
1774. " Calcination of Tin in Closed Vessels," 35

1774. Physical and Chemical Essays (Opuscules Physiques et

Chimiques). Fixed air formed in reduction of red
lead by charcoal, 95

1775. "The Nature of the Principle which combines with

Metals during their Calcination and which increases
their Weight." (Nov. 1774) Fixed air from red pre-
cipitate and charcoal, 97 ; eminently respirable air
from red precipitate alone, 97 ; fixed air an oxide of
charcoal, 98

1776. " Existence of Air in Nitrous Acid and on the means of

Decomposing and Recomposing this Acid." Corn-
position of nitric acid, 199; decomposition of nitric
acid in presence of mercury, 200 ; combination of
nitric oxide and oxygen, 200

1777. "Combustion of Kunkel's Phosphorus," 178; phos-

phoric acid, 179

1777. " Dissolution of Mercury in Vitriolic Acid and its reso-
lution into Gaseous Sulphurous Acid and eminently
respirabie air." Composition of sulphuric acid, 168

1777. " Vitriolisation of Martial Pyrites." Oxidation of pyrites
174

BIOGRAPHICAL INDEX 555

1777. "On Combustion in General." Oxygen theory of com-
bustion, 42

1777. "The Nature of Acids and the Principles of which they

are composed." The oxygen-theory of acids, 42, 176,
246, 248 ; the name oxygen, 43

1778. "Memoir on Heat" by Lavoisier and Laplace; con-

tinued during the winter, 1783-1784 ; an important
physical memoir, containing a description of the ice-
calorimeter

1783. " Memoir to prove that Water is not a Simple Substance

but that it is susceptible of Decomposition and Recom-
position" (1777) No acid formed by burning inflam-
mable air in common air or (1781) in oxygen, 117 ;
(1783) pure water prepared by burning hydrogen and
oxygen, 117 ; water produced by reduction of metal-
lic calces, 124 ; combustion of charcoal and composi-
tion of fixed air, 145

1784. " Decomposition of Water," by Meusnier and Lavoisier.

Steam decomposed by red-hot iron, 118

1784. " Combination of Oxygen with Spirit of Wine, Oil, and
different Combustible Bodies.'5 Water produced by
burning spirit of wine, 386 ; composition of spirit of
wine, of olive oil and of a candle, 146, 386
Also several memoirs on combustion, respiration and the
theory of phlogiston

1789. Elementary Treatise (Traite Eldmentaire de Chimie).
Calcination of mercury, 40 ; list of thirty elements,
449 ; alkalis as compounds, 257, 280*; salts as com-
pounds of two oxides, 394 ; organic compounds con-
tain oxygen, carbon, and hydrogen, with nitrogen
and phosphorus, 386 ; composition of organic acids,
387, 398 ; use of quantitative methods, 297 ; vinous
fermentation and the conservation of mass, 388 ;
chemical equations, 388

All quotations are translated from the complete Works
(Oeuvres de Lavoisier} published in six volumes, 1842-1893

LAVOISIER, GUYTON DE MORVEAU, BERTHOLLET AND

FOURCROY

1787. Chemical Nomenclature ( Methode de Nomencla-
ture Chimique, tr. 1788). Nomenclature of salts,
17 ; carbon, 98 ; carbonic acid and carbonates, 103 ;
hydrogen, 117; nitric acid and nitrates, 198, 199;
oxygen, 42 ; " oxygenated muriatic acid " (chlorine),
215 ; phosphoric acid and phosphates, 179; phos-
phorous acid and phosphites, 180 ; prussic acid and

556 BIOGRAPHICAL INDEX

prussiates, 243 ; sulphuric acid and sulphates, 163 ;
sulphate of potash, 18 ; sulphurous acid and sulphites,
164 ; sulphuretted hydrogen, 171 ; sulphurets or
sulphides, 172
All quotations are from the contemporary English translation

LIEBIG, JUSTUS VON, £., 1805, Darmstadt ; d., 1873, Munich.
Professor of Chemistry at Giessen and Munich

1825 ? Fails to recognise bromine as an element, 242

1831. "A new Apparatus for the Analysis of Organic Bodies."

Combustion with copper oxide, 394, 395 ; potash bulbs,
150

1832. "On the Radical of Benzoic Acid" by Liebig and

Wohler, 399. Benzoyl chloride, 407

1834. " Constitution of Ether." The ethyl radical, 402

1835. "Products of the Oxidation of Alcohol." Preparation

of aldehyde, 412
1837. Unites with Dumas, 405
1839. u Constitution of Ether and its Compounds." Radicals

NH2, NH3, NH4 ; C2H3, C2H4, C2H6, 405

LULLY, RAYMOND, b., 1235, Majorca; d, 1315, Tunis. Puri-

fication of alcohol, 387 ; preparation of ether, 387
MACQUER, PIERRE JOSEPH, b. 1718 ; d. 1784, Paris

1776. Water products by burning inflammable air, no
MALAGUTI, FAUSTINO JOVITA (1802-1878)

1837. Chlorinated ether, 410, 411, 412
MARGRAAF, ANDREAS SIGISMUND, b. 1709 ; d. 1782, Berlin
1748. Calcination of chalk, 49
1750. Gypsum from oil of vitriol and chalk, 16
MARIGNAC, JEAN CHARLES, b. 1817, d. 1894, Geneva

1860. Suggests small variations in composition of com-

pounds, 302 ; supports Prout's hypothesis, 485 .
MAYOW, JOHN, b. 1645, Cornwall ; d. 1679, London. Doctor
of Medicine

1674. Medico-physical Works (A.C.R., XVII.). Action of
oil of vitriol on nitre, 22 ; nitre prepared artificially,
19 ; damp gunpowder burned underwater, 32 ; nitro-
aerial spirit, 32 ; volume of air diminished by com-
bustion and respiration, 34 ; its elasticity not
impaired, 322 ; artificial gases collected over water,
65 ; their elastricity equal to that of air, 322 ; volatile
alkali displaced by fixed alkali, 56, 58 ; volatile acid
displaced by fixed acid, 22, 58, 498 ; salts still contain

BIOGRAPHICAL INDEX 557

acid and alkali, 394 ; sulphur from marcasite or
pyrites, 14, 162

MENDELEEFF, DMITRIJ IWANOWITSCH, b. 1839, Tobolsk,
Siberia ; d. 1907, St. Petersburg. Professor of Chemistry at
St. Petersburg

1869. On periodicity, 458 ; unknown elements, 458
1871. The periodic law, 459 ; typical elements, 460 ; series or
short periods, 460 ; odd and even series, 460 ; transi-
tion series, 461 ; long periods, 461 ; periodicity of
valency, 468 ; atomic weights of Be, 474, In, (Jr,
475, Au, 476 ; specific heat of indium, 475 ; missing
elements, 476, 478 ; eka-boron, eka-aluminium, eka-
silicon, 477, 4/8 ; horizontal and diagonal relation-
ships, 479
MEYER, JOHANN FRIEDRICH (1705-1765)

1764. Sulphuretted hydrogen inflammable, 171
MEYER, LOTHAR (1830-1895)
1864. Valency, 434

1869. On periodicity, 458 ; periodicity of valency, 468 ; atomic
volume curve, 469 ; atomic weight of indium, 475 ;
and uranium, 475
MEYER, VICTOR (1848-1897)

1879-89. Vapour density of Zn, KI, Cu2Cl2, ZnCl2, Sn.,C]4,

' 361

MiTSCHERLlCH, ElLHARD (1794-1863). Pupil of Berzelius ;
Professor of Chemistry at Berlin

1819. Isomorphism of arsenates and phosphates, 368
1821. Isomorphism of sulphates of Ca, Mg, Mn, Fe, Cu, Zn,
Co, Ni, and of spinels containing Cr, Mn, Fe, Al,
372

