Full text of "Chemical reports and memoirs ..."

See other formats

.-/'-<.

•

3 — qr-: *<z*" " r

ESQ'S

THE UNIVERSITY OF
BRITISH COLUMBIA

Presented by

H. R. MacMillan
1965

WORKS

OF THE

CAYENDISH SOCIETY.

FOUNDED 1846.

CHEMICAL

REPORTS AND MEMOIRS,

ON ATOMIC VOLUME; ISOMORPHISM; ENDOSMOSIS;

THE SIMULTANEOUS CONTRAST OF COLOURS;

THE LATENT HEAT OF STEAM AT DIFFERENT PRESSURES;

THE ARTIFICIAL FORMATION OF ALKALOIDS;

AND VOLCANIC PHENOMENA.

EDITED BY

THOMAS GRAHAM, V.P.R.S.,

CORRESPONDING MEMBER OF THE INSTITUTE OF FRANCE j
PROFESSOR OF CHEMISTRY IN UNIVERSITY COLLEGE, LONDON.

LONDON:
PRINTED FOR THE CAVENDISH SOCIETY,

T. R. HARRISON, ST. MARTIN'S LANE.

1848.

P R E F A C E

The First and Fourth Reports on Atomic Volume and Iso-
morphism, by Professor Otto, of Brunswick, contained in the
present volume, are extracted from the second edition of that
author's systematic work on Chemistry, now in course of
publication*. They contain, I believe, a mere detailed account
of the researches and speculations of H. Kopp, Schroder,
Lowig, and others, than has hitherto been given to the public.
The imperfect but advancing state of these branches renders
their critical discussion useful and opportune, and incites to
farther inquiry. The more recent memoir of M. Filhol, trans-
lated from the Annales de Chimie et de Physique, t. xxi (Sme
serie), has been added, as it contains a series of new deter-
minations of the specific gravity of many substances, by which
he is enabled to test M. Kopp's theoretical conclusions, and
to modify them in several points.

The Report on Endosmosis, by Dr. Julius Yogel of Giessen,
is the translation of a tract published separately by him, under
the title of the report. The attention of chemists and physio-
logists has lately been recalled to this subject by the researches
of Liebig on the Motion of the Juices in the Animal Body.

The Physical Investigations on Dyeing, by M. Chevreul,
which form the fifth Memoir, were read before the French
Academy, and published in the Revue Scientifique et Industrielle,
t. xii., 1847, from which they have been extracted.

The valuable memoir of M. Regnault^ On the Latent Heat
of Steam at different Pressures, is a portion of a series of
investigations undertaken by that chemist, at the instance of

* Lehrbuch der Chemie. Zum Theil auf Grundlagfe von Dr. Thomas
Graham's Elements of Chemistry, bearbeitot von Dr. Fr. Jul. Otto, Ordent-
liehem Professor der Chemie am Collegio Carol i no zu Qraunschwei

VI PREFACE.

the French Government, to elucidate the theory of the Steam-
Engine. It is extracted from a volume entitled, Relation des
Experiences entreprises par ordre de Monsieur le Ministre des
Travaux Publics, et sur la proposition de la Commission Centrale
des Machines a Vapeur, pour determiner les principales his et
les donnees numeriques qui entrent dans le calcul des Machines a
Vapeur. Par M. V. Regnault, Ingenieur au Corps Royal des
Mines, Membre de VAcademie des Sciences. Paris, Firmin-Didot,
1847. This is a volume of the Memoirs of the French Academy,
which has been published separately, and is accompanied by an
Atlas of magnificent plates.

The report by Professor E. Kop, of Strasburg, on the
Artificial Formation of Alkaloids, is also drawn from the Revue
Scientifique et Industrielle, t. xi., p. 273.

The memoir by Professor Bunsen, on the Pseudo- Volcanic
Phenomena of Iceland, was originally published in Liebig's
Annalen der Chemie und Pharmacie, bd. lxii., 1847. It con-
tains the new theory of the Irruptions of the Geyser Springs,
established by M. Descloizeaux and the Author, with many
interesting observations on the metamorphism of the palagonite
rock, and other chemical changes observed in course of progress
at this focus of volcanic activity.

The translation of all these papers has been executed by
Dr. G. E. Day.

London, July, 1848. T. G.

TABLE OF CONTENTS.

I. REPORT.

Page

On the Relation of the Volumes of Bodies in the Solid State to their
Equivalents or Atomic Weights. By Professor Otto, of Bruns-
wick 1

II. MEMOIR.

Studies on the Connection existing between the Atomic Weights, Crys-
talline Form and Density of Bodies. By M. Filhol 63

III. REPORT.

On the Laws, according to which the Mixing of Fluids and their
Penetration into Permeable Substances occur, with special refer-
ence to the Processes in the Human and Animal Organisation. By
Julius Vogel 85

IV. REPORT.

Upon Isomorphism. By Professor Otto, of Brunswick 118

V. MEMOIR.

Physical Investigations on Dyeing : On the Influence that two Colours
may exercise upon each other when seen simultaneously. By M.
Chevreul 165

VI. MEMOIR.

On the Latent Heat of Steam at different Pressures. By M. Regnault 240

VII. REPORT.

On the Artificial Formation of Alkaloids. By M. E. Kof 297

VIII. MEMOIR.

On the Intimate Connection existing between the Pseudo- Volcanic

Phenomena of Iceland. By Professor Buksen, of Marburg 323

REPORT.

ON THE RELATION OF THE VOLUMES OF BODIES IN

THE SOLID STATE TO THEIR EQUIVALENTS

OR ATOMIC WEIGHTS.

By Professor Otto, of Brunswick.

Chemists fully comprehend the nature of the relation which
exists between the equivalent of a substance and its volume
in the gaseous state. Equal volumes of different gases or
vapours do not always correspond to the same number of
equivalents; or, if we may use the expression, the equivalents
of different substances, when gasified, do not all fill an equally
large space, and hence the equivalent volumes of gases and
vapours are of different size. These gaseous volumes, however,
always bear to each other a simple ratio, such as 1 to 2, 1 to 4,
&c., and the difference is subject to known laws.

If we assume that the equivalent weights correctly repre-
sent the weight of the atoms of bodies, what has been said
of the volume of the equivalents will apply to the volume of
atoms in the gaseous state.

We are now in a position to consider whether the equiva-
lents or atoms of bodies in the solid and fluid state occupy equal
spaces, or spaces of different dimensions; whether the equivalent
volumes or atom volumes of solid bodies are of the same or
different sizes. The equivalent weights give the relations of
weight in which bodies unite to form chemical compounds. If,
now, we know the specific weights of bodies, we may determine
the relation of volume in which they combine: for we have
only to divide the equivalent weights by the specific weights
in order to obtain a quotient expressing the relations of their
volumes. Sulphur and zinc, for instance, combine chemically
In the proportion by weight of 200 of the former to 406 of the
latter; these numbers express the relation of their equivalent

2 THE VOLUMES OF BODIES

weights. The specific weight of sulphur is 1*99, and that of
zinc is 6*95; the relation of volume of these quantities of sul-
phur and zinc is therefore -f|| to -~^, or as 100 to 58; that
is to say, 100 volumes of sulphur combine with 58 of zinc.
Hence the quotients obtained by dividing the equivalent
weights by the specific gravities express the relative volumes
which the equivalents of bodies in the solid or fluid condition
occupy; they are the equivalent volumes of solid and fluid
bodies. If a designates the equivalent weight, and s the spe-

cific weight, then the equivalent volume is equal to -. To this

quotient Kopp applies the term specific volume, since it is a
relative number, like the specific gravity. But as the equiva-
lent weights are relative numbers, and we cannot alter their
designation to specific combining weights, I regard the" term
equivalent volume as a more suitable one, especially as it has been
adopted in a corresponding sense for gases and vapours. If»
instead of equivalents, we speak of atoms, the term equivalent
volume becomes naturally converted into atom volume, a word
which I first find used by Dumas, in this sense, in his Traite de
Chimie appliquee aux Arts. Schroder, who, together with Kopp,
has devoted much attention to this subject, has recently, at the
suggestion of Berzelius, employed the term molecular volume in
place of equivalent volume. In the illustration we have given,
100 is the equivalent volume of sulphur, and 58 that of zinc.
The equivalent weight of nitrate of silver, Ag O, N 05, being
2125, and its specific weight 4*36, its equivalent volume is,
therefore, *±.\\ or 478.

It will be more clearly seen what we are to understand by
the numbers indicating the equivalent volumes, if we regard
the equivalent weights as numbers of grammes. The numbers
200 for sulphur, 406 for zinc, and 1350 for silver, then indicate
that 406 grammes of zinc and 1350 of silver combine chemically
with 200 grammes of sulphur. If we consider that all state-
ments regarding specific weights, both of solid and fluid bodies,
are calculated on the assumption that that of water is unity,
and that 1 cubic centimeter of water weighs 1 gramme, then it
follows that the specific weights of bodies express the numbers of

AND THEIR ATOMIC WEIGHTS. 3

grammes that 1 c. c. of the body weighs. Thus, the numbers
1*99 and 6*95, the specific weights of sulphur and zinc, indicate
that 1 c. c. of sulphur and 1 c. c. of zinc weigh respectively
1*99 and 6*95 grammes. Now since the equivalent volumes
obtained on dividing the equivalent weight supposed to be
expressed in grammes by the specific gravity (the weight of
1 c. c. expressed in grammes), it follows that the quotient, the
equivalent volume, indicates the number of cubic centimeters
occupied by that number of grammes of the body which ex-
presses its equivalent weight. If we know that the equivalent
weight and equivalent volume of sulphur are respectively 200
and 100, and that for zinc they are 406 and 58, then we can
readily understand that 200 grammes of sulphur occupy a space
of 100 c. c; and 406 grammes of zinc a space of 58 c. c.

It is clear that the numbers for the equivalent volumes
must differ according as, in reckoning the equivalents, we use
those in which the equivalent of oxygen is assumed to be 100,
or those in which the equivalent of hydrogen is regarded as 1,
and therefore that of oxygen 8 ; but the relation of the equi-
valent volumes remains the same, exactly as the relation of
the equivalent weight is unaffected, whether we take oxygen
or hydrogen as our unit. Moreover it is obvious that those
chemists who in individual cases make a difference between
the equivalent weight and the atom weight, admitting a parti-
tion of the equivalent into two atoms, have also in these cases
to make a difference between the equivalent volume and the
atom volume. The equivalent and equivalent atom of iodine
(I) is 1585, the specific weight is 4*93; hence we calculate the
equivalent volume as \5g^f or 320. The volume atom of iodine
deduced from the density of its vapour (VI), and the atom deduced
from the specific heat of iodine (SI), both weigh, however, only
half as much as the equivalent, 792*5; the vatom volume or
satom volume amounts therefore to 749 g 35 , or 160, and is conse-
quently only half as great as the equivalent atom volume.

The accuracy of the numbers expressing the equivalent
volumes is naturally dependent on the accuracy with which
the equivalent weights and specific weights are determined.
When we consider that the former weights are only close
approximations, and that regarding the latter we often meet

B 2

THE VOLUMES OF BODIES

with very discordant statements, we cannot help acknowledg-
ing that these numbers are always to be regarded as only
approximating in some measure to the truth, and thus it be-
comes obvious how it is that different chemists have assigned
different numbers for one and the same body. The following
table contains in the first column the names of the various
elements (and of cyanogen) for which the equivalent volume
is calculated from the equivalent weights arranged in the
second column, and those specific weights which Kopp regards
as most accurately determined, in the third column. The
fourth column gives the result of this calculation, or the equi-
valent volume, in integral numbers.

Name.

Equivalen
Weight.

Antimony

1613

Arsenic

937'5

Bismuth

2660

Bromine

1000

Cadmium

696-7

Chlorine

443-2

Chromium ....

328-5

Cobalt

368-5

Copper

396

Cyanogen

325

Gold

2458

Iodine

1585-5

Iridium

1232

Iron

350

Lead

1294

Mercury

1250

Molybdenum

596

Nickel

369-3

Osmium

1242-5

Palladium

665-5

Phosphorus ....

392

Platinum

1232

Potassium ....

489

Rhodium

652

Selenium

495

Silver

1350

Sodium

290

Sulphur

200

Tellurium

802

Tin

735-3

Titanium

301-5

Tungsten

1188

Zinc

406

Specific
Weight.

Equivalent
Volume.

240

160 *

270

326

81
320

44
315
128
320

45
114

57
222

57
583

57
115
130
292
100
128
101

AXD THEIR ATOMIC WEIGHTS.

From this table we at once observe that the equivalent
volumes of the elements differ from one another to a remarkable
degree; the equivalent volume of iron is 45, that of iodine 320,
and that of potassium 583. On a closer examination we find
that the equivalent volumes of several elements are either of
the same magnitude, or stand in a simple relation to one
another. We may select the following groups from the
above table.

Cobalt ....
Copper ....
Iron

Manganese
Nickel ....

Indium ....

Osmium

Palladium

Platinum

Titanium

Zinc

Chromium

Molybdenum

Tungsten

Equiv. vol.

Selenium

44
44

Sulphur

44
44

Gold

Silver

57
57

Bromine
Chlorine

Cyanogen

Iodine

Potassium

Sodium

Equiv. vol.
115

100

128
130

326
320
315
320

292
583

It may be at once remarked that these groups in general
embrace elements which yield isomorphous combinations.

Iron may, for instance, be replaced in a compound by an
equivalent quantity (by weight) of manganese, without any
change of form occurring to the compound, since the substi-
tuted quantity of manganese has an equal volume with the
quantity of iron for which it was substituted, and therefore
fills an equal space. But, as we shall presently show, the
elements are not always held in combination with the same
equivalent volume as they have in the isolated state; we there-
fore term the latter, the original equivalent volume.

The equivalent volume of a body must vary with the tem-
perature, since bodies become expanded by warming, and
be greater at a higher than at a lower temperature. In
other words, since the specific weights of bodies diminish
as those bodies arc warmed, it follows that in calculating the
equivalent volumes, the divisor becomes less in proportion as
the temperature is increased : the quotient — the equivalent

6 THE VOLUMES OF BODIES

volume— becomes, therefore, so much the larger. According to
the atomic theory, the expansion of bodies on warming is
dependent on an increase in the relative distances of the atoms.
Atoms, together with the spaces intervening between them,
or, as we generally express it, the surrounding spheres of
heat, assume, therefore, a larger space in a higher than in a
lower temperature; that is to say, the atom- volume is larger in
the former than in the latter state.

The question now arises, at what temperature should the
equivalent volumes of bodies be submitted to comparison with
one another ? That in this case corresponding temperatures are
not equal temperatures is at once obvious, if we take into con-
sideration the very different degrees of expansion which fluid
and solid bodies undergo from an equal amount of heat. We
shall subsequently find that for fluid bodies those are corres-
ponding temperatures at which the tension of their vapours is
equally strong, and that consequently at these temperatures the
corresponding equivalent volumes must be determined, if we
wish to institute a comparison between them. In the case of
solid bodies, Kopp regards those temperatures as probably cor-
responding which lie equidistant from the fusion-points. For
such bodies as have nearly similar fusion-points the same de-
crees of temperature denote also corresponding temperatures,
and in such cases it must follow that the augmentation of the
equivalent volumes is the same for similar elevations of tem-
perature, and that a simple relation is established. Thus, ac-
cording to Kopp, an equivalent volume of zinc increases in bulk
on the temperature being raised from 32° to 212° F. almost
exactly as much as an equivalent volume of tin, while an
equivalent volume of bismuth increases to nearly double the
extent; the same elevation of temperature increases to nearly
the same degree an equivalent volume of platinum and an
equivalent volume of palladium, while, if equivalent volumes of
gold and copper are submitted to a similar elevation of tempera-
ture, the augmentation of the former is nearly twice as great as
that of the latter. The equivalent volume of gold at 32° is
fixed at 130, and that of copper at 44; by raising the tempera-
ture to 212° the former becomes increased to 130*46, and the
latter to 44*23; so that while for the copper the augmenta-

AND THEIR ATOMIC WEIGHTS.

tion is 0*23, the corresponding augmentation for the gold is
2 x 0*23. It is scarcely necessary to remark that from this a
partition of the equivalents of bismuth and gold into two atoms
might be inferred. The equivalent volumes of the elements in
the above table are calculated for the ordinary mean tempera-
ture, since the specific gravities used as divisors are the specific
gravities at that temperature.

According to Schroder*, solid bodies are in corresponding
states at such temperatures as are at corresponding distances
from their fusion-heat, and in such conditions their equivalent
volumes stand in the relations of whole numbers. At a tem-
perature removed about as far from their fusion-points as the
ordinary temperature from a red heat (400° to 1000° C.) the
equivalent volumes of solid bodies are expressed by numbers
which, taken together, are multiples of 4; these are termed
by Schroder the theoretical equivalent volumes. They are col-
lected in the following table:

Aluminum ....

.... 60 |

Molybdenum

... 68

Arsenic

.... 156

Nickel

... 44

Barium

.... 144

Osmium

... 64

Bismuth

.... 128

Oxygen

... 64

Bromine

.... 368

Palladium ....

... 52

Cadmium

.... 80

Phosphorus....

... 220

Calcium

.... 56

Platinum

... 52

Carbon

.... 36

Potassium ....

... 240

Chlorine

.... 240

Rhodium

... 56

Chromium ....

.... 72

Selenium

... 112

Cobalt

.... 44

Silver

... 128

Copper

.... 44

Sodium

... 128

Gold

.... 64

Strontium ....

... 104

Iodine

.... 424

Sulphur

... 112

Iridium

.... 52

Tellurium ....

... 128

Iron....

.... 44

Tin

... 100

Lead

.... 112

Titanium

... 56

Manganese ....

.... 44

Tungsten

... 68

Manganic acid

.... 56

Zinc...

.... 56

Mercury

.... 88

It is evident that in order to determine the equivalent volumes
of compounds we must proceed in exactly the same manner as
in determining the equivalent volumes of the elements in their

* Schroder, die Molecularvolumina der chemischen Verbindungen. Mannheim,
Fr. Bassermann, 1843. For a criticism of his Memoir, consult Kopp, Bemerkungen
zur Volumtheorie. Braunschweig, 1844.

8 THE VOLUMES OF BODIES

solid (or fluid) states ; that is to say, we must divide the equiva-
lent weight by the specific weight. The equivalent weight
of oxide of copper, Cu O, for instance, is 396 + 100 ==496;
its specific weight is 6*4; the equivalent volume is therefore
— 1"=77'5. 496 grammes of oxide of copper (1 Eq.) conse-
quently occupy a space of 77*5 cubic centimeters. The equiva-
lent weight of sulphide of copper, Cu S, is 396+200 = 596;
its specific weight is 4*16; the equivalent volume is, therefore,
fr^|-= 144. The equivalent weight of the disulphide of copper,
Cu2S, is 2x396 + 200 = 992; the specific weight is 5*97; the
equivalent volume is therefore 5.97 = 166, &c, &c.

The question now arises, in what relation do the equivalent
volumes of compound substances stand to the equivalent volume
of their constituents? Is the equivalent volume of the com-
pound the sum of the equivalent volumes of its constituents, or
does a condensation or an expansion accompany the chemical
union? If, before discussing this question, we take a glance at
the relation in which the volume of a gaseous compound stands
to the volume of its gaseous constituents, we find that in many
cases the volume of the compound is equal to the sum of the
volumes of the constituents, but that in most cases a condensa-
tion, and, in a few instances, an expansion follows. If con-
densation or expansion follows, the relation of the volume of
the compound to the volume of its constituents is invariably
expressed by very simple numbers. For instance, 1 Eq.-vol.
(2 vol.) of hydrogen gas and 1 Eq.-vol. (2 vol.) of chlorine gas
give 1 Eq.-vol. (4 vol.) of hydrochloric-acid gas; the volume
of the compound is the sum of the volumes of the constituents.
3 Eq.-vol. (6 vol.) of hydrogen gas and 1 Eq.-vol. (2 vol.) of
nitrogen gas give 1 Eq.-vol. (4 vol.) of ammoniacal gas; hence
it follows that with the chemical union there is a condensation
in the relation of 8:4 or of 2 : 1 ; 1 Eq.-vol. (2 vol.) of hydrogen
gas and 1 Eq.-vol. (1 vol.) of oxygen gas give 1 Eq.-vol. (2 vol.)
of aqueous vapour; there being in this case a condensation in
the relation of 3:2. 1 Eq.-vol. (2 vol.) of vapour of mercury
and 1 Eq.-vol. (i vol.) of vapour of sulphur form 1 Eq.-vol. (3
vol.) of vapour of sulphide of mercury, the chemical union
being, in this instance, accompanied by an expansion in the
relation of 2-^ : 3 or of 7 : 9.

AND THEIR ATOMIC WEIGHTS. 9

If a condensation occurs in the union of gaseous elements
we may, in various ways, conclude beforehand what is likely to
occur. "We may assume that the gases enter into the combina-
tion, which is then condensed, or we may suppose that one or
other of the gaseous constituents, or all of them, become con-
densed previous to their combining, and that then, without
further condensation, chemical union occurs. According to the
former view, the compound suffers condensation, while, accord-
ing to the latter, its constituents undergo this modification.
For instance, in the union of 6 vol. of hydrogen gas and 2 vol.
of nitrogen gas, we may either say that the 8 vol. of ammoniacal
gas that are produced are condensed to 4 vol., or that the
6 vol. of hydrogen gas are condensed to 3 vol. (3 Eq.-vol.)
and the 2 vol. of nitrogen gas to 1 vol. (1 Eq.-vol.) and that the
union into the 4 vol. of ammoniacal gas now follows.

Let us now see in what relations the chemical union of solid
bodies occurs; that is to say, in what relation the volume of the
compound in the solid state stands to the volume of the con-
stituents in the solid state. If there occur between the two,
simple relations similar to those between the volumes of gaseous
compounds and of their constituents, we are then in a condition
to calculate from their solid compounds the equivalent volume
of those elements which, in their isolated state, we cannot obtain
in a solid form, as oxygen, hydrogen, &c, and knowing the
equivalent volume of these elements, we likewise know their
specific weight in the solid condition, since the latter is obtained
on dividing the equivalent weight by the equivalent volume.
In an analogous manner we may determine from their gaseous
compounds the specific weight of the gases of such bodies as,
isolated, cannot be obtained in the gaseous form, as, for instance,
the specific weight of carbon vapour from the specific weight of
carbonic acid gas.

"We have already found that the equivalent volume of sul-
phide of copper, Cu S, is ^\GQ or 144. The equivalent volumes
of copper and sulphur are known, that of copper being 44, and
that of sulphur 100; hence the equivalent volume of sulphide
of copper is exactly equal to the sum of the equivalent volumes
of its constituents. 44 cubic centimeters of copper (1 Eq.) and
100 c. c. of sulphur (1 Eq.) yield exactly 144 c. c. (1 Eq.) of
sulphide of copper. The equivalent weight of sulphide of

10 THE VOLUMES OF BODIES

silver, Ag S, is 1550, and the specific weight is 6'8; the equiva-
lent volume is therefore 1^58° or 228, a number which ap-
proaches quite as closely to the sum of the equivalent volumes
of silver (130) and sulphur (100) as we have any right to
expect; 130 cubic centimeters of silver and 100 c. c. of sulphur
yield, therefore, 230 c. c. of sulphide of silver. The calculated
specific weight of sulphide of silver is consequently W^0 or
6*74, which does not differ from the observed specific weight
by quite 1 per cent. (674 : 680 = 100 : 100-9.)

The equivalent volume of disulphide of copper, Cu2 S, has
been shown to be 59.^ or 166. The sum of 2 equivalent
volumes of copper and 1 equivalent volume of sulphur is, how-
ever, 188, (namely, 44+44 + 100,) whence it follows that the
equivalent volume of disulphide of copper does not directly
correspond with the sum of the equivalent volumes of its con-
stituents. The equivalent weight of bisulphide of iron, Fe S2>
is 750 (350 + 200 + 200,) and the specific weight is 5*08; the
equivalent volume is therefore ^.50°s or 147. On adding the
equivalent volumes of the constituents, those namely of Fe and
2S, we obtain the number 245, and hence it follows, as in the
disulphide of copper, that the equivalent volume of the bisul-
phide of iron is much smaller than the sum of the equivalent
volumes of its constituents; a condensation must occur in the
chemical union of the copper and sulphur into disulphide of
copper, and likewise in that of the iron and sulphur into bisul-
phide of iron. The equivalent volume of potassium, if reckoned
from its equivalent weight (489,) and its specific weight (0*84,)
is 583. But the equivalent volume of the oxide of potassium,
K O, if calculated from its equivalent weight (589,) and its
specific weight (2*66) is 221, and therefore far smaller than
that of potassium, one of its constituents; 583 cubic centimeters
of potassium together with the oxygen required for oxidation
have been condensed into 221 c. c. of potash.

The equivalent weight of the sulphide of arsenic, commonly
known as realgar, As S2, is 1337*5 (937*5+200 + 200,) and its
specific weight is 3*56; the equivalent volume is, therefore, 375.
The sum of the equivalent volume of the constituents of realgar
amounts, however, to 360 (160 + 100+100,) and hence it follows
that the equivalent volume of realgar is greater than the sum of

AND THEIR ATOMIC WEIGHTS. 11

the equivalent volumes of its constituents ; hence, in the chemical
union of arsenic and sulphur into realgar, there must occur a
negative condensation, or, in other words, an expansion.

As in the chemical union of elements in the states of gas or
vapour, so also in the chemical union of elements in the solid
condition we find that the volume of the resulting compound
is either equal to, or smaller than, the volume of the constituents,
in which latter case (one, by the way, of very frequent occur-
rence,) condensation must have taken place, or finally, that the
volume of the compound is greater than the volume of the con-
stituents, and that consequently an expansion takes place.

After Herapath, Boullay, and Karsten had sought in vain
to find any regular law regarding the change in volume occurring
in the chemical union of solid bodies, Kopp endeavoured to give
formula? for the calculation of the specific weights of compounds,
from the specific weights and equivalent weights of their con-
stituents, whereby he was led to conclude that condensation
occurred in the resulting compound. {Pogg. Annalen, Bd. 47,
s. 133 — 153.) By interesting comparisons of the specific
weights of metals, combined with various proportions of oxygen
and sulphur, with the equivalent weights, Ammermiiller was led
to the discovery of a principle which, according to Poggendorf,
may be thus expressed : the specific weights of the compounds of
a radical with an electronegative body are to one another, either
directly as the equivalent weights (the atomic weight) of the
compounds, or as multiples or sub-multiples of them. We may
readily see that this proposition is identical with the following:
the equivalent volumes of the compounds of a radical with an
electronegative body are either equal to, or are multiples or sub-
multiples of, one another. For instance, the specific weight of
the dinoxide of copper is 5*749, and the specific weight of the
oxide is 6*4 ; but 5*749 is to 6*4 as the equivalent weight of the
dinoxide of copper is to twice the equivalent weight of the
oxide of copper : —

Spec, weights. Equiv. weights.
Cu2 0:Cu 0=5*749:6*4 = 892:496 x2

The equivalent volumes of the dinoxide and the oxide of copper
are, by calculation, respectively 58^429 and ^?^, or 155 and
77*5 ; the former is therefore exactly double the latter. The

12 THE VOLUMES OF BODIES

equivalent weights being in general determined with the greater
accuracy, we may calculate the specific weight of one com-
pound, on this principle, from that of another determined with
sufficient accuracy. The following table gives the observed
specific weights of different compounds, and the relation of the
equivalent weights corresponding to them.

Spec, weights.

Equiv.

weights.

SnO

: Sn02

6-666

. 69

835-3

: 935-3

PbO

: Pb02

801

: 919

1394-5

. 1494-5

Cu20

: CuO

5749

6-4

892

496 X

Hg20

: HgO

10*69 :

11-29

2600 :

1350 X

SnS

: SnS2

5-267 :

4-415

935-3

11353 X

MnO

: Mn02

4-726 :

3-760

444-7 :

544-7 X

MnO

: Mn203

4*726 :

4-328

444-7 :

989-4 X

2
5

Mn203

: Mn02

4-328 :

3-760

989-4 :

544*7 X

5
3

AsS9

: AsS3

3*544 :

8-459

13375 :

15375 X

Hg2Cl

: HgCl

7-14 :

5*42

2943-2 :

1693-2 x

We must assume that there are very considerable errors in
the determination of the specific weights, if we are to consider
that this table serves for the confirmation of the above rule. If
from the given relations of the equivalent weights we calculate
the specific weight of one or the other compound, it is only in
extremely few cases, as for instance in the dinoxide and oxide of
copper, that we obtain a number in any degree approximating to
the observed specific weight. The specific weights of the oxide
and peroxide of lead ought, for instance, to be exactly in the
relation of the equivalent weights of their compounds; and,
therefore, as 1394*5 to 1494-5. Since the specific weight of
the oxide can probably be determined with greater accuracy
than that of the peroxide, we had better select the latter for
calculation. We have then —

1394-5 : 1494-5 = 8-01 : x, which therefore =8*58.

In accordance with the rule in question, we calculate the
specific weight of peroxide of lead at 8*58, while by direct expe-
riment we find it to be 9*19. The specific weights of As S2
and As S3 should be to one another as the equivalent weight
of As S2 to -| of the equivalent weight of As S3 ; that is to say,
as 1337*5 : 1537*5 xf, or as 1337*5 : 1230. If again we calcu-
late the specific weight of As S3 from that of As S2, we have
1337*5 : 1230 = 3*544 : x, or x = 3*25, while from direct observa-
tion we find that its specific weight is 3'459.

AND THEIR ATOMIC WEIGHTS. 13

Schroder has laid down the following law regarding the
relation of the equivalent volume of a compound to that of its
constituents : the equivalent volume of every compound is the
sum of the equivalent volumes of the constituents of the com-
pound (elementary law) ; the equivalent volume of a constituent
in a compound may differ from that which it possesses in an
isolated condition, but in that case it always stands to the latter
in a simple relation ; that is to say, according to Schroder, in
the relations of the numbers 1:2:3:4:5:6 (the law of conden-
sation). According to this law, the positive or negative con-
densation (condensation or expansion) occurring in a chemical
union was referred to the constituents ; their equivalent volume
is changed in simple proportions, and these, positively or nega-
tively, condensed constituents unite to form the equivalent
volume of the compound, (seep. 8). In the case of sulphur, for
instance, Schroder assumed that it enters into combination with
ii -f, -J, and 1 the original equivalent volume. (Pogg. Annalen,
Bd. 50, s. 553 — 604). "We have already seen that the equiva-
lent volume of disulphide of copper is expressed by the number
166, and that the sum of the equivalent volume of the consti-
tuents is 188 (88 + 100). According to Schroder's assumption,
on the union of the copper and the sulphur the equivalent
volume of the copper remains unchanged, namely, 88 ; while
the equivalent volume of the sulphur, 100, is reduced to ^9 that
is to say, to 80. 88 cubic centimeters of copper and 100 c. c.
of sulphur condensed to 80 c. c. combine, therefore, to form 168
c. c. of disulphide of copper. The equivalent volume of bisul-
phide of iron, Fe S2, commonly known as iron pyrites, is, as
we have shewn by calculation, 147. The sum of the equivalent
volumes of its constituents amounts to 245 (45 + 100 + 100).
Schroder consequently assumed that the unchanged equivalent
volume of iron combines with the equivalent volume of sulphur,
condensed to one-half its bulk to form iron pyrites, (45+50 + 50
= 145).

Some time previously to Schroder's announcement of the
law of the relation of the equivalents of compounds to the
equivalent volume of their constituents, he made the discovery
that if from the equivalent volume of compounds which contain
a common constituent, we subtract the equivalent volume of the

14 THE VOLUMES OF BODIES

other constituent an equal remainder is, in many cases, left for
the aforesaid common constituent. The importance of this
discovery is at once obvious ; it indicates the mode by which we
can find the equivalent volume of bodies in which it cannot be
directly determined, since they are not to be obtained in the
solid state ; as, for instance, the equivalent volume of oxygen in
the solid condition.

The following table exhibits this in the case of oxides: it
contains in column A the equivalent, in column B the specific
weight of the oxide, in column c the equivalent volume calcu-
lated from the above data, in column D the equivalent volume
of the metal, and, finally, in column e the difference between the
equivalent volume of the oxide and that of the metal (c — d).

TABLE I.

Equi-

Spec.

Equiv.

Equiv. vol.

Difference.

valent.

grav.

vol.

of Metal.

C-D.

CuO

496

6-4

77-5

33-5

ZnO

506

543

92-6

33-6

CdO

796-7

6-95

1148

33 8

PbO

13945

9'5

146-8

114

32-8

HgO

1350

11-0

123

31-0

Fe203

1000

5-22

191-5

1015

3X338

Co203

1037

5-6

1854

974

3x32-5

Sb03

1913

5-56

344

240

1040

3X346

Ti02

501-5

4-18

120

640

2X32

We thus find that for the equivalent volume of oxygen we
obtain the same number, one nearly approaching to 32, if from
the equivalent volume of these oxides we subtract that of the
metal, and we may consequently well believe that the number
32 actually expresses the original equivalent volume of oxygen
in the solid form. All oxides do not, however, give the same
result, as we may see from the following table.

TABLE

II.

Equi-

Spec.

Equiv.

Equiv. vol.

Difference,

valent.

grav.

vol.

of Metal.

C-D.

Sn02

9353

6-95

134-5

101

33-5

2x 10-7

Cr203

957-0

5-21

1837

128

55-7

=■

3 x 183

AgO

1450

725

200

130

Cu20

892

5'75

155

Mo 03 '

896

264

195

3X65

Wo 03

1488

612

243

174

3X58

AND THEIR ATOMIC WEIGHTS. 15

In the two first-mentioned oxides the oxygen is contained
with an equivalent volume of about 16, (and therefore -^ x 32;) in
the four last cases with an equivalent volume of 64, or 2 x 32.

If, in accordance with these tables, we express the equiva-
lent volume of oxygen by the number 32, it affords a confirma-
tion of the above-mentioned law of Schroder. In the formation
of the oxides of copper, tin, and iron, in titanic acid, &c, the
equivalents of the metal and the oxygen remain unchanged, the
equivalent volume of the compound being the sum of the
original equivalent volumes of the constituents. In the chemical
union of tin with oxygen, forming oxide of tin, and in that of
chromium with oxygen, forming oxide of chromium, a condensa-
tion of the equivalent volume of oxygen to one-half has taken
place ; while in the chemical union of silver and oxygen, forming
oxide of silver, and in that of copper and oxygen, forming
dinoxide of copper, the equivalent volume of the oxygen be-
comes enlarged to double its bulk.

The equivalent volume of potash has already been calculated
as 222; the original equivalent volume of potassium is 583.
If from the equivalent volume of potash we subtract that of
oxygen, (namely, 32,) there remains 190 as the equivalent
volume of the potassium in the potash; this number is pretty
nearly -1-rd of 583, which expresses the original equivalent
volume of potassium, for 3 x 190 = 570. Hence, in the forma-
tion of potash, a condensation of the equivalent volume of the
potassium to -3rd of its original volume has taken place. The
equivalent volume of water is 112*5 (x-Voo) ' if we subtract
from this, 32, as the equivalent volume of oxygen, there remains
the number 80*5 to represent the equivalent volume of hydrogen,
if we assume that in the formation of water the hydrogen and
oxygen unite in their original equivalent volumes.

It is easy to see that in accordance with Schroder's law the
equivalent volumes of compounds may frequently be explained
by the assumption of different degrees of condensation; this
condensation occurring in either one or both constituents. The
equivalent volume of titanic acid, Ti02, has been already found
to be 120; on the supposition that the titanium and oxygen are
contained in their original equivalent volume, the equivalent
volume is by calculation 57+32 + 32=121, which coincides

16 THE VOLUMES OF BODIES

very well with observation. Schroder has, however, assumed
that in titanic acid the equivalent volume of titanium is ex-
panded to one and a half time its original bulk, while that of
oxygen is condensed to one-half. We then obtain ^ x 57 + -^ X 64,
or 117*5, as the equivalent volume of titanic acid; and similarly
in many other cases.

More recently Schroder has modified his views; instead of
considering' that the condensation affects the individual constitu-
ents, he now believes that the condensation occurs in the com-
pound that is formed. (See p. 9.) "When solid elements
combine, the condensations which occur stand in the relation of
whole numbers to the volume of the elements;" thus Schroder
states the law which he regards as holding good*. We must
not forget that, according to Schroder, the original equivalent
volumes of the elements are multiples of 4, if the elements occur
in corresponding conditions, (page 7). The condensations
(whether positive or negative,) which occur in the union of the
elements are expressed, according to Schroder, by the numbers
16 or 8, and by multiples of these numbers, and consequently
of the number 4. According to Schroder's table given in
page 7,

The equivalent volume of iron is ....

„ „ mercury

„ „ sulphur

„ „ chlorine

And therefore —

The volume of 2 eqs. of mercury ....
„ 1 eq. of chlorine ....

The equivalent volume of the constituents of Hg2 CI
The condensation

The equivalent volume of IIg2 CI (calomel)

Further —

The equivalent volume of mercury is
„ „ sulphur

The equivalent volume of IlgS (cinnabar) ....
The condensation

The equivalent volume of cinnabar .... .... .... 180

* Die Molecularvolume der chemischen Verbindungen von H. Schroder. Mann-
licini, 1843.

100

240

. 176

. 240

. 416

...

0
. 416

100

188

AND THEIR ATOMIC WEIGHTS. 17

Further —

The equivalent volume of the elements of iron pyrites, i

Fe S2, (44 + 200) f

The condensation .... .... .... .... .... .... 96

244

The equivalent volume of iron pyrites .... .... .... 148

In the memoir to which we have already alluded, Schroder
has established, in accordance with this law, the equivalent
volumes of various compounds — of oxides, sulphides, sulphates,
carbonates, &c, — and thence calculated their specific weights in
order to show how far his theory is confirmed by experiment.
For an element or constituent of compounds, of which the
equivalent volume cannot be determined in its original state,
from its not being obtainable in either a solid or isolable condi-
tion, he has calculated the equivalent volume according to this
law from its compounds, in a manner which we shall proceed to
investigate more closely.

With this view I insert an extract from the table that
Schroder has calculated for the Oxides. I have not deemed it
necessary to make any alteration in his equivalent weights,
although some of them have been modified by more recent
determinations; since, for our present object, they may as well
remain as they are.

Oxides. " The empirical mean for the volume of oxygen
deduced from 37 oxides is 64; I obtain it by increasing the
observed volume of the oxide by the condensation that occurs,
by subtracting from this value the volume of the metal, di-
viding the remainder by the number of atoms of oxygen, and
taking the mean of all the values thus empirically determined.
With the volume 64 (=8x8) for the atom of oxygen, I have,
in the following table, explained the volume of the oxides."

THE VOLUMES OF BODIES

■J
<

"TJ

Sh
03

-4-»
CO

s^»

l-H

-25
O

l-H

■4^

a
a>

-ta

■+-*
CO

i-H

r— H

r— 1

03
N
U
03

-^
ft

=3
e3

0)
.8

OS
CO

i— i

i-H

l-H

• l-H

»b

—

CO
CM

—

r— 1

03
CO

03
VI

03
00

'S :o

03
i-H

$ *

l-H

lO l-H

"-H

•— •

l-H

CM CM

»b

l-H

TJ"

»b »b

•XjIAEJf) 01J10

!>•

»>•

CM CO

1— 1

CO CM

-gdg pajv^nopQ

1— H

■^

o o

•atun[o^

l-H

l^»

t^.

pa.vjasqQ

TJ<

1— 1
1— 1

t—

l-H

i-H i-H

•auuip)^

•Tf

"-* CM

psDijaioaqj^

l-H

CM
CM

i-H l-H

'aoi^BsuapuoQ

§ 3

">*

»o

t—

~ 3J

•-h

l-H

i-H

r— 1

^-"

CM ©

oftheVo:
e Compor

!! II

Tji

■^

■»*

CM CM

l-H

l-H I"H

+ +

B-3

CO o

Tt«

!>.

■— I

T}<

»o

-tf

1— 1

l-H

o —

'S "to

—

■<*

l^» ©

TJ"

»o

«£

l-H

*~ '

l-H

o o

+ +

a
o

fcx>

03 ~
CM CM

i S

ft.