1828. Isomorphism of sulphates and selenates, 371
1832. Isomorphism of potassium permanganate and per-
chlorate, of potassium manganate with the chromate,
sulphate, and selenate, 371
MOISSAN, HENRI (1852-1907)

1886. Isolation of fluorine, 252

MOND, LUDWIG. Vapour-chemistry of nickel carbonyl, 361
MONGE, GASPARD (1764-1818)

1781. Combination of hydrogen and oxygen, 120
MORLEY, E. W.

1895. Densities of hydrogen and oxygen ; combining volumes
of hydrogen and oxygen, 334 ; composition of water,
128-133

MoSELEY, H. G. J. High-frequency spectra of the elements,
487

558 BIOGRAPHICAL INDEX

NERNST, W. Atomic heat of carbon, 366
NERNST AND WARTENBURG

1906. Dissociation of steam, 535
NEUMANN, FRANZ ERNST (1798-1895)

1831. Molecular heats of compounds, 367
NEWLANDS, J. A. R.

1864. Relations between equivalents, 454 ; missing elements,

454

1865. The law of octaves, 456 ; atomic weight of beryllium,

474, indium, 475, and uranium, 475 ; multiples of
oxygen, 483

1878. Atomic numbers, 456, 485
NEWTON, SIR ISAAC (1642-1726)

1704. Diamond combustible, 99 ; On atoms, 292, 389
NICHOLSON AND CARLISLE

1801. Water decomposed by electrolysis, 272
NILSON, LARS FREDRIK (1840-1899)

1879. Discovery and properties of scandium, 477
1884. Vapour-density of beryllium chloride, 475

1888. Vapour density of ferrous chloride, 361

1889. Vapour density of aluminium chloride, 361

ODLING, WILLIAM

1855. Multiple types and mixed types, 429; graphic repre-
sentations of substitution-values, 436

1864. Perissads and artiads, 442

1865. Vapour-density of aluminium, methyl and ethyl, 361

PEBAL, LEOPOLD VON (1826-1887)

1862. Dissociation of sal-ammoniac, 521
PELLETIER, BERTRAND, b. 1761, Bayonne ; d. 1797, Paris.

1785. Preparation of phosphorus, 177

Preparation of baryta, 141

PELOUZE, THEOPHILE JULES (1807-1867)

1843. Prout's hypothesis modified, 485
PLAYFAIR AND WANKLYN

1862. Vapour-density of acetic acid, 522 ; nitrogen peroxide,
524

PRIESTLEY, JOSEPH, b. 1773, near Leeds; d. 1804, North-
umberland, U.S.A. Priestley's papers, many of which appeared
first in the Philosophical Transactions of the Royal Society, were
published in six volumes as follows : Experiments and Observa-
tions on different kinds of Air, Vol. I. (1774), Vol. II. (1775),

BIOGRAPHICAL INDEX 559

Vol. III. (1777). Experiments and Observations Relating to
Various Branches of Natural Philosophy, with a Continuation
of the Observations on Air (1777) ; ditto, Vol. II. (1781), with an
index to both the Volumes, and a " Summary View of all the
most remarkable facts in this and the four preceding volumes " ;
Vol. III. (1786). In the text, following the method adopted in
Priestley's Summary, the six volumes are quoted by consecutive
numbers I. to VI. The experiments on the "Discovery of
Oxygen" are quoted from A.C.R. VII.

Experiments on Air, Vol. I. (17/4)

1772. Apparatus for manipulating gases, 86

1772. "Observations on Air in which Candles have Burned."
Sulphur burned in air over mercury, 168

1772. "Of Nitrous Air." 72. Nitrous air mixed with common
air> 73 > goodness of air, 74 ; diminished nitrous air
(nitrous oxide), 77

1772. "Of Air Infected with the Fumes of Burning Char-
coal." Fixed air produced by burning charcoal in
air over water, 95

1772. "Of Acid Air" (i.e. hydrogen chloride), 82

I773~4- "Observations on Alkaline Air" (i.e. ammonia), 82.
Alkaline air inflammable, 258 ; sal-ammoniac from
acid air and alkaline air, 83 ; sal-volatile from fixed
air and alkaline air, 84
Experiments on Air, Vol. II. (1775)

1774. " Of Vitriolic Acid Air." Preparation from vitriol and
olive-oil or mercury, 164

1774. "Of Dephlogisticated Air, and of the Constitution of
the Atmosphere." Water and air as simple elemen-
tary substances, 112; discovery of oxygen, 39;
nitrous air reduced by iron, 187 ; action of electric
sparks, acid air decomposed, 224 ; alkaline air
decomposed, 258
Experiments on Air, Vol. III. (1777)

1776. Eudiometers, 75, 90; production and nature of gases,
70 ; expansion of gases by heat. 327 ; nitrous air
dissolved by green vitriol, 72 ; properties of "nitrous
acid vapour" (nitrogen peroxide), 76, 525 ; properties
of "vitriolic acid air," 106
Experiments, Vol. IV. (1779)

"Of the Volatile Vitriolic Acid and Vitriolic Acid Air."
Oxidation to oil of vitriol, 167 ; liberation of sulphur,
1 68

"Of the Effect of the Electric Spark on Common Air."
Air diminished by sparking, 187, 189

560 BIOGRAPHICAL INDEX

Experiments, Vol. V. (1781)

Moisture produced by exploding common air and

inflammable air, 113
Experiments, Vol. VI. (1786)

1785. "Experiments Relating to Phlogiston." Metallic calces

reduced by inflammable air, 123

" Experiments and Observations Relating to Air and
Water." An inflammable gas from " finery cinder "
and charcoal, 138

1786. Iron heated in nitric oxide, 194
Philosophical Transactions only

1788. "Experiments and Observations on Acidity, Decom-
position of Water and Phlogiston " (two papers, not
reprinted). Nitrate of copper formed in an explosion
vessel, 1 88

PROUST, JOSEPH Louis, b. 1755, d. 1826, Angers. Professor
of Chemistry at Madrid

1799-1802. Law of Fixed Proportions, 297 ; composition of
carbonate of copper, 298 ; of two sulphides of iron,
298
1806. Fixed proportions as a test of chemical combination,

300

PROUT, WILLIAM (1786-1850). Doctor of Medicine, practising
in London

1815. "Prout's hypothesis," 480 ; density of hydrogen, 481

1816. Hydrogen as "protyl," 482

RAMSAY, SIR WILLIAM

1895. Preparation of argon, 192 ; properties of inert gases,

354
RAMSAY AND GRAY

1911. Atomic weight of Niton, 467
RAMSAY AND REYNOLDS

1887. Properties of purified zinc, 533
RAYLEIGH, LORD

1888. Densities of gas, 128
1895. Preparation of argon, 192

REGNAULT, ,HENRI VICTOR (1810-1878). Professor of Chem-
istry in the Ecole Polytechnique and of Physics in the College
of France ; director of the porcelain factory at Sevres.

1835. Chlorination of ethylene, 412 ; mechanical types, 413

Use of a counterpoise, 128

REICH AND RICHTER

1863. Discovery of indium, 475

BIOGRAPHICAL INDEX 561

REY, JEAN, d. 1645

1630. Tin and lead increase in weight on calcination, 26

RICHARDS, J. W. Measurement of atomic weights, 318

RICHARDS AND JACKSON. Atomic heats between 20° and
- 1 80°, 366

RICHARDS AND LEMEERT. Variable atomic weight of lead, 303

RlCHTER, J. B., £., 1762, Schleswig ; d. 1807, Berlin. Stochio-
metrie (1792—1794) Ueber die neueren Gegenstande
der Chymie (New Chemical Topics] (1791-1802).
Reciprocal proportions in the neutralisation of acids,
310

RONTGEN, WILHELM CONRAD. Ice-molecules and water-
molecules, 527

ROUELLE, GuiLLAUME FRANCOIS, b. 1703, nr. Caen ; d. 1770,

nr. Paris

1744. Definition of salt ; water of crystallisation, 25
1754. Alkalis, earths and metals as bases, 25

RYDBERG, J. R. Ordinals, 493 ; number of the elements, 493 ;
coronium and nebulium, 493

SAGE, BALTHAZAR GEORGES (1740-1824)
1777. Phosphorous acid, 179

DE SAUSSURE, HORACE BENEDICT, b. 1740, d. 1799, Geneva

1807, 1814. Analysis of alcohol and ether, 387

SCHEELE, KARL WILHELM, b. 1742, Stralsund ; d. 1786,

Koping, Sweden

1777. On Air and Fire (trans. 1780, repr. in part, A.C.R. VIII)
Air a mixture, 43 ; discovery of oxygen, 44 ; acids
and gases derived from nitre, 198 ; fused nitre con-
tains a weak acid, 199 ; decomposition of fulminating
gold, 260 ; sulphuretted hydrogen ("stinking sulphur-
eous air"), 171, 175

Chemical Essays, (tr. from the Transactions of the

Academy of Sciences, Stockholm, by T. Beddoes,
1786; translation repr. 1901, and in part A.C.R. XIII).
All quotations are taken from the modern reprints.

(a} 1771. "On Fluor Mineral." Hydrofluoric acid, 234

(b} 1774. "On Manganese." Discovery of chlorine, 211 ; proper-
ties of chlorine, 213 ; action on cinnabar, 226 ; am-
monia oxidised by manganese, 258 ; and by chlorine,
260

(c) 1779. "On Plumbago." Fixed air by combustion of plum-

bago, 101

(d) 1782. "On Prussian Blue." Prussic acid, 242

(e) 1770-86. Preparation of vegetable acids, 13

HIST. CHEM. O O

562 BIOGRAPHICAL INDEX

SCOTT, ALEXANDER

1887. Vapour density of Na, K, AgCl, PbCl2, CdBr2, CrCl3,
MnCl2, 361

1893. Composition of water, 127 ; combining volumes of

hydrogen and oxygen, 334
SEEBECK, THOMAS JOHANN, b. 1770, Reval ; d. 1831, Berlin

1808. Amalgam from ammonia, 265
SMITH, ALEXANDER. Properties of dried calomel, 530
SODDY, F. Radio-elements, 466 ; isotopes, 467
SODDY AND HYMAN. Variable atomic weight of lead, 303
STAHL, GEORGE ERNEST, b. 1660, Anspach ; d. Berlin.

Theory of phlogiston, 31

1702. Prepares a volatile acid from burning sulphur, 163

1702. Oxidation of sulphites to sulphates, 164

STAS, JEAN SERVAIS, b. 1813, Louvain ; d. 1891, Brussels

1841. "Researches on the True Atomic Weight of Carbon"
by Dumas and Stas. Composition of carbonic
anhydride, 148 ; combustion of cinnamic acid, 373

1849. " New Researches on the True Atomic Weight of
Carbon." Composition of carbonic oxide, 151

1860-82. Combining weights of nitrogen, sulphur, chlorine,
bromine, iodine, lithium, sodium, potassium, lead,
silver, 315

1860. " Researches on the Reciprocal Relations of the Atomic
Weights." Combining weights of Ag, Cl, K, Na, N,
S, Pb, 316 ; law of multiple proportions verified by
Stas's analyses, 308; Prout's hypothesis an illusion, 484

1865. " New Researches on the Laws of Chemical Propor-
tions, on Atomic Weights and their Mutual Rela-
tions." Tests the law of fixed proportions, 301, and
reciprocal proportions, 317

1876-81. " Determination of the Proportional Relation between
Silver, the Chlorides, and the Bromides." Com-
bining weights of K, Na, N, Ag, &c., 316

1887. Prout's hypothesis recognised, 485

All quotations are from the complete Works (Oeuvres
Completes) published in 3 volumes in 1894.
STRUTT, HON. R. J.

1901. Prout's hypothesis and the law of probability, 485
SUTHERLAND, W. Ice, water, and steam, 527

TAMMAN, GUSTAV. Light ice and dense ice, 528
TAYLOR, PHILIPS

1812. Spirit of wood, 404

BIOGRAPHICAL INDEX 563

TENNANT, SMITHSON, b. 1761, Wensleydale ; d. 1815, Boulogne.
Professor of Chemistry at the University of Cambridge

- Carbon from fixed air by action of phosphorus on

chalk, 99

1797. Diamonds burned with saltpetre, 99
TENNANT, CHARLES

1799. Bleaching-powder, 228

THOMSON, THOMAS, b. 1773, Crieff; d. 1852, Kilmun, Argyle-
shire. Doctor of Medicine, and Regius Professor of Chemistry
at Glasgow

1 8 30- 1831. History of Chemistry, 305

1808. Multiple proportions in oxalates of potassium and

strontium, 305

1813. Atomic weights as multiples of oxygen, 482
1825. Attempt to Establish the First Principles of Chemistry
by Experiment, 483

VALENTINE, BASIL. Said to have been a monk living in the
fifteenth century, but this is very doubtful as his
works did not appear until the beginning of the
seventeenth century
1624. Triumphal Chariot of Antimony (tr. 1661)

- Last Will and Testament (tr. 1656). Preparation of

oil of vitriol, 13, 163; of "vitriols" from metals,
17; of sugar of lead and acetate of copper, 20;
caustic alkali, 54
VELEY, V. H.

1889, 1898. Properties of pure nitric acid, 533
1893. Properties of dry gases, 533

VOLTA, ALESSANDRO, b. 1745, d. 1827, Como. Professor
of Physics at Como and at Pavia

1776. Letters on the Inflammable Air of Marshes. Discovery

of marsh-gas, 156 ; explosion of marsh-gas with air,
156
1790. Volta's eudiometer, 122

1800. Voltaic pile, 271

V/ANKLYN, J. A.

1869. Sodium fused in dry chlorine, 533
WEBER, HEINRICH FRIEDRICH

1875. Atomic heat of boron, 366
WENZEL, KARL FRIEDRICH (1740-1793)

1777. Lehre von der Verwandtschaft der Korper (Theory of

Affinity}. Fails to discover law of reciprocal
proportions, 311

002

564 BIOGRAPHICAL INDEX

WILLIAMSON, ALEXANDER W., b. 1824, Wandsworth ; d. 1904,
Haslemere. Professor of Chemistry at University College,
London

1852. "Theory of Etherification," 423; mixed ethers, 424;
the water type, 424

1852. "On the Constitution of Salts." Multiple types, 425,

427, 428 ; combining power of carbonic oxide, 435
WINKLER, CLEMENS ALEXANDER. Professor of Chemistry at
Freiburg

1886. Discovery and properties of germanium, 477, 478
WOHLER, FRIEDRICH, b. 1800, nr. Frankfort; d. 1882, Gottingen.
Professor of Chemistry at Gottingen

1827. Preparation of aluminium, 288, 399

1828. Synthesis of urea, 399

1832. "On the Radical of Benzoic Acid," by Wohler and

Liebig, 399 ; benzoyl chloride, 407
1840. S. C. H. Windier on substitution without change of

type, 414

WOLLASTON, WILLIAM HYDE, b. 1766, East Dereham ; d. 1828,
London. Doctor of Medicine

8. Multiple proportions in oxalates, carbonates, and

sulphates, 306

WURTZ, KARL ADOLPH, b. 1817, Strassburg ; d. 1884, Paris.
Professor of Organic Chemistry at the Sorbonne, Paris