!h
O

03
-

03
ft
ft
O

h
03

tj-C

«4H

<+H <+H

o o

c3
■73

03 03

o o

AND THEIR ATOMIC WEIGHTS. 19

In the first place, the description of this table demands at-
tention. " The empirical mean for the volume of oxygen, 64,
was obtained by increasing the observed volume of the oxide by
the condensation that occurs, by subtracting from this value
the volume of the metal, &c." We naturally inquire by what
means the condensations can be known, and how we are to learn
what condensation must be applied in individual cases. We
can, unfortunately, only reply that there is no certain con-
trolling point for ascertaining the condensation. The condensa-
tions that occur are, according to Schroder, such as must be
assumed in order that his own law of condensation may be best
elucidated, and that the volume of oxygen assumed by him to
be 64, may be explained in the clearest possible manner! At
least, no other answer than this suggests itself to us. As we
have already seen, very simple considerations render it so ex-
tremely probable that the equivalent volume of oxygen is 32,
that we can hardly help looking upon Schroder's alteration of
this number to 64 as unwarrantably arbitrary. Schroder was
just as much in a position to draw up a table to represent the
law of condensation if he had allowed the equivalent volume of
oxygen to remain at 32. The condensation 32 for the oxides
represented by the formula R O, the condensation 64 for R 02,
and 96 for R2 03 and R 03, would then naturally be omitted;
that is to say, in all these oxides the equivalent volume would
be exactly equal to the sum of the equivalent volumes of their
constituents. When Schroder is not in a position to satisfy
the law of condensation with the aid of the original (that is
to say, the observed,) equivalent volume of a metal, he seems
to consider himself entitled to alter the original equivalent
volume into what he terms a theoretical equivalent volume.
The oxide of chromium affords a proof of this statement. The
oxide of chromium is Cr2 03, its (old) equivalent weight is
1004, and its specific weight 5*21: hence the equivalent
volume amounts, on calculation, to 193 (5?° f )• The (old)
equivalent weight of chromium is 352, and the specific weight
5*10; the equivalent volume is, consequently, 69. If, now, we
assume, with Schroder, that the equivalent volume of oxygen is
64, then the sum of the equivalent volumes of the constituents
of oxide of chromium is 330=69 + 69 + 64 4- 64 + 64. Accord-

20 THE VOLUMES OF BODIES

ing to Schroder, the condensation amounts to 144 (9x16), and
if we subtract this number from 330, there remains 186, as the
theoretical equivalent volume of oxide of chromium. This
number differs, however, too far from 192, (the number found
in the direct manner,) and in order to obtain one more fitting to
his views, he raises the equivalent volume of chromium, in what
we must regard an arbitrary manner, from 69 to 72 ! The calcu-
lated equivalent volume is, then, 336 — 144, or 192. In this man-
ner Schroder has obtained the theoretical equivalent volumes of
the elements given in page 7, which, as we have already stated,
are also multiples of 4. He regards the equivalent volume of
chromium in the isolated state as 69, while in combination it is
72; he places the original equivalent volume of bismuth at 272,
while in combination he regards it as 256! Kopp, in his
critique on Schroder's work, has shown that we might just as
properly consider the equivalent volumes of the elements to be
multiples of 3 as multiples of 4. Although it may be deemed
almost superfluous, we proceed to make an extract from Schro-
der's tables, calculated for sulphides and carbonates.

Sulphides (Sulphurets.) "If from the observed volume
of any sulphide I subtract the volume of the metal diminished
by the condensation that occurs, and divide the remainder by
the number of atoms of sulphur contained in the combination, if
I sum up the whole of the values thus obtained and divide by
the number of observations, I obtain the whole number, 100, as
the mean of the volume of an atom of sulphur. This value is
especially deduced from 28 different bodies, and with it I have
exhibited the volume of the metallic sulphides in the following
table. It is to be observed that this value is an exact multiple
of 4."

AND THEIR ATOMIC WEIGHTS.

•fh

■•a

•

a
9

Is!

c3
E

i—5

.2
'3

ed
E

•si

d
—

• —

8
CD

—
c3

■*3

-a

co
E

l-H
»0

o3
0

d
a

"3
"-3

^-4

*CJ

'if

3
eg

O
-*>

-u

-*»*

Tt<

■t

«3

•*

— <

o
o

(Z.
O

—

—
CD

—

l-H

^»

IT*

ir^

Tl<

■Jf

•^JIABJQOtJIO

l-H

-edg p^BjnojB;}

■*<

f»

ir»

4j<

•atnn^Y

*-»

■^

t^»

t»

paAjasqo

•t

-<*

—

"<*

l-H

•^

i— i

•acunprY

"<*

pjopa-toaqj,

l-H

•<*

l-H

-H

l-H

»-H

—

»o

•uoijusuapuoQ

l-H

—

-<*

^J<

■<*

D 4)

i— 1

l-H

2 g

2 S

■5 °

l-H

!-H

l-H

i-H

l-H

O 0J

■<*

•^

l-H

-Cf

l-H

»-«

—

Cfl o

u *j

t*»

"Cf

t^.

t"»

a ep

—

"Cf

5 4*

"<*

«5

—

l-H

CJO

—

tsi

Ph:

—

•

S
I

e»H

a
d

o
©

is*

d
9

o
'3

—

v>
d
<

-**

•c

>»

d
o

—

22 THE VOLUMES OF BODIES

As in the case of oxygen so also for sulphur, we find
Schroder calculating the equivalent volume from the sulphides
(as in the other case he had done from the oxides), in accordance
with the condensation that occurs, or, in other words, with
the condensation that best suits his views, and it must be
regarded as a complete chance that the equivalent volume (100)
thus calculated is equal to the original equivalent volume.
With an equal amount of probability we might deduce the
number 108 from the sulphides as the equivalent volume of
sulphur; all that would be required being merely an altered
condensation. For sulphide of cadmium, Cd S, we should have
80 + 108, or 188, as the equivalent volume, and the condensation
would therefore be 0 ; for sulphide of copper, Cu S, we should
have 44+108, or 152, and therefore the condensation 8; for
sulphide of zinc (blende), and sulphide of mercury (cinnabar),
the condensation would be 16. For disulphide of copper it is
88 + 108, or 196; or if we assume the condensation to be 32, we
obtain 164 as the equivalent volume, which coincides far better
with the number actually observed than the number 172
calculated by Schroder. He has likewise made changes in
the original equivalent volume which we can only regard as
arbitrary, in order to obtain multiples of 4. Thus the original
equivalent volume of mercury is altered from 92 to 88, and
that of silver from 130 to 128.

Carbonates. " When from the observed volume of any
carbonate I subtract the volume of the metal and augment
the remainder by the condensation that occurs (which is always
a multiple of 16,) I obtain as a mean the number 172. This
is again an exact multiple of 4. I have consequently assumed
172 as the volume with which, in case there is no condensation,
one atom of carbon and three of oxygen combine in the car-
bonates. The volume of carbon is 36, and that of three atoms
of oxygen is 3 x 64, or 192; hence the original volume of C 03
is 228 ; the smallest condensation of C 02 in the carbonates is
therefore 228-172, or 56, which =7x8. If we start from
the original volume of carbon and oxygen, then all the con-
densations occurring in the carbonates are multiples of 8, not
of 16, but they progress from 16 to 16."

AND THEIR ATOMIC WEIGHTS.

For the sake of convenience I will present this table as
Kopp gives it in his remarks on the volume theory.

TABLE V.

Sum of the Volumes of

Conden-

Theoretical

Observed

the Componeuts.

sation.

Volume.

CaC03

56 4- 172 = 228

228

231

(Calc spar.)

CaC03

56 + 172 = 228

212

209

(Arragonite.)

AgC03

128 + 172 = 300

284

CdC03

80 + 172 = 252

236

239

FeC03

44 + 172 = 216

184

185

BaC03

144 + 172 = 316

284

KC03

240 + 172 = 412

380

383

ZnC03

56 + 172 = 228

180

176

And here we cannot refrain from observing that the assump-
tion that 172 expresses the equivalent volume of C 03 is alto-
gether arbitrary, and that we may with equal propriety
substitute another number in its place; for equally correct
results are then obtained, if we only assume that a different
degree of condensation ensues. If for instance, in place of 172
we substitute 188, or 172 + 16, as the equivalent volume of
C 03, the table assumes the following appearance.

TABLE VI.

Formula.

Sum of the Volumes of
the Components.

Conden-
sation.

Theoretical
Volume.

Observed
Volume.

CaC03

56 + 188 = 244

228

231

(Calc spar.)
CaC03

56 + 188 = 244

212

209

(Arragonite.)
AgC03

128 + 188 = 316

284

CdCO,

80 + 188 = 268

236

239

FeC03

44 + 188 = 232

184

185

BaC03

144 + 188 = 332

284

KCO3

240 + 188 = 428

380

383

ZnC03

56 + 188 = 244

180

176

Nor should we be less correct in assuming 172 — 16, or 156,
as the equivalent volume of C03; the table becoming then
modified in the following manner.

THE VOLUMES OF BODIES

TABLE VII.

Formula.

Sum of the Volumes of
the Components.

Conden-
sation.

Theoretical
Volume.

Observed
Volume.

CaC03

56 + 156 = 212

-16

228

231

(Calc spar.)
CaC03 ....

56 + 156 = 212

212

209

(Arragonite.)
AgC03

128 + 156 = 284

284

CdC03

80 + 156 = 236

236

239

FeC03

44 + 156 = 200

184

185

BaC03

144 + 156 = 300

284

KC03

240 + 156 = 396

380

383

ZnC03

56 + 156 = 212

180

176

It might be maintained that there is greater probability
in the correctness of the number 156 for the equivalent volume
of C 03 than in Schroder's assumed number 172.

It is obvious, then, that Schroder's law of condensation
is not founded on any safe basis; in leaving the subject I will
make one further remark called forth by the last table. Car-
bonate of lime as calc spar, and carbonate of lime as arragonite,
have a different equivalent volume in consequence of their
having a different specific weight; thus, in general, dimorphous
bodies are distinguished from one another by a different equivalent
volume.

Carbon as diamond has a specific weight of 3*5, and there-
fore an equivalent volume of ~^, or 21; as graphite it has a
specific weight of about 2, and therefore an equivalent volume
of 2-3 or 37. According to Schroder a condensation amounting
to 16 would have to be assumed in the diamond. Kopp is
of opinion that the earlier view of Schroder is the correct one,
namely, that the condensation occurring in chemical union is
not to be referred to the compound but to the constituents,
but he considers that this degree of condensation does not
follow any simple relations. In an admirable memoir in the
Journal f. praht. Chemie, (1845, Bd. xxxiv., s. 1,) Kopp has
explained the views which led him to the consideration of this
subject, and I cannot do better than give the following epitome
of his view, as drawn from that essay.

AND THEIR ATOMIC WEIGHTS. 25

In the first place, Schroder's discovery that an equal
residue is left on subtracting the equivalent volume of a
common constituent from the equivalent volumes of analogous
compounds, gave, as it were, a starting-point for rational con-
siderations respecting the formation of the equivalent volumes
of compounds; to considerations indeed which, according to
Kopp, are probably still far from correctly exhibiting the
natural laws which govern the combining relations of bodies
in regard to bulk, but which, however, give extremely simple
expressions for the equivalent volume and for the weight of
many compounds.

For an individual and isolated compound no view can be
shown to be even probably true of the equivalent volumes of the
constituents contained in it, since from a single observation
concerning the equivalent volume of the compound nothing can
be concluded respecting several unknown magnitudes, (namely,
the volumes which two or more constituents have when in com-
bination). But for analogous compounds, views, possessing
more or less probability, may be advanced regarding the
equivalent volumes which we have to assign to the individual
constituents.

If we revert to our former supposition that the equivalent
weights indicate so many grammes, then 1 eq. of lead (Pb)
weighs 1294 grammes, and the space which it occupies amounts
to 114 c. c, (or the equivalent volume of Pb is 114). If we
transform this quantity of lead into nitrate of lead we unite
the elements of N 06 to the lead. The lead increases in weight
by 775 grammes, and in volume by 356 c. c, since the volume
of the nitrate of lead that is formed is 476 c. c. ; the equivalent
is 2069, and the specific weight 44, and 2^4 is equal to 476.
An equivalent of silver weighs 1352 grammes, and its volume
is 130 c. c. On converting it into a nitrate the elements of
N06, amounting in weight to 775, unite with it, and its volume
increases to 356 c. c; the volume of nitrate of silver is 486,
or 130 + 356.

These phenomena are very simply embraced in the assump-
tion that the equivalent volume of lead or of silver remains
unchanged in the nitrate just as it exists in those metals in
the uncombined state; the equivalent volume of the elements

26 THE VOLUMES OF BODIES

N 06, which convert a metal into a nitrate, being taken with
it, which in the nitrates is 356.

From such considerations Kopp* has sought to determine
for various other groups of elements whose addition converts
a metal into a salt, the equivalent volumes which must be
assigned to them in such compounds. An equivalent of lead
(1294 grammes) occupies 114 c. c. ; in the process of conversion
into oxide of lead, in which 100 grammes of oxygen are taken
up, the volume becomes enlarged by 32 c. c. An exactly equal
augmentation of volume takes place if we oxidize an equivalent
of copper (396 grammes, or 4 c. c), or of mercury (1250
grammes, or 92 c. c), or of zinc (406 grammes, or 58 cc,),.
100 grammes of oxygen being in every case taken up. An
equivalent of titanium (302 grammes) occupies 56 c. c. ; if it
is converted into titanic acid, in which case it takes up 200
grammes of oxygen, its volume becomes enlarged by 64, that
is to say, by 2 x 32 c. c. Two equivalent volumes of iron
(700 grammes) occupy 90 c. c; if we convert them into per-
oxide of iron, in which case they take up 300 grammes of
oxygen, their volume becomes increased by about 3 x 32 c. c.

All the facts are therefore simply explained on the assump-
tion that the equivalent volume of the metals in question is
the same in these oxides as that which the metals have in an
isolated state; the volume of one eq. of oxygen, moreover,
being 32, (100 grammes of oxygen in these oxides filling the
space of 32 c. c.)

In this manner Kopp has sought to determine the equi-
valent volume of oxygen and similar bodies in their com-
binations. If, as we have already presumed, the equivalent
volume of the metals remain unchanged in these combinations,
this assumption leads to satisfactory results for the equivalent
volumes of other constituents also, in the compounds of anti-
mony, lead, cadmium, chromium, iron, cobalt, copper, manga-
nese, molybdenum, nickel, mercury, silver, titanium, bismuth,
tungsten, tin, and zinc; in short, in the compounds of the dense
metals. These all, with the exception of arsenic, appear to be

' Kopp, liber das specifische Gewicht der chemischen Verbindungen.
Frankfurt A. M., 1841.

AND THEIR ATOMIC WEIGHTS. 27

contained in their compounds with their original equivalent
volume.

But the case is otherwise with the light metals; those,
namely, of the alkalies and earths. These are contained in
their combinations with an equivalent volume different from
that which they possess in an isolated state. An equivalent of
potassium, for instance, (489 grammes) fills a space of 583 c. c.
If we convert this into sulphate of potash we obtain 1089
grammes of that salt, occupying a space of 420 c. c. It is impos-
sible that the potassium contained therein can fill a space of 583
c. c. We must, therefore, assume that potassium (and similarly
the other light metals, whose equivalent volume we often do
not know in the isolated state,) has, when in its combinations,
an equivalent volume different from that which it possesses in
the isolated state. When we calculate what the equivalent
volumes of such metals are in combination, we arrive at the
result that each of the light metals possesses the same equivalent
volume in all its analogous combinations ; that is to say, in all
its salts. Kopp proceeds in the following manner : —

Since the equivalent volume of the nitrates of lead and silver
is larger by 356 than that of lead or silver, we have already con-
cluded that the equivalent volume of N 06 is 356 in these salts.
The equivalent volume of nitrate of potash (K with the ele-
ments N 06) is 590. If we assume that N 06 is contained
therein with the equivalent 356, it follows that the potassium
contained in this salt must be 590 — 356, or 234.

The equivalent volume of sulphate of lead (Pb with the ele-
ments of S 04) is greater by 186 than that of lead, whence we
conclude that the elements of S 04 are contained in sulphate of
lead with the equivalent volume 186. The same occurs in sul-
phate of potash, whose equivalent volume is 420, and the equi-
valent volume of the potassium contained therein must be
420 — 186, or 234. This is the same number as that previously
found, and Kopp consequently assumes that 234 represents the
equivalent volume of potassium in its salts. By considerations
of a similar nature, Kopp has obtained the following values for
the equivalent volumes of the light metals in their salts, and
also for ammonium : —

28 THE VOLUMES OF BODIES

Equiv. volume

Ammonium

218

Barium

143

Calcium ....

Magnesium

Potassium

234

Sodium

130

Strontium....

108

If we assume that the equivalent volume of C 03 (the ele-
ments by whose union with metals the carbonates are formed,)
is 159 ; that the equivalent volume of the heavy metals is of the
magnitude we have given in page 4 ; and that, that of the light
metals is correctly given in the above table ; then the equiva-
lent volumes of carbonates of Pb, Cd, Fe, Mn, Ag, Zn, Ba, Ca,
K, Mg, Na, and Sr, agree very well with the results of obser-
vation. For instance, the equivalent volume of lead is 114,
that of carbonate of lead (Pb + C 03) is calculated at 114 + 151, or
265 ; since the equivalent of the salt is 1669, the specific weight
is 166G9 or 6*3. The equivalent volume of strontium in its salts
is 108 ; that of the carbonate of strontium is, therefore, calcu-
lated at 108 + 151, or 259, and since the equivalent weight of
the carbonate is 921, the specific weight is |-f| or 3*55.

With the same assumptions regarding the equivalent volumes
of the metals, and presupposing that the equivalent volume of
N 06 (the elements with which the metals form nitrates) is 356,
we may calculate the equivalent volume and the specific weights
of the nitrates of Pb, Ag, Am (ammonium), Ba, K, Na, Sr,
and the result will be found to accord entirely with observation.

The supposition that in the sulphates the elements of S 04
have an equivalent volume of 236, explains the equivalent
volume and the specific weight of the sulphates of Cu, Ag, Zn,
Ca, Mg, and Na. But certain other sulphates render another
supposition requisite ; namely, that the elements of S 04 are
contained in them with an equivalent volume of 186 ; this is
necessary for the sulphates of Pb, Ba, K, and Sr.

In discussing other classes of compounds in the same manner
the volume of an element proves to be equally variable in its
compounds. Such an assumption, however, does not appear to
be opposed to nature, since the fact that dimorphous bodies have
different equivalent volumes leads directly to it. As far as

AND THEIR ATOMIC WEIGHTS. 29

concerns the sulphates, it appears not impossible that there is a
change from a modification where S 04 has the equivalent volume
186, into another modification in which S 04 has the equivalent
volume 236. The emission of light which occurs on the crys-
tallization in water of sulphate of potash which has been fused
with sulphate of soda, seems to indicate that the former by its
fusion with Na S 04 undergoes a modification in which S 04
acquires the equivalent volume 236, and that on its crystalliza-
tion it passes into a denser state, in which the equivalent volume
of S04is 186.

The assumption that the equivalent volume of Cr 04 is 288,
and that of W 04 244, accurately yields the equivalent volumes
and specific weights of the chromates of Pb and K, and of the
tungstates of Pb and Cu. The metallic chlorides are the only
haloid salts whose equivalent volumes have been ascertained
with any degree of certainty. Kopp assumes* that CI has the
equivalent volume 196 in its combinations with Pb, Ag, Ba,
Na, and the equivalent volume 245 in its combinations with
Am, Ca, K, Cu2, Hg, Hg2, Sr, and these views agree well
with the observations made on the density of these salts.

If we assume the equivalent volume of the dense metals to
remain the same in their oxides as already stated, (see p.
4,) but that the equivalent volume of oxygen becomes 32, we
shall find that the equivalent volume, and consequently, also,
the specific weight of the oxides, PbO, CdO, CuO, HgO, ZnO,
SnO, Ti02, Sb03, Fe203, Co2 03, FeTi03f, agree closely
with the results of observations. Thus, for instance, we have as
the equivalent volume for Sb 03, 240 + 3x32 = 336, and since
the equivalent weight of Sb03=1913, the specific weight is
^aVe3 or ^'^' ^ne equivalent volume of oxygen cannot, how-
ever, be assumed to be 32 in certain other oxides. If, for in-
stance, we oxidize 1 equiv. tin (735 gram.), occupying 101 c. c,
to form oxide of tin, the 2 equiv. of oxygen that are added occa-
sion an augmentation of volume of 32 c. c. ; 1 equiv. oxygen
(100 gram.) occupying, therefore, a space of 16 c. c. in this
oxide. We may assume that the equivalent volume of oxygen
is 1 6 in Sn 02, as it is in Cr2 03. Finally, we may consider

* In page 5 7 1 of the work referred to.
f Titanate of Iron : Ilmenite.

30 THE VOLUMES OF BODIES

the equivalent volume of oxygen to be 64 for the oxides Cu2 O,
Ag O, Hg2 O, Mo 03. (See p. 14.)

Such is the manner in which Kopp proceeds in considering
the relations of volume of solid chemical combinations, and
which he confesses to be merely relatively probable, and not
entitled to more credit in their further developement, than the
apparent correctness of certain unavoidable deductions confer
upon them. Such, for instance, is the confirmation they give to
the Radical theory of salts, highly probable on other grounds.

We shall now turn to the equivalent volume of liquids.
Observations upon liquids have an especial interest, from show-
ing more clearly than in solid bodies, how the change of equiva-,
lent volumes by temperature is to be considered.

Kopp found, in 1841, (Annalen der Chemie und Pharm.
1842, Bd. iii. s. 99), that analogous fluid compounds exhibit
like differences in their equivalent volume ; that, for instance,
the equivalent volume of a hydrated acid is smaller by 534
than that of the corresponding ethyl combination, and smaller
by 300 than the corresponding methyl combination, and conse-
quently that the equivalent volume of every ethyl combination
is greater by 234 than the corresponding combination of methyl,
for instance, —

Equiv. Equiv. vol. Difference.
7533 7087

1106-7 1243-5 '

57G-7 4GG-8

9300 10107 l*5329

Further, —

Hydrated acetic acid .... 1*063 753*3 708*7

A + HO Mollerat.

Acetate of the oxide of methyl 0*919 at 22° 930 0 1012*0

A + MeO Dumas and Peligot.

Spec. Grav.

Hydrated acetic acid
A + HO

1063
Mollerat.

Acetate of the oxide of ethyl

0*89 at 15° C.

A + EO

Hydrated formic acid ....

Liebig.

1 '2353 at 12°

F + HO

Formiate of the oxide of ethyl

Liebig.
0912

F + EO

Liebig.

303 3

Finally, —

Alcohol 0 792 at 18° C. 578*3 729*9

EO + HO GayLussac.

Wood-spirit 0*798 at 20° 401*6 503*3

Me O + H O Dumas and Peligot.

226*6

Spec. Grav.

Equiv

Equiv. vol.

Difference

0 98 at 15°

11067

12435

]

Liebig.

0-919 at 22°

9300

1012.0

> 231-5

Dumas and Peligot.

)

1-063

7533

7087

Mollerat.

12353 at 12°

5767

466-8

} 2449

Liebig.

AND THEIR ATOMIC WEIGHTS.

Acetate of the oxide of ethyl

A-fEO

Acetate of the oxide of methyl

A + MeO

Hydrated acetic acid

E + (03 + HO) -H2
Hydrated formic acid ....

Me + (03 + H O) - H2

It is easy to see that by this law, if we know the specific
weight of the hydrated acid, we may calculate the specific weight
of the combinations of the acid with the oxide of ethyl and
methyl, and further, that from the specific weight of an oxide of
ethyl compound, we may calculate the specific weight of the
corresponding compound of the oxide of methyl. If, for instance,
to the equivalent volume of a hydrated acid we add the number
534, and divide the equivalent weight of the corresponding ethyl
compound by this sum, we obtain the calculated specific weight,
thus, the equivalent volume of hydrated acetic acid is 708*7, and
adding to this 534, we have 1242-7, with which we must divide
the equivalent weight of the acetate of oxide of ethyl ; 12427
gives 0-8906 as the calculated specific weight, which perfectly
coincides with 0*89 the number experimentally determined.
Or: the equivalent volume of the acetate of oxide of ethyl is
1243*5, from which, if we subtract 234 there remains 1009*5;
with which we must divide the equivalent of the acetate of
oxide of methyl, 930; the quotient representing the specific weight
is 0*921, while the number actually observed is 0*919. Kopp
even regards the difference as only approximative, since in their
estimation, the influence exercised by the temperature on the
specific weight and on the equivalent volume has not been
accurately considered.

At a subsequent period, Kopp examined more accurately the
manner in which the temperature is to be considered in the
determination of the equivalent volume, and found, as also did
Schroder, that the equivalent volumes of liquids should be compared
at those temperatures at which their vapours possess equal te?isio//.
Such temperatures are named corresponding temperatures.
The boiling-points of fluids under the same atmospheric pressure
are, for instance, corresponding temperatures in relation to their

32 THE VOLUMES OF BODIES

equivalent volumes, but in relation to the elasticity of their
vapours, the corresponding temperatures are not so high.
The followinp; considerations will render this statement more
obvious : —

The equivalent of ether is 462*5 ; its specific weight 0724 at 12°*5 C.
„ ,, water is 112*5; „ „ 1*000 at „

„ „ alcohol is 575; „ „ 0*7923 at 17° '8

Hence we calculate: —

The volume of 1 eq. of ether at 12°*5 = 638

»•

112-5} 7505

water „

alcohol at 17°*8 726*0

If ether and water form alcohol without the occurrence of
condensation, both in the state of vapour and that of a liquid,
then the equivalent volume of the alcohol must be equal to the
sum of the equivalent volumes of the ether and water. The
equivalent volume of alcohol at 17°*8 C. is as we have seen
726 ; but the sum of the equivalent volumes of ether and water
at 55°'5 is 750. The case would be very different if we con-
sidered the equivalent volumes of these three bodies at cor-
responding temperatures, namely at the boiling-point.

According to Gay-Lussac when the barometer stands at
[0m 76] 29-8 inches,—

The boiling-point of ether is

35°*7C. ...

. 963° F.

„ „ water

100°

212° F.

„ „ alcohol

78°*4

173° F.

Gay-Lussac has also examined the change of volume which
these substances undergo when exposed to heat: from these in-
vestigations it follows that —

The specific weight of ether at the boiling-point (35°*7) is 069739
» » water „ „ (100°) „ 0*96064

„ „ alcohol „ „ (78°*4) „ 0*73869

If now we calculate the equivalent volume of these bodies
at the boiling-point, we obtain:

The equivalent volume of ether at the boiling-point = 4| 2 3 = 663*3]

water „ „ =^^ = 117-l/78° 4

» » alcohol „ „ = ^^ = 7800

AND THEIR ATOMIC WEIGHTS. 33

And we know that at the boiling-point the equivalent volume
of alcohol is in point of fact exactly the same as the sum of
the equivalent volumes of its constituents, ether and water.

But this also holds good for other corresponding tempe-
ratures. Half the tension for which the volumes are cal-
culated, namely the tension corresponding to a column of
mercury of 0m 38, causes ether to assume the state of
vapour at 20o,7, water at 81°*7, and alcohol at 61°*4; these,
therefore, are corresponding temperatures for ether, water,
and alcohol; they lie 15°, 18°*3, and 17° respectively below
the boiling-points of these fluids, and we assume generally that
those are corresponding temperatures which lie equally distant from
the boiling-point, if they are not determined by direct experi-
ment or by calculation. At a temperature of 20°*7, the
equivalent volume of ether is 647*7; at 81°*7 that of water
is 115*7; hence the sum is 763*4; but the equivalent volume
of alcohol at 61°*4 is 764*7, and therefore equal to this sum.

According to Schroder the equivalent volumes of fluids
stand to one another in the relation of whole numbers. Thus,
according to him:

The equivalent volumejof ether at the boiling-point being 663'3,which = 17 x 39*03
„ „ water „ „ 117-1 = 3x3903

„ „ alcohol „ „ 780*0 =20x3903

the numbers 17, 3, and 20 express the relative relation
of the equivalent volumes of ether, water, and alcohol, and
thus the equivalent volumes of all fluids at the boiling-point
must be simple multiples of the number 39*03. Kopp has
examined the value of this theory in the Annalen der Chemie
und Pharmacie, 1844, bd. 50, s. 115, to which I must refer
my readers.

We have already seen that the ethyl compounds have
usually an equivalent volume greater by 234 than the corres-
ponding methyl compound. Now since ethyl = C4 H5, and
methyl = C2 H3, ethyl is distinguished from methyl by an
excess of C2 H2, and to this association of elements we must
ascribe the excess of the equivalent volume (234) occurring in
its compounds. But in the different ethyl and methyl com-
pounds which consist only of carbon, hydrogen, and oxygen,
all the equivalents (atoms) C, O, and II, must have constant

34 THE VOLUMES OF BODIES

equivalent volumes (atom-volumes), otherwise the constant
difference of composition, C2 H2, would not give rise to a
constant equivalent volume. Now in order to be able to
determine the equivalent volumes of C, H, and O, for all
corresponding temperatures, we must have three fluids con-
sisting of these elements, for which the equivalent volume
at any distance from the boiling-point is accurately known.
Water, ether, and alcohol are fluids of this nature. Kopp
has, however, explained why it is that from these three fluids
we cannot deduce the three unknown magnitudes. (Annal. d.
Chem. und Pharm., 1844, bd. 50, s. 78.) The consideration
of these points leads, however, to the very important result
that H+O (water) and 4C + 5H + 0 (ether), taken together
at corresponding temperatures occupy the same space as
4C + 6H + 20 (alcohol), although of the two first fluids one
equivalent in a state of vapour occupies two volumes; while
one equivalent of alcohol in a state of vapour occupies four
volumes. Hence, for C, H, and O, at equally great distances
from the boiling-point, we have the same equivalent volumes
in a liquid which is condensed from two volumes of vapour,
as in another which is condensed from four volumes. At
corresponding temperatures C, H, and O have the same equi-
valent volumes in all fluids, even if when converted into vapour
they exhibit different degrees of condensation (Op. cit. s. 79).
An assumption led Kopp to a mode by which he might, at
least approximately, determine the equivalent volumes of
C, H, and O, when in combination; namely, the hypothesis
that the equivalent volume of a compound does not undergo any
considerable change, probably none at all, if one equivalent of
oxygen be substituted in it for one equivalent (2 volume-atoms) of
hydrogen, and the measurement of the equivalent volumes of
the compounds be always determined at an equally great
distance from the boiling-point, (or more correctly, at those
temperatures at which there is an equal tension of the
vapour).

Alcohol (C4 H6 02) and acetic hydrate (C4 H4 04) differ
in their formulae by the latter containing 2 eq. of oxygen
more, and 2 eq. of hydrogen less than the former; in other
words, 2 eq. of hydrogen of the former are replaced in the

AND THEIR ATOMIC WEIGHTS. 35

latter by 2 eq. of oxygen. Now C4 H4 04, representing
750 grammes of acetic hydrate, the specific gravity at 16°
being according to Mollerat 1063, occupies a space of 705 c. c,
at a temperature which is distant 104° from the boiling-point of
acetic acid, (120° C., Liebig); but C4 H6 02, representing 575
grammes of alcohol, occupies a space of 695*4 c. c, at a tempe-
rature which is also 104° from its boiling-point (78°*4 according
to Gay-Lussac), that is to say, at — 25° *6, and hence it follows
that acetic hydrate and alcohol at corresponding temperatures
have a similar equivalent volume, since the equivalent volume
of 2 H is equal to that of 20.

If we represent the equivalent volume of oxygen by (O),
and that of hydrogen by (H), then from what we have already
stated (0)=(H). But from observations it follows that the
equivalent volume of water (H O) at its boiling-point is 117,
and hence it follows that (H) =5S'5} and (O) =58*5, in liquid
combinations at their boiling-points.

The equivalent volume of carbon (C) can only be found
in the following manner. In accordance with observations,
the equivalent volume of ether (C4 H5 O) at its boiling-
point is 663, which is the sum of 4(C) + 5(H) + (0). But
from what has been already stated, 5(H) + (0)=5 x 58*5
+ 58*5=351, and this subtracted from 663 gives 312 as the
value of 4(C). Hence (C)=78. This is the equivalent
volume of carbon in liquid combinations at their boiling-points.
We arrive at the same number if, from the equivalent volume
of alcohol at its boiling-point, 780, we subtract 6(H) + 2(0),
or 468, and divide the remainder by 4.

For the boiling-point we have, therefore, the following
equivalent volumes,

(C)=78, (H) = 58*5, and (0) = 58*5.
It is of assistance in calculations to regard the equivalent
volumes of carbon, hydrogen, and oxygen as products having
a common factor. The numbers we have just found may be
very simply expressed in the following manner.

(C)=78 = 8 x 9*75; (H) and (0) = 58'5 = 6 x 9*75.
And hence, at the boiling-point the equivalent volumes of
carbon, hydrogen, and oxygen stand in the simple relations
of 8 : 6 : 6.

36 THE VOLUMES OF BODIES

We must now consider the question, does this relation hold
good for corresponding temperatures below the boiling-points,
and does the same or a different relation hold good for all
corresponding temperatures under the boiling-point? or (in
other words) have C, H, and O, in liquid compounds, a similar
or a different dilatability? Kopp confesses his inability to
answer this question.

The assumption that the given relation holds good at all
corresponding temperatures is supported by no analogy; in
the metals in their isolated state this regularity is not observed,
if, in their case, we regard those as corresponding temperatures
which are equally distant from the fusion-point. In order
that this regularity might occur, it would be requisite that
every simple metal, with nearly the same fusion-point, should
exhibit a similar capacity of expansion between 0° and 100°,
which, however, is not the case.

This assumption, however, which Kopp in the most decided
manner explains as perfectly arbitrary, permits the experi-
mental approximate determination of the equivalent volumes
of C, H, and O, and leads to a formula by which the density
of many liquid compounds may be expressed with remarkable
accuracy.

We may by this assumption regard the diminution of the
equivalent volumes of C, H, and O as a diminution of the
common factor, 9*75. The amount by which this factor dimin-
ishes for a temperature distant by a definite number of degrees,
D°, from the boiling-point, has been attempted to be deter-
mined by Kopp, in the following manner. According to
Gay-Lussac the equivalent volume of ether, which boils at
35°*7, is at — 19°*3==612*9; this temperature is 55° from the
boiling-point; hence, in this case, D=55. The equivalent
volume of ether at the boiling-point calculated from the equi-
valent volume of C, H, and O: C4H50: is 4 x 78 + 5 x 58*5
+ 58=663. The equivalent volumes of C, H, and O must,
therefore, at the boiling-point and at —19° -3, stand to one
another as 663 : 612*9. Hence we have,

663 : 612-9 :: 78*0 :x; x=72\ (C)
663 : 612-9 : : 58-5 : x; #=54-08 (H)
663 : 612-9 : : 58*5 : x; #=54*08 (O)

AND THEIR ATOMIC WEIGHTS. 37

Hence, at — 19°*3, C has an equivalent volume of 72*1, while
that of H and O is 54*08. These numbers may, however,
be expressed as 8 x 9*01 and 6 x 9*01 ; and hence for a depres-
sion of temperature of 55° (d = 55).

The common factor undergoes a diminution from 9*75 to
9*01, or has become 0'74 smaller; and hence for one degree
the change amounts to °5^4, or 0*0134, and the equivalent
volume of carbon in a compound, at D° below its boiling-point,
is expressed by the formula, (C)=8x(9*75 — dx 0*0134); and
the equivalent volumes of hydrogen and oxygen by the formula
(H) and (O) =r 6 x (9*72 -dx 0*0134.)

From the equivalent volume of alcohol at the boiling-point
and at 3°*4, we calculate that the number by which D must
be multiplied to express the change of the constant factor
is 0*0108.

The boiling-point of alcohol is 78°*4, and the equivalent
volume, calculated according to the formula C4 H6 02, is 782,
or 4x78 + 6x58*5 + 2x58*5; at 3°*4 (d is consequently in
this case 75°), the equivalent volume is 715*5. From a mean
of these and other observations, Kopp determines this magni-
tude at 0*01, and hence the equivalent volumes in which
carbon, hydrogen, and oxygen are contained in a compound
at d° below the boiling-point, are expressed by

In a compound which boils at 80° the equivalent volume of
carbon at 60° (or 20° below the boiling-point) =8 x (9*75-20
X 0*01)^:8 x 9*55=76*4; in a compound which boils at 137°,
the equivalent volume of oxygen at 13° = 51*06. d is therefore
in this case 137° - 13°= 124°, and hence we have 6 x (9*75
— 124x0*01). Hence, generally, the equivalent volume of a
liquid compound consisting of a equivalent of carbon, b of
hydrogen, and c of oxygen, at a temperature which is D° from
the boiling-point of the compound is

(8« + 6b + 6c) x (9*75 - 0*01 d)
The following illustrations tend to show how far this formula
gives results coinciding with those yielded by observations.

38 THE VOLUMES OF BODIES

Let us calculate the equivalent volume and the specific gravity
of carbonic ether for 19°. Carbonic ether boils at 126°; D is
therefore in this case =126°— 19°= 107°. As carbonic ether
is expressed by the formula C5 H5 03, its equivalent volume
at 19° must be, according to the above formula,

(5 x 8 +5 x 6 + 3 x 6) x (975 - 0-01 X 107°)=88 x 8-68 = 764
and since the equivalent weight is 737*5 the specific gravity
at 19°= 7736745 = 0-97. Ettling has found it to be 0-975.

For hydrated caryophyllic acid, C20 H12 04, whose equi-
valent is therefore 2050, and whose boiling-point is 243°, the
above formula gives the specific gravity 1*08 at 8° C, while
Ettling actually found it to be 1*079.

If one of the three elements is altogether absent in the
compound we must proceed as if there were 0 equivalents
present. A compound, C2H, which boils at 85°, has at 15°
the equivalent volume (2 x 8 + 6) x (9*75-0*01 x 70)= 199 ;
and since the equivalent is 162*5, the specific gravity at 15°
is 119295, or 0*82. Kopp has not only endeavoured to deter-
mine the equivalent volumes of carbon, hydrogen, and oxygen,
but also of chlorine, nitrogen, and sulphur.

In relation to these various points I must refer to page 98
of his excellent memoir, and will only add that the following
values correspond the best with the results of actual obser-
vation.

(CI) = 28 X (9*75 - 0-01 d)
(N) = 18 X (9'75 - 0*01 d)
(S) = 15 X (9*75 - 0-01 d)

In order to facilitate the calculations, the equivalent volumes
which are assumed for C, H, O, CI, N, and S, in a liquid com-
pound at a temperature of D° below the boiling-point, have
been collected by Kopp into a table which I here transcribe
without changing the old equivalent weights of CI, N, and S.

AND THEIR ATOMIC WEIGHTS.

HandO

Equiv. 75.

Equiv. = 125.
Equiv. = 100.

Equiv. = 4-12-6.

Equiv. = 176.

Equiv. = 201.

78-0

58 5

273

175-5

146-25

77*2

57'9

2702

173-7

144-75

76'4

57*3

267-4

171*9

143-25

75-6

567

264-6

170-1

141-75

74*8

56-1

261-8

1683

140-25

74 0

555

2590

1665

13875

73-2

549

256-2

164 7

13725

72-4

543

253-4

162-9

135*75

71*6

53'7

2506

1611

134-25

70'8

531

247 8

159-3

132-75

100

70-0

52-5

2450

1575

131*25

110

692

519

242-2

155-7

12975

120

68'4

51-3

2394

153-9

12825

130

6) -6

50*7

236-6

152- 1

126-75

140

66*8

50*1

233-8

150 3

12525

150

66 0

49*5

231-0

148-5

123-75

160

652

48-9

228-2

1467

122-25

170

64'4

48*3

225-4

144-9

120*75

180

63-6

47'7

222-6

1431

11925

190

62-8

47 1

219-8

141-3

11775

200

62*0

465

217-0

1395

11625

210

61*2

45-9

214-2

1377

11475

220

60-4

453

211*4

135-9

113-25

230

59-6

44*7

208-6

1341

11175

240

58*8

441

205-8

132-3

11025

250

58-0

43-5

2030

130-5

108-75

260

572

429

200-2

128-7

107-25

270

564

423

1974

126-9

10575

280

55*6

41-7

194-6

1251

104-25

290

54-8

41-1

191-8

1233

10275

300

54'0

40*5

189*0

121-5

101-25

The application of this table hardly requires explanation.
Let us suppose, by way of illustration, that we employ it to
determine the specific gravity of chlorobenzide (Mitscherlich)
at 7° C. Chlorobenzide is C12 H3 Cl3; its equivalent is 2265;
its boiling-point is 210° C, and hence d = 210-7, or 203. If,
for the sake of convenience, we make Dm 200, we calculate
the equivalent volume as 12 x 62 + 3 x 46*5 +3 x 217, or 1534*5,
and if we divide the equivalent weight by the equivalent
volume we obtain for a quotient 1*476 as the specific gravity.
Mitscherlich found it at the given temperature, 7° C, to be 1 *457.
Or, suppose we have to calculate the specific gravity of mer-
captan at 21° C. Mercaptan, the hydrosulphide of the sul-
phide of ethyl, is represented by the formula C4 H6 S2, and its

40 THE VOLUMES OF BODIES

equivalent is 777; its boiling-point has been fixed by Liebig at
36°*2C. In this case d is 36°*2-21°, or 15°*2 C. If D-15
the equivalent volume of mercaptan is 4 x 76*8 + 6 X 57*6 + 2
x 144=941; and if we divide the equivalent weight by this
number we obtain 0*826 as the calculated specific gravity,
while Liebig actually found it to be 0*835.