Chemistry a French science, 34

1849. Methylamine and ethylamine, 422

1873. Vapour-density of phosphorus pentachloride, 523

SUBJECT INDEX

Acetamide, 440
Acetate of copper, 20

of lead, 20

of lime, 20

of magnesia, 21

of soda, 20
Acetates, 20
Acetic acid, 12

composition of, 391
dissociation of, 522
formula of, 393
structure of, 430, 440
vapour-density of, 522

anhydride, 419, 432

ether, 402, 417
Acetyl compounds, 432

radical, 405, 409, 430, 431
Acid, acetic, 12, 391, 393

-air, 82

-anhydrides, 250, 419, 432

benzoic, 13, 393

boric, 288

butyric, 430

chlorosulphonic, 430

citric, 13, 391, 393

formic, 430

gallic, 393

hydriodic, 239

dissociation of, 239

hydrochloric, 218

hydrofluoric, 234

lactic, 13

Acid, mucic, 391, 393

muriatic, 14, 215

nitric, 14

nitrobenzoic, 431

oxalic, 13, 391, 393

propionic, 430

prussic, 242

-spirit of nitre, 14

succinomc, 393

sulphovinic, 402

sulphuric, 13

tartaric, 13, 391, 393

trichloracetic, 411

uric, 13
Acids, 12

action of, on alkalis, 15
on chalk, 15, 52
on metals, 16
on insoluble salts, 50

anhydrides of, 250, 419, 432

hydro-, 246

organic, composition of, 391
formulae of, 393

oxy-, 246

strength of, 21

vegetable, 12

without oxygen, 176
Aerial acid, 101
Affinity, laws of, 311
chemical, 501

tables of, 499

theory of, 311
Air absorbed during combustion

32

an elastic fluid, 320
an element, 112

565

566

SUBJECT INDEX

Air, composition of, 191

condensation of, 324

density of, 67

dephlogisticated, 40

diminished by breathing, 34
by phosphorus, 43
by rusting iron, 43
by sparking, 189

elastic force of, 322

eminently respirable, 40

factitious, 65

fixed, 50

goodness of, 74

inflammable, 66

-measure, Priestley's, 74

nitrous, 72

phlogisticated, 31

-pump, Boyle's, 32

rarefaction of, 324

spring of the, 320

vital, 40

Alchemistic Period, i
Alcohol, action of sulphuric acid
on, 154

analysis of, 387, 388

chlorination of, 408

oxidation of, 412

purification of, 387

structure of, 401

vapour-density of, 401
Aldehyde, 412, 438
Alkaline air, 82

decomposed, 257

earths as oxides, 280

decomposed by electric

current, 281
Alkalis, 3

caustic, 53, 54

mild, 53, 54

regarded as elements, 257
Allotropy, 101
Alum, 288
Alumina, 60, 288
Aluminium, 288

chloride, dissociation of, 361,

380

Alums, isomorphism of, 370
Amalgam from ammonia, 265
Amber, 6
Amide-radical, 401

Amines, primary, secondary,

and tertiary, 423
Ammonia, 57

composition of, 263
decomposed by electric cur-
rent, 265

by sparking, 258, 262
dissociation of, 517
gas collected by Priestley, 82
inflammable, 258
oxidised by calx of gold, 259
by " manganese," 258
by nitric acid, 261
properties of, 83
solid, 287
supposed to contain oxygen,

264

Ammonium, 267
amalgam, 265
carbamate, 269, 519

structure of, 430
chloride, 267

a molecular compound, 442
dissociation of, 528
hydroxide, 287
nitrate, decomposition of, 79
oxide, 287

salts, vapour-densities of, 519
Amphi-salts, 247
Analysis, qualitative, 512

quantitative, 512
Anhydrides of organic acids, 419,

432

Aniline, 433

Apparatus for calcining metals
in air over mercury (Lav-
oisier), 34

for collecting gas from red
lead and charcoal (Lavoisier) ,
96
for combining nitric oxide and

oxygen (Gay-Lussac) , 196
for combustion of carbonic

oxide (Stas), 152
of graphite and diamond

(Dumas and Stas), 149
of hydrogen and oxygen

(Cavendish), 114
of hydrogen and oxygen
(Morley), 131, 132

SUBJECT INDEX

567

Apparatus for combustion of
organic compounds, 391,
392, 395
of spirit of wine (Lavoisier),

r47

for composition of nitric oxide
(Gray), 207

of water (Dumas), 126
for decomposing water by iron

(Lavoisier), 119
for density of oxygen (Morley),

129
for expansion of gases by heat

(Gay-Lussac), 328, 329
for experiments on gases

(Hales), 85 ; (Priestley), 85
for exploding hydrogen and

oxygen (Cavendish), 115 ;

(Monge), 120
for heating mercury in air

(Lavoisier), 41
for measuring goodness of air

(Priestley), 75

for preparing fluorine (Mois-
san), 252

vitriolic acid air (Priestley),

165

for proving dissociation of sal-
ammoniac, 521

for separating gas from muri-
atic acid (Cavendish), 82
for sparking air over mercury

(Cavendish), 189
for studying gases (Mayow),

33

for weighing gases (Caven-
dish), 68
Aqua regia, 14

fortis, 14
Argon, 192
Arsenic, atomic weight of, 351,

355

Artiads. 442
Association, 527
Atomic heat of boron, 18

of carbon, 365
Atomic heats, 364

at high temperatures, 366
at low temperatures, 365,
366

Atomics heats, law of, 362
of solid elements, 365
periodicity of, 471
Atomic numbers determined by

experiment, 487, 491
(Newlands), 485, 486
Atomic symbols (Berzelius), 295

(Dalton's), 294
Atomic theory, 291
(Dalton's), 292
Atomic volume curve, 472
volumes, periodicity of, 469
weight of arsenic, 351, 355
of beryllium, 474
of carbon, 350
of chlorine, 349
of hydrogen, 347
of indium, 475
of mercury, 351
of nitrogen, 349
of oxygen, 348
of phosphorus, 351
of sulphur, 350
of uranium, 476
weights by Cannizzaro's me-
thod, 347

correction of, 473, 476
exact, 354
of argon and potassium, 476,

491
of cobalt and nickel, 476,

490

of metals, 360
of tellurium and iodine, 476
Atomicity, 433.
Atoms, 291

indestructible, 292
positions in space, 441
Aurum Fulminans, 58
Avogadro constant, 341
Avogadro's Hypothesis, 338
an approximation, 346
Azote, 1 86, 45

B

Balanced actions, 124, 498, 503,

504, 508
in gases, 510

568

SUBJECT INDEX

Balanced actions, influence of

precipitation, 504
influence of vaporisation,

508
Barium, 281

sulphate, structure of, 418
Base, 17, 25
Bases, organic, 433
Benzamide, 400, 417
Benzene, formula for, 440
Benzoic acid, 13
formula of, 393
radical of, 399
structure of, 431
Benzoyl chloride, 400, 407, 417,

418

compounds, 400, 432
radical, 399, 400
Beryllium, atomic weight of, 474
Bicarbonates, analysed by Berg-
man, 109
analysed by Cavendish, 109,

69

discovered by Black, 108
Black-lead (see Graphite), 101
Black oxide of manganese, 211
Bleaching, 228