To this table Kopp appends several interesting remarks.
He thinks that it may be assumed with equal probability that
the equivalent volumes which pertain to the elements in
combination are the same as those elements possess in an
isolated state; for instance, that sulphur in a liquid combina-
tion at 100° below the boiling-point is contained in the same
equivalent volume which it has when isolated, in a liquid
state, at 100° below its boiling-point. If we have already
found that the elements when in combination have a different
equivalent volume from that which they have in an isolated
state, we must to this condition add that the equivalent volumes
of the elements are not determined for corresponding tem-
peratures. It is very probable, according to Kopp, that for
solid bodies those temperatures are. approximatively corres-
ponding which are equally distant from the fusion-point. We
say, for instance, that the equivalent volume of potassium in
combination, as in its sulphates, is different from that which
we observe for potassium in the isolated state, since the equi-
valent volume of sulphate of potash is smaller than that of
potassium. We have not at present the means of accurately
determining this point, since we are only acquainted with the
equivalent volumes of potassium and its compounds at mean
temperatures. In all probability the equivalent volume of
potassium is the same in combination as in its isolated state,
for equal distances from the fusion-points. If we could observe
the equivalent volume of potassium at a degree of temperature
which is as far distant from the fusion-point of this metal as the
difference of temperature between the fusion-point of sulphate
of potash and the temperature at which we observe the equi-
valent volume of this salt, we should find that the equivalent
volume of potassium is much smaller than that of the salt,
while the observations which are instituted for one and the
same, for the mean temperature, give the very reverse. The

AXD THEIR ATOMIC WEIGHTS. 41

equivalent volume which we assign to potassium in its sulphates,
and which is different from that which it has been observed
to have in its free state, will approximate to that which the
potassium possesses at a degree of temperature as far distant
from its fusion-point as the temperature of the fusion-point
of sulphate of potash is from the temperature at which the
specific gravity of the salt was observed. According to Kopp
strong objections may be brought against this mode of obser-
vation, which cannot be removed until the fusion-point and
the expansion of many inorganic substances have been accu-
rately determined.

If the formula that we have given for the calculation of the
equivalent volumes of C, H, O, CI, N, and S at equal distances
from the boiling-point of their compounds were absolutely
correct in regard to their form, and in reference to all the
numerical calculations contained in it, it might then, in addition
to its principal object, also serve, by the known atomic com-
position and a specific gravity observation at a known tem-
perature, to lead to the determination of the distance of the
observed temperature from the boiling-point, or, in other words,
of the boiling-point. But since the formula cannot be regarded
as undoubtedly certain, and the estimation of the numbers
contained in it is not altogether to be depended on, it would
be a matter of considerable hazard to attempt to determine
the boiling-point by this means. (Kopp.) The reasons of the
uncertainty of the formula are thus explained by Kopp. The
form of the formula depends on the assumption: (1) that the
equivalent volumes which are taken for C, H, O, CI, N, and S
in liquid combinations, always stand in the same relations to
one another at corresponding temperatures. This assumption
is certainly the simplest, but still it is a perfectly arbitrary
one; (2) that those should be regarded as corresponding tem-
peratures which are equally distant from the boiling-point.
This assumption, as far as actual practice is concerned, gives
results which, in most cases, approximate pretty closely with
the truth, but, strictly speaking, it is not correct, and conse-
quently leads to false theoretical results, as may be seen by the
following remarks.

According to the formula the contraction which a liquid

42 THE VOLUMES OF BODIES

undergoes on cooling x° from the boiling-point, is always pro-
portional to the magnitude x. Experience, however, most
decidedly shows us that in by far the greater number of liquids
this is not the case; for a definite interval of temperature the
contraction is almost always smaller, the more distant this is
from the boiling-point. Moreover, according to the formula,
all liquids, if we assume for them equal volumes at the boiling-
point, at equal distances from the boiling-point occupy equal
volumes. If we assume that the volume of any liquid at its
boiling-point is 9*75, then at a distance of D° from the boiling-
point it will be 9*75 — 0*01d. Moreover this conclusion is not
confirmed by experience, but is much oftener opposed by it,
if, for the numerical determination 9*75 and 0*01, other magni-
tudes are assumed as more correct.

That the numerical determinations in the formula are
uncertain is a fact not concealed by Kopp, in various parts
of the treatise to which we have so often referred; he adds,
in relation to the point, the assumption that the equivalent
volume of hydrogen is equal to that of oxygen, the basis of all
our determinations, is far from being so confirmed by direct
observations as to be regarded entirely accurate; it is possible
that the relation of the equivalent volumes is nearly, without
being perfectly, the same, just as in the estimation of the
equivalents of elements simple relations are often tolerably
approximative, but after the most accurate determinations are
not positively correct.

The practical applicability of the formula appears from a
large number of results to be well established. The formula
gives the specific gravity of liquid compounds for any distance
of temperature from the boiling-point with an accuracy that
deserves attention.

Aubergier has ascertained empirically that the volatile
oils isomeric (polymeric) with oil of turpentine have the same
specific gravity at temperatures equally distant from their
boiling-points. The formula leads to the same result, since it
requires that all liquid substances of equal empirical atomic
construction (an equal per-centage composition) should have
the same specific gravity at temperatures equally distant from
the boiling-point. Hence, at the ordinary temperatures at

AND THEIR ATOMIC WEIGHTS. 43

which observations are made, the specific gravity of polymeric
substances boiling at high temperatures must be greater than
that of those with lower boiling-points.

The formula may also serve for the estimation of the
density of the vapours of a compound. The latter is done
in order to obtain a check in ascertaining the relation in which
the atoms of the elements unite to form a compound, and at
the same time as a means of knowing how many elementary
atoms are contained in one atom of the compound. Our
formula, in its present state, can only be used as a check.

In addition to the knowledge of the atomic relations in
which the elements unite to form a combination, the deter-
mination of the boiling-point is sufficient to enable us to
calculate the specific gravity of the liquid, and the observation
with which this calculation is compared is easily and simply
made. But the formula also affords the means of ascertaining
the atomic weight of a liquid compound and of determining
how many elementary atoms are embraced in one atom of
the compound, if the magnitudes contained in it are so accu-
rately determined, that the boiling-point enables us indirectly
to ascertain the composition, and the density at a given tem-
perature. Ldwig believes that it may be mathematically
demonstrated, (1) that the values assumed by Kopp as the
equivalent volumes of carbon, hydrogen, and oxygen, in liquid
organic compounds are false, and (2) that the equivalent
volume of an element is not, as Kopp believes, equally great
in all combinations. (Pogg. AnnaL, 1846, Bd. lxviii., s. 51, &c.)
We have already mentioned that the compound molecule C2 H2
has in its combinations a constant equivalent volume, which,
if we compare the equivalent volumes of the ethyl and methyl
compounds without reference to correspondence of temperature,
amounts to about 234. If we take, as a foundation, the equi-
valent volumes established by Kopp for carbon and hydrogen
in combinations at their boiling-points, namely 78 and 58*5,
the equivalent volume of this compound molecule is 2x78 +
2 x 58*5, or 273. Lowig now ascertains the equivalent volume
of C2 H2 directly for the boiling-point, by establishing a com-
parison between the combinations which differ in their com-
position by C2 H2, or by a multiple of it, in reference to their

44 THE VOLUMES OF BODIES

equivalent volume at the boiling-point, and arrives at discrepant
results. Since the contraction that occurs on cooling is not
accurately established for other liquids here to be considered,
as it is for water, alcohol, and ether, he compares, in this point
of view, the compound ethers, wood spirit, and fousel oil with
alcohol, but acetone and the simple oxides with ether*, in order
to calculate the equivalent volumes for the boiling-point. The
following table contains the data for these comparisons :

« p Below the Equiv. Vol.

" * ' Boiling point. Atomic Vol.

0*6974 0° 663

0-925 37° 458

0*739 0° 780

0-9085 34° 1018

0-890 59° 1236

0-7938 40° 503

0-8137 118° 1351

Valerianate of oxide of methyl, C12H1204 .... 0*8806 100° 1646

If now we proceed to calculate the equivalent volumes of
these compounds for the boiling-point in accordance with the
contraction of alcohol, we obtain the following values: —

Ether

.... C4 H5 0

Acetone

.... C2 H3 0

Alcohol....

... C4 H6 02

Formiate of oxide of ethyl

•••• C6 H6 O4

Acetate of oxide of ethyl

-• C8 H8 O4

Wood-spirit

.... C2 H4 02

Fousel oil

.... C10H12O2

Equiv. Volume.

Ether

... C4 H5 0 ..

663

Acetone .... ....

... C2 H3 0 ..

480

Alcohol

- C4 H6 O2 ••

780

Formiate of oxide of ethyl ....

... C6 H6 04 ..

1073

Acetate of oxide of ethyl ....

... C8 H8 04 ..

1323

Wood-spirit

... C2 H4 02 ..

528

Fousel oil ....

... C10H12O2 ..

1529

Valerianate of oxide of methyl

"• C12H12 04 •

1832

If we subtract the equivalent volume of one of these
compounds from that of another differing from it by C2 H2,
or some multiple of C2 H2, the difference naturally indicates
the equivalent with which C2 H2, 2C2 H2, &c, are contained
in the latter compound. For instance:

* According to Gay-Lussac the following diminution occurs in 1000 parts :—

Water. Aleohol. Ether.
From 0° to 15° from the boiling-point downwards 10*50 .... 17*51 .... 2423
„ 15° to 30° „ „ 8*35 .... 17*23 .... 2209

„ 30° to 45° „ „ 7-65 .... 16*11 .... 19*06

„ 45° to 60° „ „ 5*92 .... 15*11

„ 60° to 75° „ „ 4*26 .... 14*15

AND THEIR ATOMIC WEIGHTS. 45

Eqniv. Vol.

Alcohol C4 H6 02 = 780

Wood-spirit C2 H4 02 = 528 C2 H2 = 252

C2 H2 = 252

Valerianate of oxide of methyl C12 H12 04 = 1832

Acetate of oxide of ethyl .... Cs H8 04 = 1323 C2H2=^=254

C4 H4 = 509

Fousel oil C10H12O2 = 1529

Alcohol C4 H6~ 02 = 780 C2 H2 = ^-9 = 250

C6 H6 = 749

'10 Hl<

Wood-spirit C2 H4 02 = 528 C2 H2 = ?-^ = 250

Fousel oil C10H12O2 = 1529

C8 Hs = 1001

Hence, in this manner we obtain for C2 H2, the constant
volume 250, while, according to Kopp, it amounts to 273. As-
suming that the former number is the correct one, we obtain
for H O the following volumes:

Eqniv. Vol.

Alcohol .... C4 H6 02 = 2C2H2 + 2HO = 780

-2C2H2 =500 HO = ^j-° = 140

2H O = 280

Wood-spirit.... C2 H4 o2 = C2H2 + 2HO = 528

- C2H2 =250 HO = ?-p = 139

2H O = 278

Fousel oil .... C10H12O2 = 5C2H2 + 2HO = 1529

-5C2H2 =1250 HO=^=139'5

2H O = 279

For the equivalent volume of oxygen we obtain the fol-
lowing values:

Equiv. Vol.
Acetate of oxide of ethyl .... Cs Hs 04=4C2 H2 + 04=1323

-4C2H2 =1000 0=^=80-8

04= 323

Formiate of oxide of ethyl .... C6 H6 O4=3C2H2+O4=1073

-3C2H2 = 750 0=^=80-8

04= 323

46 THE VOLUMES OF BODIES

Equiv. Vol.

Valerianate of oxide of methyl C12 H12 04=6C2 H2+ 04= 1 832

-6C2H2 =1500 0=^-2=83

04= 332

Kopp's values, however, are, for water 117, and for oxygen
58*5. Such coincident results for the volumes of C2 H2, H O,
and O, as are given above are, however, according to Lowig,
only obtained when we happen to institute comparisons between
compounds with similar rational formulas. If we compare
combinations of different rational formulas, we obtain different
values; thus —

Equiv. Vol.

Fouseloil C10H12O2 = 1529

Acetate of oxide of ethyl .... C8 H10O2*= 1323

C2 H2 = 206 instead of 250

2 eq. Ether C8 H10O2 = 1326

1 eq. Wood-spirit C2 H4 02 = 528 C2H2 = £§-i = 266

C6 H6 = 798

Ether C4H50 = 2C2H2 + HO = 663

- 2C2H2 =500

HO = 163 instead of 117

Acetone C3H30 = 1£C2H2 + O = 480

- 1£C2H2 = 375

O = 105 instead of 80*8

If we subtract the equivalent volume of ether (663) from
the equivalent volume of alcohol (780), there remains the
number 117 (already determined) for HO. If, from the
formula for alcohol 2C2H2 + HO, we subtract 500, as the
number for C4 H4, there remains, as has been already shown,
280 for 2HO. If the volumes of the equivalents of oxygen
and hydrogen were equal, they must amount to 2 x 117, or 234.
But 280 — 117 = 163, and, as we have clearly seen, H O actually
occurs in ether with the volume 163. Hence, according to

* Since, according to Kopp, hydrogen and oxygen have an equal equivalent
volume, O may be arbitrarily substituted for H. The formula for acetic ether,
C3 H8 04, is in this way converted into Cs H10 02.

AND THEIR ATOMIC WEIGHTS.

Lowig, it is completely proved that if we separate alcohol
into 2C2 H2 + 2H O, the one equivalent of water must occur
in the compound with the volume 117, and the other with the
volume 163.

If we assume that the equivalent volume of H is repre-
sented by 58*5, and C2 H2 is taken as equal to 250, we obtain
for C the volume 66'5 in place of 78, for 250-117=133,
and i|^=: 66-5.

On comparing with one another the compounds which have
different rational formula? we obtain for C and H the most
discrepant values.

Ether

Equiv. Vol
.... C4 H5 0 = 663

"Wood-spirit ....

.... C2 H4 02 = 528 (

C2 = 135

Alcohol

.... C4 H6 02 = 780

Acetone

.... C,H,0 = 480

CH,0= 300

c _ ljLi _ 67.5

-H,0 = 234 C = 66

4 equiv. Acetone
1 equiv. Fousel oil

C12H1204= 1920
C10H12O2 = 1529

C2 02= 391
- O, = 117

C =

274

= 137

= 274

4 equiv. Acetone
■ 1 equiv. Acetic ether

C12H1204= 1920
C3 Hs 04 = 1323

C4 H4 = 597 instead of 500
— H4 =234 C = 3-¥- = 90

= 363

4 equiv. Acetone
■ 1 equiv. Formic ether.

C12H1,04= 1920
C6" H6~ 04 = 1073

= 847 instead of 750

C6 H6

- H6 =351 C = 4JL1 = 82

= 496

48 THE VOLUMES OF BODIES

Fouseloil
—2 eq. Acetone ....

Equiv. Vol.
.... C10H12O2=1529
.... C6 H6 02 = 960

C4 H6 = 569
-C4 H4 = 500

H = ^ = 345

3 equiv. Ether....
— Valerianic ether

H2 = 69

.... C12H1404 = 1989]
.... C12H1204 =1832

H= 1l1 =78«

H2 04 = 157

The illustrations already adduced are sufficient, according
to Lowig, to establish the fact that Kopp's values are not
correct, and that the volumes of an elementary atom cannot
be equally great in all compounds. It is, however, equally
well proved that this is based upon regular laws, since otherwise
we could not always obtain, for a series of cases, the numbers
250 for C2 H2, 280 for 2H O, and 80*8 for O.

The fact, that the specific gravity of the carburetted hydro-
gen generally remains nearly constant, however much the propor-
tion of the equivalents (atoms) of carbon and hydrogen may differ,
and that the specific gravity is always considerably increased
if only a few equivalents of oxygen are combined with it,
suggested to Lowig the idea that the specific gravities of
carbon and hydrogen in organic compounds might be equal.
In order to put this idea to the test, he compared the com-
binations belonging to one group, as, for instance, alcohol,
ether, acetic acid, and acetone; wood spirit and formic acid;
fousel oil and valerianic acid; oil of bitter almonds and benzoic
acid. Ether and acetone were first compared.

Equivalent volume of ether C4H5 0 = 663

„ acetone C3H3 0 = 480

C H„ = 183

If now the specific gravity of carbon is the same as that
of hydrogen, then must |- of the volume 183 be carbon, and
-| hydrogen (in C H2, 2 parts by weight of H occur with 6 parts
by weight of C). If we take the equivalent atom of hydrogen
(=12-5) as our unit, then C H2 = 8 units (since 8 x 12*5 = 100

AND THEIR ATOMIC WEIGHTS.

parts by weight). The space occupied by the unit of weight
(12*5) is therefore found on dividing 183 by 8, and =22*87.

Ether contains 24 units by weight of carbon and 5 of
hydrogen, =29, and acetone contains 18 of carbon and 3 of
hydrogen, =21.

If we multiply the unit of space (22*87) by 29 and 21
we obtain the exact equivalent volumes of ether and acetone:

29 x 22*87 = 663, and 21 x 22*87 = 480
winch proves, according to Lowig, that the equivalent volumes
of ether and acetone are equal to the space which belongs
to their radicals, and that a condensation occurs which is
equal to the equivalent volume of oxygen. 1 equivalent
volume of ether (663) + 1 equivalent volume of water (117)
is equal to the equivalent volume of alcohol (=780). Now
wood-spirit and fousel oil have a similar chemical constitu-
tion to alcohol. On subtracting 117 for HO from the equiva-
lent volume of these combinations, we obtain the equivalent
volume for oxide of methyl, C2 H3 O, and for oxide of amyl,
C10Hn O.

The equiv. vol. of wood-spirit = 528
„ for H O = 117

C2H30 = 411

The equiv. vol. of fousel oil = 1529
„ for H O = 117

C10HnO = 1412

In accordance with ether and acetone the space 411 must
correspond to the volume of C2H3, and the space 1412 to the
volume of C10Hn. Now, 2^lg7 = 18 units of space, and
1 ,* 3 ^ - 62 U.S. But C2H3=15 units of weight, and
C10 Hn = 71 u. w.; hence, in methyl 15 u. w. = 18 u. s., and
in amyl 71 u. w. = 62 u. s.

Hence, in methyl there are 3 more units of space than there
are units of weight. Methyl contains 3 eq., and, according to
Lowig, it is, at all events, most simple to assume (and indeed
any other assumption would be impossible) that in methyl the
equivalent volume of hydrogen is as large again as in ether and
acetone. If we suppose that this expansion of the hydrogen
also takes place in fousel oil it follows that the volume of the
carbon in fousel oil is one-third more condensed than in ethyl
or methyl. Hence we obtain:

THE VOLUMES OF BODIES

u s

^10

u s

H„

rr:

^2^3,

CioHn,

Hence the volumes of the atoms of carbon and hydrogen
are to one another in ethyl as 6 : 1 ; in methyl as 3 : 1, and in
amyl as 2 : 1 ; and if it is assumed that in methyl the atoms
of carbon and hydrogen are in their normal state, and if, for
the unit of space, we take 22, as a number easily divisible,
then —

(1.) The atom- volume of hydrogen = 44
(2.) „ „ carbon = 132

And the condensations are —

\ for the hydrogen atom = 22 = 1 atom-volume*
\ „ carbon atom = 88 = 1 atom- volume

With 6 and 4 units of space for the equivalent of carbon, and
2 and 1 for the equivalent of hydrogen, the calculated specific
gravity, according to Lowig, exactly accords with the result
of experiment, if we base it upon Gay-Lussac's contraction.
According to him it is so far mathematically established that
the elements in combinations have different volumes.

According to Lowig, alcohol and the acetic hydrate have
an equal equivalent volume, and so also have wood spirit and
the formic hydrate, and fousel oil and the valerianic hydrate.
Since, in the formation of acetic acid by the oxidation of
alcohol, when 2 atoms of hydrogen are abstracted from the
ether of the alcohol, the remaining atoms do not change their
volume, the condensed atom of oxygen remains unaltered.
From C4 H5 O (ether) we have C4 H3 0 (oxide of acetyl).
But on oxidizing C4 H3 O to C4 H3 03, that is to say, to acetic
acid, this additional atom of oxygen corresponds to 1 unit of
space.

H5
O

24 u s

5 „
0 „

Ether

C4
H3
O

24 u s

3 „
0 „

Oxide of acetyl

* In some combinations the condensation f (= 33,) also occurs. {Pogg.
Ann., bd. lxiv. s. 292.)

AND THEIR ATOMIC WEIGHTS. 51

Oxide of acetyl = 27 u s
Oo = 2 „

Acetic acid = 29

If this oxidation takes place in compounds in which the atom-
volume of hydrogen = 2 u. s, the atom of oxygen which enters
the compound corresponds to 2 u. s.

C2 = 12 u s

H3 = 6 „

O = 0 „

Oxide of methyl = 18 „

C2 = 12 u s

H 2 „

O = 0 „

02 = 4 „

Formic acid = 18 „

It follows from the laws of gaseous combinations that the
gaseous oxide of ethyl unites without condensation with the
organic acids, as with the vapour of water, and coinciding with
the combination of liquid oxide of ethyl with water; there
doubtless occurs the same relation in the union of the liquid
oxide of ethyl with the liquid acids, an assumption which is
in all respects confirmed by the specific gravity of this com-
pound.