-powder, 228
Blue vitriol, 17

Boiling-point of dry nitrogen
peroxide, 530

of nitrous anhydride, 530
Borax, 288
Boric acid, 288
Boron, 288
Boyle's Law, 325
Bromine, 240

properties of, 241
Bronze, 7

Age, 7
Burning, air diminished by, 34

-glass (Lavoisier), 100, 34;
(Mayow), 33; Priestley, 39
Butyric acid, 430

Calces, 26

metallic, reduced by charcoal,

95, 139
by hydrogen, 123

Calces, metallic, reduced by po-
tassium, 277
Calcination, 26

of mercury, 40

of tin in sealed flasks, 35
Calcined tartar, 3
Calcium, 281

chloride, 282

fluoride, 283

hydroxide, 283

oxide, 283

sulphate, 21
Calc-spar, 105
Calomel, 225

Calomel, dissociation of, 380, 529
Calx of gold, 259

of zinc reduced by charcoal,

137

Caput mortuum, 60
Carbohydrates, 390
Carbon, 98

allotropy of, 101

atomic weight of, 350

compounds, isomerism of, 441

perchloride of, 407

quadrivalent, 437, 438

skeleton, 439

tetrahedral formula for, 441
Carbonated baryta reduced by

charcoal
by iron, 140
Carbonates, 103

isomorphism of, 369

reduction of, 140, 141
Carbonic acid, 103

anhydride, 103, 251

composition of, by volume,

143

by weight, 145
decomposed by sparking,

144

dissociation of, 516
reduced by charcoal, 142
Carbonic oxide, combustion of,

by copper-oxide, 151
composition of, by weight,

151

by volume, 143
density of, 139
discovery of, 137

SUBJECT INDEX

569

Carbonic oxide from carbonic
anhydride and charcoal,
142

from smithy scale, 138
from steam and charcoal,

142
preparation from oxalic acid,

151

prepared by reducing car-
bonates, 140, 141
and oxygen do not explode

when dry, 531
Carbonyl chloride, 440
Caustic potash isolated, 55
Chalk, 5, 48

action of acids on, 52
burning of, 49
, composition of, 378
dissolved by fixed air, 103
formula of, 379
not dissolved by liquid hydro-
gen chloride, 533
Charcoal, absorption of gases by,

149

fixed air production from, 95
heated in dry oxygen, 531
purification of, 149
purified by chlorine, 219
Charles's Law, 327
Chemical affinity, laws of, 501
change, conditions of, 528
change as reversed electrolysis,

534

combination, laws of, 296

compounds, 301

equations, 389
Chile saltpetre, 19
Chloral, 408
Chlorate of potash, 227
Chloric acid and chlorates, 228
Chlorides metallic, 225
Chlorine, 212

an element, 220

as an oxide, 213, 218, 222

atomic weight of, 349

discovery of, 211

displaces sulphur and oxygen,
226

dry, does not attack sodium
or other metals, 533

Chlorine not an acid, 214

not decomposed by charcoal,

219, 221
plays the part of hydrogen,

410

properties of, 212
radical, 420
-water, decomposed by light,

213

Chlorosulphonic acid, 430
" Choke-damp," 64
Citric acid, 13

composition of, 391
formula of, 393

Classification of the elements
(Davy), 450; (Dumas),
452; (Lavoisier), 449;
(Mendeleeff), 457, 463;
(Newlands), 455
metals and non-metals ;

451

periodic, 462
Clay, 288
Coal gas, 158
Colcothar, 13, 60
Combination, chemical, 301
Combining-power, 434, 435
volumes of hydrogen and

nitrogen, 335
and oxygen, 115, 121, 127,

130. 33°
weights, 312

Combustion in dry gases, 531
of a candle, 146
of charcoal, Lavoisier's, 145
of hydrogen and oxygen, 117,

131

of olive oil, 146

of organic compounds, 388,
391, 392, 395

of phosphorus, 178

of spirit of wine, 146

oxygen and chlorine as sup-
porters of, 450

tube, Morley's, 131

with copper oxide, 394
Common differences, Dumas's,

454

Complex molecules of gaseous
elements, 338, 352

570

SUBJECT INDEX

Composition by volume of am-
monia, nitric oxide and
muriatic acid, 342
of ammonia, sulphuric acid,

carbonic anhydride, 331
of hydrogen chloride, 335
of oxides of nitrogen, 332
of sal-ammoniac and sal-
volatile, 331
of water, 341
of air, 191
of ammonia, 263
of artificial sulphide of iron,

i?4

of carbohydrates, 390
of carbonic anhydride by

volume, 143
of carbonic anhydride by

weight, 145, 148
of carbonic oxide by volume,

143

of chalk, 378
of fulminating gold, 261
of iron pyrites, 174
of lead nitrate, 379

sulphate, 314, 374, 375, 377

and sulphide, 317
of muriatic acid, 216
of nitric acid, 199, 202
of nitric oxide, 194, 207, 332
of nitrogen peroxide, 195
of nitrous oxide, 193
of oils and resins, 391
of olefiant gas, 155
of organic acids, 391
of oxides of copper, 307

of iron, 367

of lead, 307

of sulphur, 307
of potassium chlorate, 315,377
of prussic acid, 243
of sal-ammoniac, 264
of silver chloride, 354
and chlorate, 317
of silver iodate and iodide, 317
of slaked lime, 283
of sugar, starch and wood, 390
of sulphides of iron, 307
of sulphuretted hydrogen, 171
of sulphurous anhydride, 168

Composition of the alkalis, 283
of water, 120, 122, 124, 125,

127, 128, 133, 415
Compound ethers, 401
organic radicals, 398
radical, 245

acetyl, 405, 409, 430, 431

amide, 401

ammonium, 267

amyl, 422

benzoyl, 399

butyryl, 430

carbonyl, 431

cyanogen, 245, 399

ethyl, 402, 431

formyl, 430

methyl, 404, 431

methylene, 404

phenyl, 422

picramyl, 409

propionyl, 430

sulphuryl, 428, 429, 430, 439
radicals, metallic and non-
metallic, 270

Compounds, copulated, 433
heterogeneous, 421
homogeneous, 421
molecular, 441
Condensation, 418
Conjugated radicals, 430
Conservation of mass, 389
Copper, 7

gain in weight on burning, 28
native, 7

nitrate formed in an explosion-
vessel, 1 88
oxide reduced by hydrogen,

125
use of, in combustion of

organic compounds, 394
water not decomposed by, 118
Copulated compounds, 433
Copulation, 417
Coronium, 493
Corrosive sublimate, 225
Counterpoise, used by Regnault,

128

Crystal, 6

Cyanogen chloride, 243, 255.
407

SUBJECT INDEX

57i

Cyanogen gas, 245
radical, 245, 399

Density of carbonic oxide, 139
of gases, 340, 345
of hydrogen, 129, 481
of nitrogen, nitric oxide, and

oxygen, 337
of oxygen, 128

Dephlogisticated air, 40
marine acid (chlorine), 212

Diamond, 5

allotropy of, 101
combustion of, 99

Diatomic gases, 352
radicals, 433

Didymium, 303

Diminished nitrous air (nitrous
oxide), 77

Dimorphism, 381

Dissociation, 514
of acetic acid, 523
of aluminium chloride, 361,

380

of ammonia, 517
of ammonium chloride, 521
of calomel, influence of moist-
ure on, 529

of carbonic anhydride, 516
of dry nitrous anhydride, 530
of ferric chloride, 361, 380
of ferrous chloride, 361, 380
of hydriodic acid, 239
of nitrogen peroxide, 525
of sal-ammoniac, influence of

moisture on, 528
of stannous chloride, 361, 380
of water, 515, 535

Dissolution, 301

Dualistic theory, 394

Dulong and Petit's Law of
Atomic Heats, 362

Dutch Liquid, 155

Earths prepared from salts, 57
Ekaaluminium, 476

Ekaboron, 476
Ekamanganese, 478
Ekasilicon, 476
Elastic fluids, 71
Electric battery, 272, 273

in dissolution of metals, 534
current, 271

decomposes alkaline earths,

287

ammonia, 265
potash, 273
salts, 273
soda, 278
water, 272

Electro-chemical theory, 397
Electrolysis, chemical change as

reversed, 534
Elements, 112

and compounds, 221
Boyle's views on, 448
classification of (Davy), 450;
(Dumas), 452 ; (Lavoisier),
449 ; (Mendeleeff), 457, 465 ;
(Newlands), 455
four, of Aristotle, 162, 448
molecular complexity of, 338,

352

weights of, 352
symbols for, 295, 296
Elixir of life, 2
Emerald, 6
Empirical formulae, 373

of salts, 374
Epsom salts, 59, 19
Equivalent of chlorine, 316
of potassium, 316
of silver, 316
Equivalents, 310, 311, 312, 313;

314, 315

and atomic weights, 314
determined by Stas, 315
of acids and bases, 311
relations between, 454
table of, 313

Ethane, 431

Ether, acetic, 402
analysis of, 387, 388
chlorination of, 410
hydrochloric, 402
preparation of, 387

572

SUBJECT INDEX

Ether, structure of, 401

sulphuric, 402

vapour-density of, 401
Etherification, 423
Ethers, compound, 401

mixed, 424
Ethylamine, 422
Ethyl chloride, 402, 427, 432,

44°

compounds, 403, 432
radical, 402
Ethylene, 155
chloride, 155

chlorination of, 407
preparation from alcohol, 387
Eudiometer, 75

Priestley's, 90 ; Volta's, 122,

123

Even series, 460

Expansion of gases by heat, 325,
327

F

Fats, 6

Fermentation, vinous, 388

Ferric chloride, dissociation of,

361, 380
Ferrous chloride, dissociation of,

361, 380

Finery cinder, reduced by char-
coal, 138
Fire-air, 44
" Fire-damp," 64
Fixed air, 50

a product of combustion, 94
an acid, 101
an oxide of carbon, 97
density of, 67

dissolves chalk and mag-
nesia, 103
dissolves iron, 107
dissolves zinc, 103
from burning charcoal, 95
from combustion of dia-
mond, 99
of graphite, 101
from mercuric oxide and

charcoal, 97

from red-lead and charcoal,
96

Fixed air from reduction of

calces, 95
in chalk, 50
in common air, 51, 94
in magnesia, 94
in mild alkalis, 53, 94
in sal-volatile, 56, 94
in water, 51, 94
not produced by combustion

of hydrogen, 116
produced by breathing, 385
by burning charcoal, 386
by fermentation, 386
properties of, 94, 385, 386
quantity of, in various com-
pounds, 69
reddens litmus, 102
solubility of, 70
Fixed proportions, 296

as a test of chemical com-
bination, 300
Flame in vacuo, 30
Fluids, elastic, 71
Fluorine, 233

isolation of, 252
Fluor-spar, 233

crystals of, 234
Formic acid, 430
Formula of chalk, 379
of cinnamic acid, 373
of lead nitrate, 380
of lead oxide, sulphide and

sulphate, 374

of potassium chlorate, 378
Formulae, empirical, 373
graphic, 439
molecular, 373
of salts, 374

of volatile compounds, 352
structural, 439
Fuller's earth, 5
Fulminating gold, 58, 260
composition of, 263

Galena, crystal of, 173
Gallic acid, formula of, 393
Gallium, 477
Garnets, isomorphism of, 370

SUBJECT INDEX

573

Gas pingue, 64
sylvestre, 64

Gases, balanced actions in,
Boyle's Law, 325
Charles's Law, 327
collected over water, 65
density of, 340, 345
expansion by heat of, 325, 327
Hales' apparatus for experi-
ments with, 84
kinetic theory of, 324
Mayow's apparatus for collect-
ing, 33, 84

method of weighing, 68
Priestley's apparatus for work-
ing with, 86
work on, 70
properties of, 320
specific heats of, 353
stored over mercury, 70

Gay-Lussac's Law of Volumes,

334

Generator for gases, 84
Germanium, 478
Glass, manufacture of, 6
Glauber's salt, 18
Gold, 7

fulminating, 58, 260
Graphic formulae, 439
Graphite, combustion of, 101

allotropy of. 101
Green vitriol, 9, 17
Gum, composition of, 390

formula of, 393
Gun-barrel, 88, 118
Gypsum, 16, 18, 19

II

Halogens, 247
Haloid salts, 247
Hardness of water, 105

permanent, 107

temporary, 106
Heat capacity of gases, 353
Hepatic gas, 171
Hot-cold tube (Deville's), 516,

5i7

Hydracids, 247
Hydrides, 227

Hydriodic acid, 239

properties of, 240
Hydrocarbon, 155
Hydrochloric acid and chlorides,
218

displaced by oxalic acid, 509
Hydrocyanic acid and cyanides,

245

Hydrofluoric acid, 234, 235
Hydrogen, 117

and oxygen do not explode

when dry, 531
as standard of atomic and

molecular weights, 343
atomic weight of, 347
chloride, 218

composition by volume, 222
decomposed by sparking

over mercury, 224
gas collected by, Priestley,

82

liquid, properties of, 533
properties of, 82
density of, 481, 129
radical, 420
removed without replacement,

407

theory of acids,
Hydrol, 527
Hydro-salts, 264
Hypochlorous acid and hypo-

chlorites, 228
Hypoiodous acid, 254

Ice, dense, 528

light, 528

-molecules, 527

Indestructibility of matter, 389
Indicators, 15

Indium, atomic weight of, 475
Inflammable air, 66

density of, 67

Inorganic chemistry, develop-
ment of, 383

types, 422
lodic acid, 254

anhydride, 254
Iodide of nitrogen, 239

574

SUBJECT INDEX

Iodides, metallic, 238
Iodine, 237

an element, 237

chloride, 238

discovery of, 235

oxide, 238

properties of, 236
Iron, 7

Age, 7

dissolved by fixed air, 107

gain in weight on burning, 28

pure, does not act on copper

sulphate, 533
does not dissolve in acids,

533

does not rust, 533
smelting of, 7
water decomposed by, 118
Isomeric change, 399
Isomorphism, 368

complete and incomplete, 381
of alums, 370
of carbonates, 369
of garnets, 370

of manganates, chromates,
selenates and sulphates, 371
of perchlorates and perman-
ganates, 371
of phosphates and arsenates,

368

of spinels, 382
perfect and imperfect, 382
Isomorphous mixtures, 370

K

Kinetic Theory of Gases, 324
Kopp's Law of Molecular Heats,
367

Lactic acid, 13

Laughing gas, 77

Law of fixed proportions, 296,

297

accuracy of, 301
of isomorphism, 369
applications of, 371
exceptions to, 372

Law of multiple proportions, 303

accuracy of, 308
of octaves, 456
of periodicity, 458
of reciprocal proportions, 309

accuracy of, 316
of volumes, 334

an approximation, 334

rejected by Dalton, 335

Laws of chemical combination,

296
Lead, 8

atomic weight of, 303
calcined, 27

in air over water or mer-
cury, 34
nitrate, composition of, 379

formula of, 380
sulphate, composition of, 314,

374. 375. 377
and sulphide, composition

of, 317

formulae of, 377
Lime, 7, 48

an oxide, 283

dry, properties of, 529, 533
Limestone, 5, 7, 48
Lime-water, 49
Litharge, 8
Litmus, 15

turned red by fixed air, 102
Liver of sulphur, 170
Long periods, 460
Lunar caustic, 19

M

Magdeburg hemispheres, 320
Magnesia, 19

dissolved by fixed air, 103
Magnesium, 281
Malic acid, 13
Manganese, 210
Marble, 5
Marcasite, 8

crystals of, 8

nodules of, 9

Marsh gas, composition of, 157
decomposed by sparking,
157

SUBJECT INDEX

575

Marsh gas, discovery of, 156
properties of, 156
structure of, 440
type, 437

Mass action, 501, 502, 504, 511
Matter indestructible, 389
Mercury, 7

atomic weight of, 351

calcination of, 40

oxide heated alone, and with

charcoal, 97
red oxide of, 39

trough, 88 ; Lavoisier's, 147
Metalepsy, 406
Metalloids, 451
Metals, action of nitric acid on,

79

and non-metals, 451
burning of, 26
not attacked by dry chlorine,

533
not dissolved by dry acids,

specific and atomic heats of,

363, 472

Methyl alcohol, 403, 405

Methylamine, 422

Methyl compounds, 405, 432
radical, 404

Mice, 12, 39, 86, 87

Milk of sulphur, 171

Milk-sugar, formula of, 393

Minium, 8

Missing elements (Mendeleeff),

476 ; (Newlands), 455
total number of, 492

Mixed types, 429

Moisture, influence of on dis-
sociation, on combustion, and
on chemical change, 528 — 533