The specific gravity of acetate of oxide of ethyl at 64°,
or 10° below its boiling-point, is 0*840, and consequently its
equivalent volume is q.^q? or 13*10. In accordance with
Gay-Lussac's contraction for alcohol at 10° below the boiling-
point, the unit of space has the volume 22*62, and hence we
calculate for acetic ether the unit of space ==^|^|:=: 58 U.S.
The sum of the unit of space is then (29) for ether, and (29)
for acetic acid*.

If we assume for water 5 units of space, which certainly
is not exactly accurate, since 5x22*87=114 instead of 117,
then, according to Lowig, these correspond to the equivalent
volume of

Alcohol C4H50,HO .... 34usand -^ =22*87

Ether C4H50 .... 29 „ and -^ =22*87

Acetone C3H30 .... 21 „ and ^ =22-80

* I will remark that we should reckon 1G units of space instead of 18 for the
formic acid calculated from formic ether. (Lbwig, Pogg. Annul. , C4, s. 51G.)

E 2

52 THE VOLUMES OF BODIES

Formic ether
Acetic ether
Wood-spirit
Fousel oil

E 0, Fo 03 .

... 47 u s and -*£p- = 22-83

4 7

EO, A

... 58 „ andi-J-J-5- = 22-81

Me 0, H 0 .

... 23 „ and *££ = 22-83

Am 0, H 0 .

... 67 „ and J-f-p- = 22-73

Valerianate of oxide of methyl Me 0, Va03 .... 80 „ and ^|^ = 22'8l

The volume of the unit of space is therefore found expe-
rimentally to be of perfectly equal size.

The following paragraphs contain the laws which Lowig
has established for the relations of volume in the combinations
of C, H, and O and the radicals consisting of C and H,
together with the nomenclature he has formed for them, by
which these laws may be generally expressed:

(1.) The radicals consisting of carbon and hydrogen he
divides into two classes:

(a) Into hydroisocarbonyles, or radicals which contain either
an equal number of atoms of carbon and hydrogen, or in which
one element exceeds the other by one atom.

(b) Into hydropolycarbonyles, in which the atoms of carbon
exceed those of hydrogen.

(2.) The Hydroisocarbonyles are divisible, according to the
relation of the atoms of carbon and hydrogen, into

(a) Dyhenyles, or radicals containing one atom more of
carbon than of hydrogen, C2 H, C4H3, C6 H5.

(b) Perisyles, or radicals which consist of an equal but
odd number of atoms of carbon and hydrogen, as C H, C3 H3,

C7 H7.

(e?) Artyles, or radicals consisting of an equal and even
number of atoms of carbon and hydrogen, C2 H2, C4 H4.

(3.) Associated radicals belonging to the same groups form
an ascending series, in which each term contains C2 H2 more
than its predecessor, and the collective terms of a group agree
with one another in their elementary properties.

(4.) The Dyhenyles unite with 3 at. of oxygen to form
organic acids: C.2 H, 03, C4 H3, 03, C6 H5, 03. The Dyotryles
form oxides with one atom of oxygen, which comport them-
selves as bases. The acids of the dyhenyles unite with the

AND THEIR ATOMIC WEIGHTS. 53

oxides of the dyotryles, and both unite with one at. of water
to form hydrates.

(o.) If we distinguish by the symbols A1? A2, A3, &c,
the terms 1, 2, 3, &c, of the series of acids formed by the
union of a clyhenyle with 3 at. of oxygen, or C2 H, 03,
C4 H3, 03, C6 H5, 03; and likewise by the symbols Bp B2, B3,
&c, the corresponding oxides of a dyotryle, or C2H3, O,
C4 H5, 03, and by 1, 2, 3, &c, the hydrates of the dyhenyle
acids, or C2 H, 03 + HO, C4 H3, 03 + H O; it follows from
the relation of the atoms of carbon and hydrogen, that the
combinations Aj+Bj must have the same elementary com-
position as the acid hydrate 2; and, generally, we obtain the
following series of metameric combinations.

A1 + Bx = 2
A2 + Bx = Al + B2 = 3
A3 + Bx = A2 + B2 = A, + B3 = 4
A4 + Bx = A3 + B2 = A2 + B3 = Al + B4 = 5
&c.a &c.

(6.) The atom-volume of hydrogen is to that of carbon
in the

First term C2H, C2 H^ C2 H2 as 2 : 6

Second term C4H3, C4 H5, C4 H4 as 1 : C

Higher term .... .... C8H7, C10Hn, C10H10 as 2 : 4

If the weight of the unit of weight = 12*5 then is

The volume of the unit of space = 22*87 at the boiling-point, and the

Specific gravity of the unit of space = ^ g5 = 0*546 at the boiling-point.

(7.) If we combine the hydroisocarbonyles with 1 at. of
oxygen, there takes place a condensation equal to an atom-
volume of oxygen. Hence, the volume of the radical, when
combined with 1 at. of oxygen, remains unchanged, whilst
an augmentation of the specific gravity corresponding to a
unit of weight of oxygen ensues. If, however, the oxides of
the dyhenyles take up 2 atoms of oxygen, the volume of these
atoms of oxygen is equal to the volume of the atoms of
hydrogen in the compound. Consequently the atom-volume
of oxide of ethyl, C4 II5 O, is equal to the atom-volume of
acetic acid, C4 H3 03 ; and the atom-volume of oxide of amyl,
C10HHO, is equal to the atom- volume of valerianic acid,

54 • THE VOLUMES OF BODIES

C10 H9 03. Hence it follows generally that the dyotryles, their
protoxides, and the dyhenyle acids, formed from these oxides with
3 at. of oxygen, have equal atom-volumes.

(8.) If the acids and oxides of the hydroisocarbonyles
combine with 1 atom of water to form hydrates, the atom-
volume of the compound corresponds with the atom-volume
of the constituents; and hence, at the boiling-point the increase
of volume amounts to 117. In the conversion of wood spirit into
the hydrate of formic acid, of alcohol into the hydrate of acetic
acid, and of amylic alcohol {f ousel oil) into the hydrate of vale-
rianic acid, there is consequently no change in the atom-volume.

(9.) If the dyhenyle acids combine with the oxides of the
dyotryles, then also is the atom-volume of the compound equal
to the atom-volume of the constituents. Since in methyl and
in ethyl the atom-volume of the hydrogen, and in amyl that
of the carbon also is different, it follows that the atom-volume
of metameric compounds of methyl, ethyl, and amyl cannot
exactly coincide with one another. Formiate of oxide of ethyl
and acetate of oxide of methyl, valerianate of oxide of ethyl
and acetate of oxide of amyl, are the only compounds of this
nature which have, on these grounds, an equal atom-volume.

The equivalent or atom-volume of chlorine is fixed by
Lowig at 10*5 units of space from observations ranging
between 8 and 10, (Pogg. Annal., bd. 64, s. 527,) 36 u. w. are
therefore 10*5 u. s., and he thinks that we may assume that
36 u. w. have become condensed to — hr the original bulk.

The specific gravity of perchloride of formyl (chloroform),
at 40° C. below its boiling-point, is 1*48, and since the atomic
weight is 1490, the atomic volume is -^T^* or 10°7> aQd
2^° J =46 units of space.

Now C2 = 2 x 6 =12 units of space
H = 1 X 2 =2 „

Cl3 = 3 X 10-5 = 32 „

C2 H Cl3 =z 46 units of space

For the atomic volume of iodine, as determined from iodide
of ethyl, Lowig obtains 17*46 units of space, and for the atomic
volume of bromine, as determined from the bromide of ethyl,

AND THEIR ATOMIC WEIGHTS. 55

13*8 units of space, and the latter is exactly the mean of the
atomic volumes of chlorine and iodine. The atomic volume
of sulphur he calculates at 2 u. s. and 8 u. s., and that of
nitrogen at 7 u. s.

Rhodanethyl (sulphocyanide of ethyl), E Rn, or E, C2 N S2,
has at 15° C*., that is to say, 80° below its boiling-point, the
same specific gravity as water; hence, as the atomic weight
is 1087*5, the atomic volume is 10817'5, or also equal to 1087*5,
which yields 1^q.78°, or 52 units of space.

Now C2 = 2 x 6 = 12 units of space
S2 = 2 X 2 = 4 ,,

N = 7

E =29 „

E C2 N S2 =52 units of space

From all these experiments Lowig finally draws the con-
clusion: that 1, the atomic volumes of organic combinations
stand in a simple relation to their atomic weights. 2. That
the normal relations exhibited by organic combinations in
their gaseous condition, are only the result of the normal
molecular conditions of solid and liquid combinations.

Time must decide upon the importance of Lowig's laborious
work. He has further developed what Schroder had previously
established regarding the atomic volumes of solid compounds,
namely, that the constituent or ultimate elements of the latter
are contained with a different atomic volume. All the cal-
culations of Lowig rest upon the hypothesis that liquid organic
compounds are expanded by heat in the same manner as
alcohol; and it is therefore singular that the tenability of
these views has not been in any degree investigated.

I cannot leave the subject of the connection between the
equivalents and the volume of bodies in a solid and liquid
condition without making mention of a few ingenious sugges-
tions by Schroder on the relations which, according to this
chemist, exist between the equivalent volume and some few
other properties of bodies. Solid and liquid bodies are, as is
well known, not equally expanded by heat.

* The volume of the unit of space at 80° to 90c below the boiling-poiut.

56 THE VOLUMES OF BODIES

On calculating the expansion of the elements for their
equivalent volumes and not for equal volumes, we obtain
numbers, that appear to be in part equal to each other, and
in part to stand in a simple relation. Schroder is of opinion
that the equivalent volumes of simple bodies suffer expan-
sions by heat which stand in simple relations to each other
when the equivalent volumes themselves have also simple
relations. The equivalent volume of platinum and palladium
is for instance 57, the equivalent volume of lead 114, con-
sequently exactly twice as large; the expansion of the equi-
valent volume of platinum and palladium from 0° to 100° is
0*162*, that of the equivalent volume of lead is 0*974, con-
sequently 6x0*162. The fact that expansions do not stand
in the same relation as the equivalent volumes, gives rise to
some doubt as to the validity of this law. (Pogg. Ann., Bd. Hi.,
s. 282, &c.)

The equivalent volume of water is 112*5; its expansion
from 0° to 100° amounts to 0*045668; the expansion of the
equivalent volume is therefore =5*13, that is, 32x0*16; the
expansion of steam from 0° to 100° amounts, as in the case
of all gases, to 0*366, whence it follows, that the volume
112*5 of a gas expands by 41 from 0° to 100°. This amount
is eight times the expansion of water (8x5*125) and 8x32
times of the expansion of platinum, measured for equivalent
volumes. Schroder is of opinion that the expansions of bodies
are comparable in all conditions of aggregation, provided their
equivalent volumes stand in simple relation to each other.

In connection with this view, we may mention that of
Persoz, according to which an equal volume of gas or a
multiple of that volume can be obtained from an equal
volume of simple and compound bodies in their transition
from the fluid or solid to the gaseous condition, (by which
he finds the hypothetical specific gravities of the gases of
all bodies,) and that the hypothetical specific gravity of the

* We obtain this number by multiplying three times the linear expansion
from 0° to 100° by the equivalent volume at 0°. The linear expansion of
platinum from 0° to 100° = 0*000984, three times this number, the cubic
expansion, = 0*002952 ; this multiplied by 57 gives the equivalent volume
of platinum.

AND THEIR ATOMIC WEIGHTS. 5J

gas of a body thus obtained is to that of steam, as the specific
gravity of that body is to that of water. But Gmelin, (in
his Handbuch, 4te Aufl., s. 58 u. 74,) has shown that Persoz
not only starts with purely arbitrary views, but even with
such as might be proved to be false, and that he confirms his
calculations in part with observations altogether incorrect.
1 vol. of solid iodine gives 437 vols, of vapour; 1 vol. of
phosphorus 310 vols.; 1 vol. of arsenic 440 vols.; 1 vol. of
sulphur 230 vols.: 1 vol. of mercury 1500 vols, of vapour,
whence the untenability of Persoz's view is made apparent.
At the time Schroder traced the condensation induced by
the union of solid bodies to their constituents, and not, as
he now does, to the aggregate, and when he assumed the
equivalent volume of the oxygen to be 32 (see page 13,) and
not 64 as he now does, he endeavoured likewise to explain the
different specific heat with which oxygen and the elements
generally enter into combination as occasioned by the different
condensation of the elements in these compounds. {Pogg. Annal,
Bd. lii., s. 269, &c.) He declared, by way of further confirm-
ing this proposition, that, if the atom-volume of a body suffer
condensation according to simple relations, the specific heat of
this atom will likewise change according to simple fixed rela-
tions.

It is well known that oxygen is contained with the specific
heat of 30 in the oxides R O, and many oxides K2 03, and R 03;
in all these oxides the equivalent volumes of the oxygen
amounts to 32, as shown by the first table, at page 14. The
specific heat of the oxide of tin, Sn 02 is 87 ; if we deduct
the specific heat of Sn, that is, 40, there will remain 47 for
the specific heat of 02, consequently 23 for that of O, that is
nearly |^x30; oxygen is therefore contained in oxide of tin
with ^ of the specific heat, with which it is contained in the
first-named oxides. According to the second table at page 14,
oxygen occurs in the oxide of tin with the equivalent volume
16, that is, ix32; consequently with half as large an equi-
valent volume as the one, with which it appeared in the first-
named oxides. According to Schroder, therefore, this conden-
sation of volume to -^ corresponds with the diminution of spe-

THE VOLUMES OF BODIES

cific heat to j. The same holds good, according to his statement,
for titanic acid, in which he likewise assumed the equivalent
volume of the oxygen to be only 16, although it may safely be
taKen as o~.

The specific heat of iron pyrites : Fe S2 is 97. If from
this we deduct the specific heat of Fe, that is, 40, there will
remain a specific heat of 57 for S2, consequently the specific
heat of 28 for S, whilst the specific heat of S in a free con-
dition, is equal to that of iron, namely 40. The number 28
is again about £ x 40. The equivalent volume of iron pyrites
is 147, if from this we deduct the original equivalent volume
of the iron 45, there will remain for S2 the equivalent volume
102 and for S the equivalent volume 51, whilst in a free
condition, for S, it amounts to 100. Thus, by the union of
iron with sulphur to form iron pyrites, a condensation of ± the
equivalent volume of the sulphur has taken place, and "here
again consequently a diminution of the specific heat to £ cor-
responds to this condensation of A. *

As the equivalent volume of oxygen in Mo 03 and Wo O
amounts to 64, that is 2 x 32, (see page 14, second table) it
might be supposed that the specific heat of O in these com-
binations would be greater than 30; such, however, is not the
case, it amounting only to 26. The specific heat of Mo O
and Wo03 is for instance 118, and if we deduct from this
the specific heat of Mo and Wo, that is 40, there will remain
for 03 the specific heat 78, and consequently for O that of 26
It is also worthy of notice that oxygen is contained in oxide
oi iron with an equivalent volume of 32, and in oxide of
chromium with the equivalent volume of 16; notwithstanding
that the specific heat of both oxides is equally great.

The study of the equivalent volumes of bodies in a solid
and liquid condition deserves not less attention than the studv
of the equivalent volumes of bodies in a gaseous state, and,
indeed, the former possesses a peculiar interest from the relation
in which the equivalent volume in a solid condition stands to
isomorphism. It has occupied a large space in this essay from
the few definite points that could be advanced upon the subject

AND THEIR ATOMIC WEIGHTS. 59

and the many different views that required to be enumerated.
That the relation, in which the equivalent volume of a com-
bination stands to the equivalent volume of its constituents
should not be accurately known, cannot excite surprise, if
we bear in mind that the data on which the establishment
of these relations is based, are either wholly wanting, or are
deficient, and, without doubt, in some cases, erroneously con-
ceived. Thus exact details are wanting for most bodies with
reference to the enlargement of volume (dilatation) suffered
by the action of heat; the specific gravities are generally
known only with some approximation to the truth, and the
data obtained for the fusion and boiling points deserve, for
the most part, still less confidence than those concerning the
specific gravities. As it was impossible to find the law of
equivalent weights until the older and frequently most in-
correct numerical data of the per-centage composition of the
combinations had been replaced by others that were more
accurate, so it will likewise be impossible to find the law of
the equivalent volumes until correct data are obtained res-
pecting dilatation, the fusion and boiling point, and the specific
gravity of bodies.

This would seem to be the most fitting place to say a few
words upon Gmelin's atomic numbers. We have named equi-
valent volumes of bodies in a gaseous and liquid or solid condi-
tion, the relative volumes, which are filled by an equivalent of
the bodies, in a gaseous, or liquid, or solid condition. In the
comparison of the equivalent volumes in a gaseous condition
the equivalent volume of oxygen has been set down as =1;
the equivalent volume of hydrogen was therefore =2, that
of the vapour of alcohol =4, &c; in comparing the equivalent
volumes in a solid and liquid condition, the quotients obtained
by the division of the specific gravities by the equivalent
weights have been directly made use of, since no such simple
relation between the equivalent volume as observed in the
gaseous condition has as yet been established. If now, in
accordance with the atomic theory, we substitute the word
atom for that of equivalent, we obtain the word atom-volume
from the word equivalent volume, and if further we use the
words atoms and atom-volume in the definition given for the

60 THE VOLUMES OF BODIES

equivalent volumes of bodies, we shall have as follows: atom-
volumes of bodies in a gaseous, liquid, or solid condition are
the relative volumes occupied by an atom of the body in a
gaseous, liquid, or solid condition. According to this definition,
the volume of an atom of iodine in a state of vapour is 400
times greater than the volume of an atom of iodine in a solid
condition, for 1 vol. of solid iodine yields above 400 vols, of
vapour of iodine; 1 vol. of fluid mercury yields about 1500
vols, of vapour of mercury, the atom of mercury consequently
in a gaseous state fills a space 1500 times greater than that
which it occupies in a solid form.

The atoms of bodies are impenetrable, and not capable of
expansion ; they are not in immediate contact with each other,
do not fill space uniformly, but are kept at a certain distance
from each other, which changes with the density of the body. If
we suppose 1500 atoms to be contained in a certain volume of
fluid mercury, the distance of the atoms from one another will
increase so much on their transition from the fluid to the gaseous
form, that a volume of the vapour of mercury, equal to a
volume of fluid mercury, will only contain one atom of the
metal. We are very generally accustomed to say, that the atoms
are surrounded by spheres of heat, which enlarge by heating,
and on a transition from the solid to the gaseous condition, and
diminish by cooling, and on a transition from the gaseous to the
solid state*.

Gmelin considers it better, instead of giving the relative
volumes taken up by an equal number of the atoms of bodies,
conversely to give the number of the atoms contained in an
equal volume of the bodies, and thus speaks of atom-numbers

* Lowig says, in his Chemie der Organischen Verbindungen, 1844, Bd. i., s.
91, " According to the atomic view, the individual atoms, of which a connected
mass consists, cannot perfectly fill space ; it is to be assumed that each individual
atom consists of a solid part surrounded by a sphere of heat, standing in the
same proportional relation to the individual atoms, in which the atoms combine
together, so that the quantity of the casing of heat is equivalent to the quantity
of the atomic weight. If the materiality of heat be admitted, then the atoms
will combine with heat, according to definite proportions, in the same manner
as these atoms combine together." It is, however, clear that heat cannot be
anything material if we think of it as existing in the manner specified between
the atoms of bodies ; if it were anything material it would likewise consist of
atoms, and what would then intervene between these atoms ?

AND THEIR ATOMIC WEIGHTS. 61

instead of atom-volumes. I do not consider that the least
advantage to science is gained by this change. It will be evi-
dent in what relation the atom-number stands to the atom-
volume, of which it forms, as one may say, the converse. The
atom-volume shows the relative volumes of the atoms of bodies;
if, therefore, we divide the number 1 by the atom-volumes, we
must obtain the atom-numbers. The atom-volume is the quo-
tient yielded by the division of the specific gravity by the
atomic weight, and the atom-number is the quotient obtained
by the division of the atomic weight by the specific gravity.
Thus the more atoms of a substance are contained in a definite
space, and the denser the atoms are, the greater must be the
specific gravity; the specific gravity of a substance is, there-
fore, the product of the atom-number and the atom-weight.

If we assume that 1 equiv. of hydrogen, nitrogen, chlorine, &c,
fill 1 vol., 1 equiv. of oxygen, phosphorus vapour, and vapour
of arsenic will fill % vol., and 1 equiv. of sulphur only -$ vol.
If, therefore, 1 vol. of hydrogen, &c, contain 1 at. of hydro-
gen, an equal volume of oxygen, that is, 1 vol. of oxygen
must contain 2 at. of oxygen, and 1 vol. of vapour of sulphur
contain 6 at. of sulphur. Thus Gmelin distinguishes between
1 atom-gases, (hydrogen, iodine, bromine, chlorine, nitrogen,
mercury,) 2 atom-gases, (as phosphorus, arsenic, oxygen,) 6
atom-gases, (as sulphur,) and he maintains that in elastic fluid
compounds there are besides -3, 2", -3 , and ± atom-gases *. On a
substance assuming a gaseous form, its atoms become sur-
rounded by spheres of heat, whose volumes are as 1 (in the
6 atom-gases,) as 3, (in the 2 atom-gases,) as 6, (in the 1 atom-
gases,) as 9, (in the |- atom-gases,) as 12 (in the -i- atom-
gases,) as 18, (in the 1 atom-gases,) as 24, (in the 1 atom-
gases.)