Molecular compounds, 441
formulae, 373

heat of nitrogen peroxide, 526
heats of compounds, 367
weights determined by Avo-

gadro's hypothesis, 340
exact, 356

Monatomic gases, 352, 354
radicals, 433

Mortar, 6

Mucic acid, composition of, 391

formula of, 393
Multiple proportions in car-
bonates

in gases, 304, 305, 306
in oxalates, 305, 306
in oxides of carbon, 308
of copper, iron, lead,

sulphur, 307
in sulphates, 306
in sulphides of iron, 304, 307
types, 427

Muriates of copper, gold, iron,
lime, magnesia, potash, soda,
20
Muriatic acid, 14

composition of, 216
contains hydrogen, 215
radical, 218, 222, 225

N

Natural series, Dumas's, 453

Nebulium, 493

Neodymium and praseodym-
ium, 303

Neumann's Law of Molecular
Heats, 367

Nitrate of potash, 19
of silver, 19
of soda, 19

Nitrates, 199

Nitre, 5, 19

acid spirit of, 14
decomposed by heat, 198
Egyptian, action of vinegar

on, 5
mother of, 58

Nitric aeid, 14, 73, 198, 199
analysis of, 200
composition of, 199, 202,

379

displaced by vitriol, 22, 498
formed during explosion of

hydrogen and oxygen,

116, 188
produced by sparking air,

1 88

pure, properties of, 533
structure of, 439

576

SUBJECT INDEX

Nitric acid, synthesis of, 200
oxide, absorbed by ferrous

salts, 72
and oxygen do not unite

when dry, 530
combination with oxygen,

201

composition of, 194, 207, 332
density of, 520
properties of, 72
Nitrites, 199
Nitro-aerial spirit, 32
Nitrobenzene, 417
Nitrobenzoic acid, 431
Nitrogen, 45, 121

atomic weight of, 349
from fulminating gold, 261
iodide, 239
peroxide, 73, 197
composition of, 195
dissociation of, 525
dry, boiling-point of, 530
molecular heat of, 526
properties of, 75
solid, 525

vapour-density of, 525
Nitrous acid, 73, 199

gas, 73
Nitrous air (see also Nitric

oxide), 72
combination of oxygen

with, 73
diminished (nitrous oxide),

77

Nitrous anhydride, dry, boil-
ing-point of, 530
dry, properties of, 530
fumes (nitrogen peroxide), 73
oxide as an anaesthetic, 80, 81
composition of, 193
from ammonium nitrate, 79
properties of, 78, 79
Noble metals, 8
Nomenclature chemical, 17 ;
ammonia, 57; ammonium,
267 ; anhydrides, 250 ; bar-
ium, 281 ; bromine, 240 ;
calcium, 281 ; carbon, 98 ;
carbonic acid and carbonates,
103 ; carbonic oxide, 140 ;

chloric acid and chlorates,
228 ; chlorides, 218 ; chlorine,
212 ; crystals, 6 ; cyanides,
245 ; cyanogen, 399 ; fluorine,
233 > hydriodic acid, 239 ;
hydrochloric acid, 218 ; hy-
drocyanic acid, 245 ; hydro-
fluoric acid, 235 ; hydrogen,
117 ; iodides, 238 ; iodine,
237 ; magnesium, 281 ; ni-
trates, 199 ; nitric acid, 14,
198, 199 ; nitric oxide, 194 ;
nitrites, 199 ; nitrogen, 45,
191 ; nitrogen peroxide, 73,
197 ; nitrous acid, 199 ; ni-
trons oxide, 193 ; oxygen, 42 ;
phosphoric acid and phos-
phates, 179 ; phosphorous
acid and phosphites, 179 ;
potassium, 275 ; prussic acid
and prussiates, 243 ; salts, 16 ;
sodium, 278 ; strontium, 281 ;
sulphides and sulphurets, 172 ;
sulphuric acids and sulphates,
14, 163 ; sulphurous acid and
sulphites, 164 ; see also under
Compound radicals.

Nordhausen oil of vitriol, 13

Nuclei, derived, 416
fundamental, 416

O

Octaves, Newlands's, 456
Odd series, 460
Oils, 6

composition of, 391
of bitter almonds, 399
of vitriol, 9, 13

contains sulphur, 163
displaced by nitric acid, 498
sulphur in, 14
Olefiant gas, 155

composition of, 155
decomposed by sparking,

156

discovery of, 154
preparation from alcohol,

387
Ordinals, 493

SUBJECT INDEX

577

Organic analyses, alcohol and

ether, 388
of Berzelius, 392, Liebig,

395 ; Gay-Lussac 390
bodies containing metals,

435

chemistry, rise and develop-
ment of, 384

compounds, carbon and hy-
drogen in, 385
formulae of, 391
Organo-metallic compounds, 435
Orthrin, 401
Oxalic acid, 13

composition of, 391
displaces hydrochloric acid,

509

formula of, 393
Oxidation, 124
Oxide of iodine, 238
Oxides, 43

of nitrogen, composition of,

204

reduced by potassium, 277
Oxyacids, 247
Oxygen, 42

as standard of molecular and

atomic weights,
atomic weight of, 348
density of, 128
discovery of, 38, 43
displaced by chlorine, 226
from nitre, 44
from nitric acid, 44
from red lead, 39
from red oxide of mercury,

39, 44
Oxymuriatic acid (chlorine),

215
Oxy-salts, 264

Palladium tube, 130
Pearl-ashes, 3, 68
Pearls, 6

Periodic acid, 254
Periodic classification, 462

Law, 459
Periodicity, 457, 458

of atomic heats, 471

HIST. CHEM.

Periodicity of atomic volumes,

469

of valency, 468
Peroxide of manganese, 211
of potassium, 283
of sodium, 283
Peroxides, 211
Perissads, 442
Philosopher's stone, 2
Phlogisticated acid of nitre, 198 ;

air, 31

Phlogiston, 31

Phosphates and arsenates, iso-
morphism of, 368
Phosphonium iodide, structure

of, 433

salts, vapour-densities of, 519
Phosphoric acid, and phosphates,

179

displaces sulphuric acid, 509
Phosphoretted hydrogen, 180
from phosphorous acid, 181
not an acid, 181
Phosphoric anhydride, 179
Phosphorous acid and phos-
phites, 1 80
Phosphorus, atomic weight of,

35i

burning of, 178
distilled in dry oxygen, 531
preparation of, 177
properties of, 178
smouldering of, 179
supposed to contain hydrogen,

265
pentachloride a molecular

compound, 442
dissociation of, 523
vapour- density of, 523
Picramyl radical, 409
Plaster of Paris, 16
Plumbago (see Graphite), 101
Polyatomic radicals, 433
Polymerisation, 351
Polymorphism, 372
Polysulphides, 177
Potash, 3

action of heat on, 55
as an element, 257
caustic, isolated, 55

P P

578

SUBJECT INDEX

Potash decomposed by electric

current, 273
by iron, 275
Potassium, 275
amalgam, 276
chlorate, 227

composition of, 315, 377
formula of, 378
chloride, 276
decomposes water, 277
ethoxide, 423
hydroxide, 284
iodate, 254
iodide, 239
oxidation of, 275
properties of, 2.75
Prehistoric Period, I
Principles, three alchemistic, 162,

448

Producer gas, 142
Prom, 400
Propionic acid, 430
Protyl, 482
Prout's hypothesis, 480

and the law of probabilities,

485

tested, 483, 484
Prussian blue, 242
Prussiates, 243
Prussic acid, 242
anhydrous, 24
composition of, 243
structure of, 440
Pyrites, 8

composition of, 174
contains sulphur, 163
crystals of, 9
oxidation of, 174
sulphur in, 14
vitriolisation of, 174
Pyrolusite, 210

Quadrivalency of carbon, 437,

438
Quartz, 288

crystals of, 6
Quicklime. 49

R.