If we set down the atom-volume of oxygen as =1 instead
of that of hydrogen, we shall of course obtain other atom-
numbers. Oxygen, phosphorus, arsenic, &c, will then be 1

* For making use of the expressions \ atom-gas, \ atom-gas, &c., we must
ask the indulgence of the severe critic, as much as when we say that 1 atom
of mercury in a gaseous form fills a space 1500 times greater than it occupies
when in a liquid condition.

62 VOLUMES OF BODIES AND THEIR ATOMIC WEIGHTS.

atom-gases, hydrogen, nitrogen, chlorine, &c., 1 atom-gases,
and sulphur a 3 atom-gas.

The atom-numbers 1, 2, 6, &c, for hydrogen, oxygen, and
sulphur, are reduced numbers, expressive of the relation in
which the atom-numbers resulting from the division of the
specific gravity of the gas by the atom-number stand to each
other. Gmelin has also calculated such reduced atom-numbers
for the solid and liquid elements; and they may easily be
deduced from the Table at page 4, which gives the atom-
volumes of the elements. Potassium has the greatest equiva-
lent and atom-volume, it must, therefore, have the smallest
atom-number ; and it is evident, that the elements whose atom-
volume is equal, must also have equal atom-numbers.

II.

MEMOIR.

STUDIES ON THE CONNECTION EXISTING BETWEEN

THE ATOMIC WEIGHTS, CRYSTALLINE FORM,

AND DENSITY OF BODIES.

By M. Filhol.

Among the objects of research, that have most frequently
attracted the attention of physicists and chemists, we may
mention the study of the relation existing between the atomic
weight, the density and the crystalline form of different bodies,
whether simple or compound. The researches made in this
department of science by Dumas, Boullay, Kupfer, Persoz,
Schroder, Kopp, and Ammermuller, have made us acquainted
with a series of results as remarkable as they were un-
expected.

As the atomic weights, which have served to establish the
remarkable approximations referred to in the memoirs of the
authors alluded to, have, for the most part, been very consider-
ably modified during the last few years, I have thought that it
would not be wholly devoid of interest, to examine whether the
laws established by these chemists were still valid, notwith-
standing the changes that had been introduced ; it has likewise
appeared to me, that a critical resume of all that has been
published up to the present day, and a consideration of the
new views developed in these works, might not be deemed
wholly unimportant.

THE VOLUMES OF BODIES

With these views, I purpose, in the present memoir, —

1st. To consider the researches of the different authors
whom I have already cited, and to examine into the relations
they have established between the atomic weight, the specific
weight, and the crystalline form of bodies, in order to see
whether they can be retained without alteration.

2nd. To discuss the value of the views developed by M.
Kopp, relative to the manner in which we ought to consider
the chemical constitution of salts.

3rd. And in conclusion, to unfold several of my own views
on this subject.

I purpose devoting a separate and special memoir to the
consideration of the relation existing between the specific
volume of isomorphous, or homeomorphous bodies and their
crystalline form. I will, however, before proceeding further,
give the atomic weights and the densities on which my calcula-
tions have been based.

I have made choice amongst the different determinations of
atomic weights of those which appeared to me to merit most con-
fidence, without in any way being biassed by the greater or less
support which any of the numbers might lend to the views I
was desirous of developing. The numbers which I selected are
as follows: —

ATOMIC WEIGHTS.

Barium

... 854*85

Magnesium

... 157-75

Cadmium

... 696-17

Manganese

... 344-44

Calcium

... 250-00

Mercury

... 1250-90

Carbon....

... 75-00

Potassium

... 488-84

Chlorine

... 44302

Silver ....

... 134901

Chromium

... 249*83

Sodium

... 287*17

Copper

... 396-00

Sulphur

... 20000

Iodine ....

... 1585-57

Strontium

... 545-60

Iron

... 35000

Tin

... 735-03

Lead ....

... 1294-50

Zinc ....

... 406-59

I have myself determined a great number of densities, of
which I will give a table. My calculations have been made
with the most careful attention.

AND THEIR ATOMIC WEIGHTS.

Specific Gravities. By M. Filhol.

Iodic Acid

4 250

Anhydrous

Sul-

Arsenious „ ....

. 3'884

phate

of Silver

5-410

Arsenic „ ....

.. 4-250

Lead

6-300

Nitrate of

Soda....

2-260

Protoxide of Barium

. 5456

Strontium ..

.. 4-611

Baryta
Strontia

3-200
2 857

Calcium

.. 3' 180

Lime

2-240

Zinc

.. 5612

Lead....

4581

Lead

.. 9-361

Hydrate of Potash

2 044

Binoxide of Copper

.. 6-322

Soda
Baryta

2-130

Alumina

.. 4-154

(BaO, HO)
Strontia

4495

Anhydrous Chlo-

(SrO, HO)

3-G25

ride of Potassium ..

.. 1-994

Baryta

(Ba O, 9H O)

Sodium

.. 2-240

1 -656

Barium

.. 3-750

Strontia

Strontium .

.. 2960

(SrO, 9HO)

1-396

Calcium

.. 2-240

Zinc ...

3-053

Iron

.. 2-528

Lime

2 078

Iodide of Potassium .

. 3056

Crystallized Chlo-

Sodium

. 3-450

ride

of Barium

Barium

... 4-917

(BaCl + 2HO)

2-664

Lead

.. 6-384

Strontium ....

1 603

Mercury (bin) 6*250

Calcium

1-635

Silver

.. 5-500

Magnesium ....
Iron....

1-558
1-926

Monosulphide of Potassium .

. 2130

Sodium

... 2-471

Crystallized

Sul-

phate

of Soda ...

1-520

Carbonate of Potash

... 2-267

Lime

2331

Soda

... 2-509

Magnesia

1 751

Baryta

... 4565

Alumina
Iron....

1-569
1-904

Anhydrous Sul-

Zinc...

2036

phate of Potash

2-625

Copper

2286

Soda

... 2-629

Strontia

... 3-770

Crystallized

Ni-

Lime

3102

trate

of Strontia

2113

Magnesia

... 2628

Lime

1-780

Zinc

... 3-400

Anhydrous

Borax

2-367

Copper

... 3530

Iron

... 2-841

Crystallized

Borax

1 -692

66 THE VOLUMES OP BODIES

Amongst these densities there are a great number which
agree with those determined by Boullay, Mohs, Karsten, &c.,
others, on the contrary, differ very widely from them; as,
for instance, those of baryta and strontia, but I have so
frequently, and always with the greatest care, repeated the
process of weighing these two oxides, that I believe I may
affirm that the error is not on my side.

I will now proceed to examine the works of each of the
authors to whom I referred at the beginning of this Memoir,
taking them in the order in which they were published.

Kupfer published in 1824 {Annates de Chimie et de Physique,
2 me Serie, tome xxv., page 337,) a Memoir in which he gave
the exposition of a method, by which, according to his views,
the density of a large number of bodies may be determined.

Suppose p and p' to be the atomic weights of two different
substances, s and s' their densities, and y and tf the volumes
of their primitive form, the semi-axis being assumed to be

TO S 7) S

equal to unity; then we have — = — — .

In order to obtain exact results by this formula, it is neces-
sary that the two substances to be examined should belong to
the same system of crystallization.

Kupfer admits only four primitive forms, viz. ; the regular
octohedron, the octohedron with a square base, the octohedron
with a rhombic base, and the rhombohedron.

Kupfer's memoir contains the calculations of a very large
number of densities, all of which agree in a most remarkable
manner with the results yielded by experiment; and this agree-
ment is much the more extraordinary as there are serious
grounds, amounting almost to positive evidence, for believing
that the relation indicated by Kupfer cannot be true.

Suppose, for instance, that in the equation — — ^-4-,

y y

y becomes equal to y'*, we shall have p s = p /, that is to
say, the densities will be inversely as the atomic weights:
which is by no means in conformity with experience. On
examining with particular attention the memoir of which I

* Many examples of this may be found in Kupfer's Memoir.

AND THEIR ATOMIC WEIGHTS. 67

speak, we shall soon detect the cause of this surprising agree-
ment, which results from Kupfer multiplying or dividing the
atomic weights of either of the two bodies under examination
sometimes by whole numbers, and sometimes by fractions,
and by this means always concludes in satisfying the conditions
of the problem. Thus the atomic weight of phosphate of
lime is multiplied by -|, that of the emerald by ^, that of

amphigene by i, and that of Siberian idocrase by -|, &c. The
necessity of thus having recourse, in a great number of cases,
to the multiplication of atomic weights by numbers chosen in
so arbitrary a manner, affords reason to doubt the exactness
of the preceding formula; and I am about to show that other
considerations lead to the proof that it is not accurate.

Dumas discovered long since a simple relation existing
between the atomic and the specific weights of a large number
of bodies; and this relation is moreover connected with the
crystalline form, as it only exists in isomorphous bodies. If
p be the atomic weight of a body, d its density, v its atomic

volume, we shall have v = ^.

Dumas was the first to demonstrate that the value of v was

sensibly the same for all bodies, whose isomorphism had been

established by Mitscherlich; and far from shaking the validity

of this law, the changes that have subsequently been made

in several atomic weights have contributed to prove its

accuracy. We shall shortly see that Kopp has proved its

applicability to compound, no less than to simple bodies.

Thus, then, for perfectly isomorphous bodies ^ = ^, or

what is the same, pd'=p' d. Consequently Kupfer's formula,
which supposes pd ss p' d'9 is necessarily false.

The study of the relations existing between the atomic
weight, the density, and the crystalline form of compound
bodies, is rendered very difficult, owing to the combination
of bodies being almost always accompanied by a considerable
change of volume, and that it is very rare for the volume of a
compound to be represented by the sum of the volumes of
its components. We may, therefore, easily conceive that the

F 2

68 THE VOLUMES OF BODIES

density of a compound cannot be determined by calculation
until the law regulating the change of volume experienced
by the body in the act of combination shall be accurately
known. Several chemists have endeavoured to discover it,
and I will give a rapid analysis of their labours.

The first memoir written on this subject is by Boullay.
After having arranged in one table the actual density of
a large number of compound bodies, this chemist compares
their density to that which would be possessed by these
bodies, if their elements had combined without experiencing
any change of volume; he then proceeds to demonstrate that
several cases may present themselves whilst the bodies are com-
bining.

1. The density of the compound may be the same as that
of the components.

2. It may be different.

On comparing the actual density of several sulphides with
their calculated density, he then finds that in this class of
compounds, the combination is always effected with conden-
sation; the inverse holds good for iodides. Here the volume
of the combination is greater than that of the components.
We must, however, according to Boullay, except iodide of
potassium, (and I would also add to this the iodides of sodium
and barium).

M. Boullay finally examines if the condensation is the same
for those bodies which yield the same formula of atomic com-
position, and here he can discover no analogy.

The memoir that I have been considering is essentially
characterised by the absence of all preconceived views, the ex-
periments having invariably been made upon compounds whose
elements were known in their solid state. M. Boullay appears
to have adopted the best means of discovering the laws of
condensation or of dilatation, but I will subsequently show,
that in making choice of other compounds, and proceeding in
other respects in the same manner as M. Boullay, we arrive
at curious approximations, that he had not observed. It must
not be forgotten that sulphur, which presents so many ano-
malies in its physical properties, may occur in different states in
certain sulphurets, and this fact probably explains the reason

AND THEIK ATOMIC WEIGHTS. 69

why M. Boullay did not observe any relation between their
co-efficients of condensation.

Next in point of date to the memoir I have been consi-
dering, succeed the researches of M. Persoz.

Without entering into any unnecessary detail, I would
call attention to the fact that M. Persoz calculates a priori the
density of bodies in the following manner: he supposes that
1 equiv. of oxygen weighing 100 grammes, and occupying a
volume of 70 liters, 1 equiv. of any other body will, if reduced
to vapour occupy a space either of 70 liters, or 70 multiplied
by 2, 4, 8, 16, 32 — or by 3, 6, 12, 24; this being assumed, he
seeks the hypothetical weight of 1 liter of the vapour of the
body whose density he wishes to determine, and then establishes
the following relation:

0-8003 : P : : 1 : x

Weight of Weight of Density Density

1 liter of aqueous 1 liter of the of soiurht

vapour. vapour of the body. the water.

M. Persoz presents in support of his views, a table in which
the density of several bodies is expressed with a surprisingly
close approximation to the results yielded by experiment. But
as soon as we attempt to apply his method in a general manner,
we meet with so many exceptions as to render it impossible
to admit the validity of the proposed law : I will only instance
a very small number of cases; many similar ones might, how-
ever, be found.

Calculated Density. Observed Density.

Iron C-24 7*73

Tin 6'55 7*28

Copper 7*16 8-90

Silver 12*03 10'47

Manganese 6*00 8*00

Arsenic 8*35 596

Sulphur 1-78 1-99

M. Baudrimont states the relation indicated by M. Persoz
differently: according to him the molecular and specific weights
of bodies of a determined nature are certain multiples or
submultiples of one another, and these relations are the same
for isodynamic bodies. We may, therefore, conclude that
isodynamic bodies sensibly occupy the same volume with

THE VOLUMES OF BODIES

weights proportionate to their molecules. To demonstrate
this law, M. Baudrimont compares the equivalents of bodies
and their density to the same unit. The unit he takes is
56*24, the half of the number representing the equivalent
of water; by dividing the chemical equivalents of bodies by
the semi-equivalent of water, he obtains numbers that bear a
simple relation to the densities. The following table will
give some idea of this.

Bodies.

Atomic
Weight.

HO
— =1-

Density
observed.

Relation.

Diamond ....

1504*5

2-66

3-55

2 : 3

Chrome

349 83

621

5-90

Manganese

344-14

612

8 02

Iron

350-00

622

7778

Zinc

406*59

7'22

7-21

Nickel ....

36914

6-38

841

Cobalt

36899

665

8 53

>1 : 1

Copper

396-00

7-03

8-89

Silver

674-5

11-99

10-47

Palladium ....

665-9

11-84

11-80

Platinum ....

1333-5

21-93

21-60

Gold

1228-31

2183

1925

Tungsten ....

1183-00

21-03

17 60

)

Cadmium ....

69677

12-38

860

Arsenic

468-24

8-32

596

( 3 : 2

Molybdenum

598-52

10-62

7*40

Lead

1294-5

2301

11-35

]

Antimony

806-45

14-34

6-70

Tin

735-30

1307

729

>2 : 1

Selenium ....

494-58

8-79

4-30

Phosphorus

196*4

3-49

1-77

Iodine

792*5

14-08

4-94

3:1

Bromine ....

499-8

8-88

3187

Sodium

287-17

5-10

0-97

5 : 1

Potassium ....

48994

871

0-86

10 : 1

The relations indicated in this table are by no means
rigorous, and very considerable errors must have been made
in the determination both of the atomic weights and the den-
sities before such results could have been obtained. I have
endeavoured to discover the changes that must necessarily be
made in the densities in order to arrive with exactitude at
these relations, and the following table gives the result of my
calculations.

AND THEIR ATOMIC WEIGHTS.

Tin*lip<*

Density

T^nHip^

Density

UUUItji

obtained.

calculated.

JJUUlL.^i

obtained.

calculated.

Diamond

355

3-99

Cadmium

8-60

8-25

Chrome

5 90

621

Arsenic

5-96

554

Manganese ....

8-02

6-12

Molybdenum....

7*40

6'74

Iron

7-778

622

Lead

11-35

11-50

Zinc

7-21

722

Antimony

6*70

7'17

Nickel

8-41

638

Tin

7'29

6-53

Cobalt

8-53

6 55

Selenium

4-30

4-39

Copper

889

7*03

Phosphorus ....

1-77

1-74

Silver

10-47

11-99

Iodine

4-94

4-69

Palladium

11-80

11-84

Bromine

3*187

2-96

Platinum

21-60

21-93

Sodium

0-97

1-02

Gold

19-25

21-83

Potassium

0-86

0 871

Tungsten

1760

21-03

The difference existing between the density calculated and
the density observed for carbon, manganese, iron, nickel,
cobalt, copper, silver, gold, and tin, is too great to be attributed
to errors of observation. M. Baudrimont, moreover, states
that this simple relation is not observed for sulphur, mercury,
and bismuth, consequently out of 28 bodies we find 12 which
yield numbers widely remote from the relation indicated in
the preceding table.

A law which admits of such numerous exceptions, and
supposes so many errors in its general results, cannot, in my
opinion, be received. Moreover, M. Baudrimont himself
admits that these numbers are merely approximative, and
that in wishing to submit them to a general law we may
deviate from the truth by the corrections we attempt to make.
Here it must be observed, and I think not unjustly, that in
order to obtain strictly accurate results, we ought to be able
to take some account of the comparative specific heat in dif-
ferent bodies, of their molecular condition, and, above all, of
their conditions of polymorphism.

M. Kopp has published a series of very important memoirs
on specific volumes; he first attempted to determine by cal-
culation the density of several bodies whose determination had
not been made by experiment, and presented great difficulties;
as barium, strontium, calcium, and magnesium. It is a very
singular fact, that he arrived at different numbers for the

72 THE VOLUMES OF BODIES

density of these metals, according as to whether he calculated
according to that of the protoxide or that of the sulphate.
Thus—

The density of Barium deduced from Ba O = 3*60

„ „ .... SO3BaO = 5'50

M. Kopp takes the number 4*55, the mean of the two
preceding numbers. He likewise finds —

The density of Strontium deduced from Sr O = 2 -37

„ „ .... S03Sr 0 = 335

and he adopts the mean number 2*86.

In introducing into the calculations established in his
memoir the specific volumes of barium and strontium, deduced
from the preceding densities, M. Kopp arrives at results,
which do not accord with those yielded by experiment, and
he thence concludes that barium and strontium do not combine
with their primitive specific volume, but with the following
volume.

Barium v = 143

Strontium v — 108

I have already remarked that the numbers given by Karsten
as representing the density of baryta and strontia were in-
correct; consequently the results obtained by Kopp in calcu-
lating from them the preceding densities must likewise be
false.

M. Kopp does not explain in his memoir how he was led
to represent the specific volumes of barium and strontium by
the numbers 143 and 108, but they agree so well with the
results yielded by experiment in all the tables he has calcu-
lated, that we may reasonably consider them to be exact; and
I will proceed to demonstrate that they are so.

In accordance with M. Kopp's views, we shall find all
the facts established, if we admit that all metals excepting
potassium, sodium, barium, strontium, magnesium, calcium,
and aluminum enter into combination with a change of
volume. Those that I am about to cite combine with the
volumes specified, which differ considerably from their actual
volume.

AND THEIR ATOMIC WEIGHTS. 73

. , .v , _ Volume in the

Actual V ohime. ^ ,.

Combination.

Potassium 583 234

Sodium 291 130

Barium 188 143

Strontium 191 108

Calcium 156 60

Magnesium .... .... 156 40

I have found, in opposition to the views entertained by
M. Kopp, that barium and strontium are comprehended in
the class of metals which may be considered to retain their
primitive volume in combinations ; it must be understood that
in making these calculations I have taken the data furnished,
as I suppose correctly, in M. Kopp^s memoir. On referring to
the tables in his memoir (Annales de Chimie et de Physique, 3me
serie, tome iv., p. 462), it will be seen that the facts can be
perfectly well received, on the supposition that the volume
of oxygen contained in the oxides is equal to 16, 32, or 64.

If we assume this law to be correct, and seek the density
of barium and strontium according to the densities that I have
already given for baryta and strontia, we shall have: —

Actual Specific Volume. Volume of the Metal.

BaO .... 175 175-32=143

SrO .... 140 140-32=108

Thus, the numbers 143 and 108, which M. Kopp derived from
experiment, are deduced by calculation from the densities of
baryta and strontian, as given above.

It must necessarily be admitted, that the other metals, as
potassium, &c, actually enter into combination with a volume
different from their actual volume.

I will not at present pause to consider the views developed
by M. Kopp with regard to the relations existing between the
crystalline form and the specific volume, as my object is to
enter fully into the subject in a special memoir, which I
purpose at a subsequent period to lay before the Academy.
I will now only direct my attention to an enquiry into the
correctness of M. Kopp's opinions relatively to the manner in
which we ought to consider the constitution of salts, and
examine how far his views on the subject are supported by the
study of specific volumes.