Radicals, compound, 245

conjugated, 430

derived, 416

diatomic, 433

fundamental, 416

monatomic, 433

polyatomic, 433
Radio-active elements, 466
Radio-elements, 466
Rare-earth elements, 465, 492
Rathbone Place water, 103
Receiver for gases, 84, 88
Reciprocal proportions, 309
in acids and bases, 310
Red calx of mercury, 39

lead, 8

precipitate, 60, 44
Reduction, 124
Residues, 417
Resin, composition of, 391
Reversible actions, 499, 5OI> I24
Rock-crystal, 6
Rouge, 13

Sal-ammoniac, 5

composition of, 264
crystal of, 4
dissociation of, 521
dry, properties of, 529, 533
influence of moisture on the

dissociation of, 528
prepared from acid air and

alkaline air, 83
volatile spirit of, 57
" Sal mirabile," 18
Sal sylvii, 20
Sal-volatile, 56

prepared from alkaline air

and fixed air, 84
Salt, 3

Boyle's definition of, 3
common, 3
Epsom, 59
Glauber's, 18
sedative, 288
spirit of, 14

SUBJECT INDEX

579

Saltpetre, 5, 8, 19

crystal of, 4

Salts as binary compounds of
two radicals, 270

crystallisation of mixed, 505,

5ii

decomposed by electric cur-
rent, 273

formulae of, 374

insoluble, 1 7

insoluble, action of acids on,

507

neutral, 25

nomenclature of, 16

preparation of, 16

structure of, 394
Sand, 288
Scandium, 477
Sceptical chemist, 448
Sedative salt, 288
Selenite, 16, 18, 19
Semi- water gas, 143
Short periods, 459
Silica, 288
Silicon, 288
Silver, 7

chlorate and chloride, com-
position of, 317

chloride, composition of, 354

iodate and iodide, composition

of> 317
Slaked lime, 49

composition of 283
Smithy scale, reduced by char-
coal, 138
Soda, 3

decomposed by electric cur-
rent, 278

Sodium-amalgam, 279
chloride, 286
decomposes water, 279
hydroxide, 284
properties of, 278
sulphate, 21
Solid solutions, 370
Solutions, solid, 370
Specific heats, 364

at high temperatures, 366
at low temperatures, 365,
366,472

Specific heats of metals, 363

of solid elements, 365
Spectra, 475

high-frequency, of the ele-
ments, 487, 488, 489
Spirit of hartshorn, 57

of salt, 14

of wood, 403
Spiritus nitro-^neus, 32
Spring of the air, 320
Stalactites, 105, 106
Stalagmites, 105
Stannous chloride, dissociation

of, 361, 380
Starch, composition of, 390

formula of, 393
Steam-molecules, 527
Strontium, 281
Structural formulae, 439
Structure of alcoholand ether, 401

of benzene, 440

of organic compounds, 398

of salts, 394
Substitution, 406

-value, 433

Succinic acid, formula of, 393
Sugar, 6

composition of, 390

formula of, 393

of lead, 20
Sulphate of copper, 18

of iron, 18

of lime, 19

of magnesia, 19

of potash, 1 8

of soda, 18
Sulphates, 17

decomposition of, by heat, 169
Sulphides, burning of, 173

conversion to sulphates, 174

metallic, 172
Sulpho-salts, 248
Sulphovinic acid, 402
Sulphur, association and disso-
ciation of, 351, 524

atomic weight of, 350

combustion of, 163

crystals of, 163

displaced by chlorine, 226

dissociation of, 524

58o

SUBJECT INDEX

Sulphur dissolved by alkalis, 171
distilled in dry oxygen, 531
from pyrites, 162
in oil of vitriol, 163
liver of, 170
milk of, 171
native, 162
polymerisation of, 351
supposed to contain hydro-
gen, 265
Sulphurets, 172
Sulphuretted hydrogen, 171
an acid, 175
composition of, 171
decomposed by sparking,

172
Sulphuric acid, action on alcohol,

154

and sulphates, 163
composition of, 168, 169,

375. 376

displaced by nitric acid, 498
displaced by phosphoric

acid and by silica, 509
structure of, 439
anhydride, 169, 250
ether, 402
Sulphurous acid, and sulphites,

164

structure of, 430
anhydride, 250

composition of,
Sulphuryl chloride, 439
Supporters of combustion, 451
Sylvine, 20
Synthesis of organic compounds,

399
Syrup of violets, 15

Tables of affinity, 499
Tannin, formula of, 393
Tartar, 3
calcined, 3
vitriolated, 18
Tartaric acid, 13

composition of, 391
formula of, 393
Theory of types, 411

Tin, 7

calcined, 26

in air, over water or mer-
cury, 34

in sealed flasks, 29, 35
Tinstone, 7
Topaz, 6

" Tournesol " (see litmus), 15
Transition periods, 460
Triads, Dobereiner's, 452
Triatomic gases, 352
Trichloracetic acid, 411
Trimanganese, 478
Turpentine, 6
Types, mixed, 429
multiple, 427
simple inorganic, 421

ammonia, 422

hydrochloric acid, 425

hydrogen, 425

marsh gas, 437

water, 424
theory of, 411
Typical elements, 460

U

Uranium, atomic weight of, 475
Urea, synthesis of, 399
Uric acid, 13

V

Valency, 355, 434

fixed or variable, 441

of carbon, 437, 438

periodicity of, 468
Vapour densities, 340, 345,

354
of ammonium and phos-

phonium salts, 519
of compounds of iron, 362
of metallic elements and

compounds, 360, 361
density of acetic acid,
of nitrogen peroxide, 525
of phosphorus penta-

chloride,
of sulphur, 524

SUBJECT INDEX

581

Vapour-density, variation with

temperature, 522
Verdigris, 16

Vicarious replacement, 370
Vinegar, 6, 8
distilled, 12

" Vinegar upon nitre," 4
Vinous fermentation, 388
Vital air, 40
Vitriol, blue, 17, 1 8
green, 9, 17, 18
Nordhausen oil of, 13
oil of, 9, 13

displaced by nitric acid, 498
Vitriolated tartar, 18

decomposed by nitric acid,

498
Vitriolic acid air (sulphurous

anhydride), 164
converted into oil of vitriol,
167

preparation of, 165
properties of, 166
Vitriols, 17

preparation of, 18
Volatilespiritofsal-ammoniac,57
Voltaic pile, 271
Volta's eudiometer, 122

W

I Water absorbed and liberated
in the formation of salts, 269
a product of combustion, 113
an element, 112
an oxide, 117
association of, 527

Water by reduction of metallic
calces, 123

composition of, Berzelius and
Dulong, 124 ; Cavendish,
115; Dumas, 125; Humboldt
andGay-Lussac, 122; Monge,
120; Morley, 128; Scott, 127

decomposed by charcoal, 142
by electric current, 272
by iron, 118
by potassium, 277
by sodium, 279

dissociation of, 515

-gas 143

hardness of, 105

-molecules, 527

of crystallisation, 25

structure of, 439

type, 424
Waters, aerated, 102

chalybeate, 108

mineral, 102
Waxes, 6

Willow- tree, van Helmont's, 112
Wine, 6, 8

X-rays, wave-lengths of, 487,
488, 489

Zinc oxide reduced by charcoal,

T37

Zinc, pure, does not dissolve in
acids, 530

PKINTEI) IN GREAT BRITAIN BY

RICHARD CLAY AND SONS, LIMITED,

BRUNSWICK STREET, STAMFORD STREET, S.E.j
AND BUNGAY, SUFFOLK.

LOAN DEPT.

STAV;

o u) LD

rt!

THE UNIVERSITY OF CALIFORNIA LIBRARY