74 THE VOLUMES OF BODIES

All the arguments advanced by M. Kopp are based upon a
very curious observation made by M. Schroder, which I must
explain before I proceed further.

M. Schroder observed that when, after having calculated
the specific volume of a series of analogous combinations
(oxides, chlorides, sulphates), he deduced from the volume of
each of the combinations the volume of the corresponding
constituents, the same number was often obtained to express
the volume of the common constituents. (See Annates de Chimie
et de Physique, 3me serie, tome iv., p. 474.)

Thus, on subtracting from the specific volume of nitrates
the volume of the metals entering into their composition, we
always obtain as the remainder the number 358, which ex-
presses the specific volume of N 06, in the same manner as
for carbonates, the volume of C 03— 151, &c.

We will now examine the consequences deduced by M.
Kopp from his calculations. Two theories present themselves
in explanation of the chemical constitution of salts. According
to the first and more ancient one, a salt produced by the
combination of an oxacid with a metallic oxide must be con-
sidered to include each of the two elements that have entered
into its formation : thus, sulphate of potash equals S 03 K O.
According to the second theory, the oxygen of the oxide forms
with the acid a new compound, which will combine with the
metal : thus, sulphate of potash would be S 04 K.

Many arguments may be advanced for and against each of
these theories, I will, however, limit my remarks to the dis-
cussion of the support which, according to M. Kopp, is derived
by the second of these theories from a study of specific
volumes.

We will then consider, as this chemist proposes, both the
specific volumes of the salts, formed by an oxygen acid, and
those of the best known metals, as for instance the nitrates of
lead and silver.

We obtain by experiment for the specific volume of nitrate
of silver, v=:488; for that of nitrate of lead, v=472; for the
specific volume of silver, v=:130; for that of lead, v=114; for
the specific volume of oxide of silver, v = 1 94 ; for that of oxide
of lead, i?r=146.

AND THEIR ATOMIC WEIGHTS.

What theory can need more hypothesis to express the
volume of the salts in question? The theory of hydracids, sup-
posing the metal to be contained as such in these salts, deter-
mines the volume of N (X

AgN06
Ag

r6*

488
130

NO,

= 358

Pb N 06 = 472

Pb =114

NO,

= 358

The specific volume of N 06 is thus of the same value in
both cases; here only one supposition is required to express
the density of these salts, and the specific volume of N 06 is
established in both cases as being equal to 358.

The theory of oxygen acids does not admit of similar sim-
plicity; it supposes the metal contained in the salts to be
in the condition of an oxide, and in order by this means to
determine the specific volume of S 03, we have —

AgON05
AgO

488
194

PbONO.
PbO

472
146

NO,

= 294

NO:

= 326

The theory of oxygen acids thus yields a different remainder
for the volume of N" 05 in the different salts; consequently
the theory of hydracids is necessarily simpler than the former
with regard to the explanation of density, and considered
from this point of view has the greater appearance of pro-
bability.

These arguments are plausible, but a very little consi-
deration will be sufficient to show that they are much less
valuable than we might at first sight be inclined to believe.
And all the preceding reasons are actually based upon an
incorrect density (that of the nitrate or the oxide of silver),
or upon an exception, as I am about to show. If we suppose
for instance, that the theory of oxygenous acids be correct,
and proceed with regard to other salts as we have done with
the nitrates of silver and of lead, we shall have :

N 05 Ba O = 501
BaO = 175

NO,

= 326

N 06 Sr O = 466

140

SrO =

NO, =

326

THE VOLUMES OF BODIES

On examining other classes of salts ;

Vol.

Sulphate of Zinc .... S 03 Zn O = 294

ZnO

SOa

204

Vol.
Sulphate of Copper S 03 Cu O = 280

Cu O = 76

S 03 = 204

Carbonate of Baryta C 02 Ba O = 294

Ba 0=175

C02 = 119

Carbonate of Zinc C 09 Zn O = 209

ZnO

CO, = 119

The following tables will serve to prove that the examples
I have cited have not been arbitrarily chosen, but that all the
salts comprised in M. Kopp's tables, (excepting the salts of
silver,) lead to the same result. I believe I ought to remark,
that if we suppose an error to have been made in the deter-
mination of the density of oxide of silver, the salts of this
metal would be comprehended under the general law. An
error of this nature seems not improbable, for the numbers
given by different authors for the density of oxide of silver
vary very considerably from one another, as we may see by
the following statements: —

Density of oxide of silver ..

7*14 according to Herapath.
725 „ Boullay.

8 26 „ Karsten.

However this may be, I am about to show that the theory of
oxygen acids leads to results which are quite as satisfactory
as those obtained by the theory of hydracids. (I would remark
here that the following calculations are all based upon the
densities and chemical equivalents adopted by M. Kopp.)

Sulphates. The theory of hydracids furnishes a good
explanation of facts, if we suppose that the compound S 04
experiences two modes of condensation in its combinations.
Thus, in certain sulphates, its volume equals 186, whilst in
others it is equal to 236.

The theory of oxygen acids likewise accords well with the
facts observed, if we admit that the compound S 03 experiences
two modes of condensation, and that its specific volume equals
154 in some sulphates, and 204 in others.

AND THEIR ATOMIC WEIGHTS. 77

Sulphates. S03 = 154; O = 32. (First Series.)

Chemi-

Density

of the

Salt.

Specific

Spec.Vol.

Calculated

Sulphates

cal

volume

Volume

of the

of Sul-

Volume

Equiva-

of the

combined

phuric

of the

lent.

Salt.

Metal.

Oxygen.

Oxide.

Acid.

Salt.

1895

632

300

114

146

154

300

Baryta ....

1458

443

329

143

175

154

329

Potash ...

1091

2-60

420

234

266

154

420

Strontia

1148

390

294

108

140

154

294

S 03 = 204; O = 32. (Second Series.)

Copper .

Zinc

Lime

Magnesia
Soda

997

534

280

204

1004

342

294

204

857

2-90

296

204

759

275

276

204

892

244

366

130

162

204

280
294
296
276
366

Chromates. The density of Chromates may be well ex-
pressed by assuming the number 196 as the specific volume of
Chromic Acid.

Chromates
of

Chemi-
cal
Equiva-
lent.

Density
of the

Salt.

Specific

Volume

of the

Salt.

Specific
Volume
of the
Metal.

Specific
Volume

of
Oxygen.

Spec.Vol.

of the
combined

Oxide.

Spec.Vol.

Chromic

Acid.

Calculated

Volume

of the

Salt.

Lead

Pota3h ....

2046
1241

5*98
2-69

342

462

114
234

32
32

146
266

196
196

342

462

TlJNGSTATES.

oxygen = 32.

Volume of the acid = 212; volume of

Tungstates
of

Chemi-
cal
Equiva-
lent.

Density

of the

Salt.

Specific

Volume

of the

Salt.

Specific
Volume
of the
Metal.

Specific
Volume

of
Oxygen.

Spec.Vol.

of the
combined

Oxide.

Spec.Vol.

Tungstic

Acid.

Calculated

Volume

of the

Salt.

Lead

Lime

2877
1839

8-04
605

358
304

114

32
32

146
92

212
212

358
304

THE VOLUMES OE BODIES

Nitrates. O e= 32; NO. = 326.

Nitrates
of

Chemi-
cal
Equiva-
lent.

Density

of the

Salt.

Specific

Volume

of the

Salt.

Specific
Volume
of the
Metal.

Specific
Volume

of
Oxygen.

Spec.Vol.

of the
combined

Oxide.

Spec.Vol.

Nitric

Acid.

Calculated

Volume

of the

Salt.

Lead

Ammonia
Baryta ....
Potash ....

Soda

Strontia

2071
1004
1634
1267
1068
1324

4-40
1-74
3-20
2-14
219
2-84

472
576
501
592
488
466

114
218
143
234
130
108

32
32
32
32
32
32

146
250
175
266
162
140

326
326
326
326
326
326

472

576
501
592
488
4C6

Carbonates. O = 32; CO

2 e 119.

Chemi-

Density

Specific

Spec.Vol.

Calculated

Carbonates

cal

Volume

of the

Volume

Equiva-

of the
Salt.

of the

combined

Carbonic

of the

lent.

Salt.

Metal.

Oxygen.

Oxide.

Acid.

Salt.

Cadmium

1073

4-63

232

113

119

232

Iron

715

367

195

119

195

Manganese

722

3-70

195

119

195

Lead

1670

6-30

265

144

146

119

265

Zinc

779

3*73

209

119

209

Baryta ....

1233

419

294

143

175

119

294

Lime

632

3-00

201

119

201

Magnesia

534

2-80

191

119

191

Potash ....

866

225

385

234

250

119

385

667

237

281

130

162

119

281

Strontia

923

3-56

259

-108

140

119

259

Lime ....1
Magnesia J

1166

2-90

402

J 60
\ 40

32
32

119 )
119 J

402

The preceding calculations will, if I am not deceived, show
that no arguments can be drawn from the study of specific
volumes either for or against the theories in question, and that
the latter remain precisely the same as they were before these
observations were made.

To the labours of M. Kopp succeed those of Animermuller,
who has established a law whose expression, modified by
Poggendorf, is as follows:

The specific weights of the combinations of a radical with
an electro -negative body are to one another as the corresponding
atomic weights, either directly, or after multiplying them by
certain simple fractions.

AND THEIR ATOMIC WEIGHTS. 79

This law is verified in a great many cases, but an objection
may be advanced against it which I think of the greatest im-
portance. Thus, if we examine the tables drawn up by M.
Ammermuller, we shall not fail to perceive that the fractions
by which he multiplies the atomic weights are very far from
being simple fractions, and this is so fully the case that by
pursuing his method of working out the calculation we are
always certain of arriving at a result in conformity with the
one required by the experiments, and that even when we have
started from results, which are evidently false. This may be
proved by M. Ammermiiller's own calculations. He calculates
the density of binoxide of lead from that of the protoxide,
supposed to be 8*01, and he thus arrives by calculation at a
fairly correct determination of this density. But the real density
of protoxide of lead is, at the least, 9*3; M. Ammermuller has,
therefore, arrived at accurate results, although the data on
which his calculations were based are evidently incorrect. This
instance will, I think, be sufficient to show that this law cannot
be regarded as valid.

The last author whose work I would consider is M.
Schroder, and the conclusions he arrives at are to this effect: —

1. That the specific volume of a body may be equal to the
sum of the volumes of its elements. (This proposition is based
upon the observations made by M. Boullay.)

2. That any element may occur in different combinations
under different conditions, in such a manner that its specific
volume varies in the same relation as the numbers 1, 2, 3, 4, 5, 6,
&c. This law comprehends that of condensation.

3. That the condensation of any one of the elements of a
compound body may vary according to the same relation of
the combination. (This fact forms the basis of isomerism.)

The fourth law relating to isomorphism agrees with that
of M. Dumas, of which I have already spoken.

The truth of the first law is evident, and cannot be contested.

The second law does not appear to be sufficiently demon-
strated: it may, however, be supported by numerous examples.
Thus, we have already seen that the condensation of oxygen
in the oxides or in the oxysalts is such that all the facts may
be explained if we assume the volume of the oxygen to be

THE VOLUMES OF BODIES

equal to 8, or to a multiple of 8. M. Kopp has shown that
the condensation of chlorine in the chlorides is such, that the
vol. of the chlorine is 196 for some, and 245 for others. Now
these numbers, 196 and 245, are to each other as 4 to 5; but
even here the law is often at fault. Whether, therefore, in
the series of the sulphates, we adopt the views of Kopp, or
those which I have developed, the numbers that express the
atomic volume of S 0Q, or of S 0A, in the two cases of con-

's'

densation, are not exact multiples of one another.

I am now about to show, that if we study the changes of
volume which occur during combination, setting aside all
preconceived views, and without seeking to ascertain if the
dilatation or the condensation have effected either of the com-
ponents (which it seems to me to be very difficult to appre-
ciate), we arrive at results that are not wholly unimportant.

Let us take the series of the chlorides, and let —

P represent the atomic weight of the chlorine.

D „ its density.

P' „ the atomic weight of the metal.

D' „ the density.

A „ the density of the chloride.
If we suppose that the elements have combined without a
change of volume we shall evidently have :

P + P (dd')
~~ Pd' + P'd'
If D represent the density of the chloride, as yielded by expe-
riment, we shall always find D> A, and — ^ — will express

the co-efficient of condensation.

Let us take for the density of the chlorine the number 1 *35.
Faraday has found 1*33, Kopp admits 1*38.

Chlorides

Density

Density-

Co-efficient of

observed.

calculated.

Contraction.

Potassium

1-994

1-039

0-478

Sodium

2-240

1169

0-477

Barium

3*75

2825

0-265

Strontium

2-960

2*266

0-234

Calcium

2-240

1-702

0-240

Lead ....

5541

3*928

0-291

Silver....

5-530

3-920

0-291

AND THEIR ATOMIC WEIGHTS.

Let us similarly examine the sulphates.

Sulphates

Density

Co-efficient of

calculated.

observed.

Contraction.

Magnesia

2*266

2-628

0137

Zinc ....

2-920

3-400

0-140

Copper

2-991

3530

0152

Baryta

3393

4-440

0-236

Strontia

2-901

3-77

0-230

Lime ....

2-366

3102

0245

Lead ....

4-729

6*300

0-249

Supposing the density of C 02 to be 0*83, we obtain the
following results on applying this calculation to carbonates.

Carbonates

Density

Co-efficient of

calculated.

observed.

Contraction.

Baryta

2*566

4-565

0-437

Strontia

1-949

3-610

0-460

Lime ....

1-415

2-720

0-477

Lead ....

3-491

6470

0459

Cadmium

2-783

4-455

0 380

Magnesia

1-293

2-966

0563

Manganese ....

1-264

3-691

0657

Zinc ....

1-860

4-445

0581

Let us now proceed to consider from the same point of
view, the hydrated and crystallized salts.

Hydrates

Density

Co-efficient of

calculated.

observed.

Contraction.

Baryta (BaO+HO) ....

3718

4-495

0*172

Strontia (SrO + HO) ....

2-989

3-625

0-175

Lime(CaO + HO)

2292

2-078

0093

Magnesia (brucite)

2336

1-920

0178

Alumina (gypsite)

2400

1-989

0171

Alumina (AL03 + 8HO)

1-786

1*459

01 83

Zinc(ZnO + HO)

3434

3053

0110

Baryta (Ba 0 + 911 O) ....

2288

1-656

0-276

Strontia (Sr O + 911 O) ....

1-911

1-396

0-269

THE VOLUMES OF BODIES

Crystallized

Density

Co-efficient of

Sulphates of

observed.

calculated.

Contraction.

Soda ....

1-520

1-375

0*093

Magnesia

1-751

1-434

0-180

Lime ....

2-331

2-099

0-099

Zinc ....

2036

1-658

0-171

Iron ....

1-904

1-546

0188

Copper

2-286

1-829

0-199

Crystallized
Chlorides of

Density
observed

Density
calculated.

Co-efficient of
Contraction.

Strontium

Barium

1-920
3160

1-603
2-640

0165
0-166

These tables evidently show that in a very large number
of cases, setting aside all hypothesis, we find that the co-efficient
of the condensation of the components that are most analogous
to one another is the same in a chemical point of view.

These approximations further show that in a certain number
of cases, the relation existing between the densities calculated
for certain components of the same class is identical with that
which exists between the densities yielded by experiment; and
hence arises the possibility of calculating, a priori, the density
of certain salts from that of their elements and that of a salt
of the same species, having a great analogy with the one,
whose specific weight we wish to determine. In fact we have
A : h'::D:D'.

I will give a few examples of this fact: —

Chloride of sodium

„ potassium....

1-169 : 1*039

.... A = 1*169, D = 2-24
.... A' = 1*039,
2-24 : 1-990.

The density of chloride of potassium is given, by experi-
ment, in the following numbers, 1*94; according to Kopp 1*92;
to Karsten 1*994.

Sulphate of baryta ....
„ strontia ....

3-393 : 2 901

.. A = 3-393, D = 4-440
. A' = 2-901,
4-44 : 3 796.

AND THEIR ATOMIC WEIGHTS. 83

Experiment gives for sulphate of strontia 3*95; according
to Breithaupt 3*59; to Karsten 3*77.

It will easily be seen that these instances might be greatly
multiplied. Unfortunately, the relation which I have indicated
is not absolutely general, and cannot be considered to con-
stitute a law. It would have been easy to obtain identical
numbers for the expression of the co-efficient of condensation
of a large number of compounds by merely altering slight
quantities in the densities used in my calculations, but I have
left the numerous discrepancies that will be observed, in pre-
ference to altering any of the results yielded by experiment.

I would remark, in resuming the subject under discussion,
that the study of the relations which I have considered is far
from being so much advanced as may be supposed; it is easy,
by modifying a few of the numbers serving to establish the
calculations made on the subject, to obtain very promising
approximations, but their perfect agreement with the results
of experiments should excite some distrust, for it is extremely
difficult in cases where we are unable to take account of
numerous causes of error, to arrive at results which are as exact
as the greater number of those given in some of the works I
have analysed.

I am, however, far from wishing to depreciate the merit
of those memoirs, and I trust that the authors will be convinced
that nothing but a desire of throwing some degree of light
upon a difficult subject, in which almost everything remains
to be done, would have led me so unsparingly to expose the
weak side of their several theories. The best means of arriving
at truth is, in my opinion, to abstain as much as possible from
all hypothesis, to endeavour above all to determine with great
exactness the densities which form the basis of the calculations,
and to make a comparison, whenever it is possible, between
regularly crystallized bodies, since form has a great influence
on density. Setting aside the equality of volume in isomor-
phous bodies, the views I have discussed in this memoir, what-
ever be the ingenuity with which they may be advanced, all
admit of being contested.

I have endeavoured in the most impartial manner to show
whatever was defective or good in each one. There are some

84 VOLUMES OF BODIES AND THEIR ATOMIC WEIGHTS.

which certainly ought to be rejected, and others which must
be very considerably modified before they can be brought into
harmony with the results yielded by experiment, and others
again which, although correct in a very great number of cases,
are not so in all; these, it appears to me, ought not, however,
to be over-hastily rejected, since a more careful study may
show that they are capable of being made to accord in every
case with experiment.

III.

REPORT.

ON THE LAWS, ACCORDING TO WHICH THE MIXING
OF FLUIDS AND THEIR PENETRATION INTO PER-
MEABLE SUBSTANCES OCCURS, WITH SPECIAL
REFERENCE TO THE PROCESSES IN THE HUMAN
AND ANIMAL ORGANISM.

By Julius Vogel.

The phenomena, which* we are accustomed to embrace under
one general conception of animal life, are composed of an
infinite number of individual processes, which stand in the
most varied combinations and mutually reacting relations to
each other. To these belong those mixtures of different fluids
either in immediate and direct contact with each other, or
separated by organic partitions (animal membranes), through
which they pass and mix together. Many of these mixtures
are so simple, and are so entirely identical with those processes,
which every one in ordinary life has constant opportunities
of observing, that they must be considered as scarcely deserving
any further consideration; others, on the contrary, present
many peculiarities, appearing at first to be highly paradoxical,
and even in direct opposition to the laws of hydrostatics.
Among these are the processes termed Endosmosis and Exos-
mosis, in which two fluids, that mix by means of a partition,
so completely change their volumes, that the one fluid increases
while the other, in a corresponding manner, diminishes. Cases
of this nature have met with special and careful observation,
and attempts of different kinds have been made to explain

86 ENDOSMOSIS.

the phenomena occurring in this process. The deeper, how-
ever, that we penetrate into these apparently paradoxical
phenomena, the more does the mysterious obscurity that en-
shrouded them disappear; and the conviction forces itself upon
the observer, that they are connected in the most natural
manner with those simple cases, which give rise to no surprise,
from our being familiarised to them by daily experience. A
satisfactory explanation of the whole must, however, indis-
pensably be preceded by an endeavour, starting from the
simplest cases, to bring all phenomena occurring in the mixing of
fluids under common laws. But these, like all other natural
laws, cannot be established a priori, but must be obtained from
the experience gained by experiments, which again presuppose
observations. The conditions that may occur in the mixture of
fluids are very numerous, and consequently the field opened to
us is extensive, and the toil required most arduous. To this
must be added, that it is not sufficient here to ascertain the in-
fluence of the general conditions at work, but that it must be
considered rather the special task of science accurately to deter-
mine the amount of influence exercised in each individual case;
that is, to establish it numerically. Such an exact mathema-
tical mode of treating the subject is not only possible, but
appears to be necessary, and simply in accordance with the
requirements of science; it is, however, attended by difficulty,
and presupposes an extended series of careful investigations,
since each experiment of the kind, as the following consider-
ations will show, necessarily presents many sources of error;
it is only, therefore, such numerical results as have been
derived from a large number of observations that can be
received as approximating to a standard of correctness. The
problem is one which can only be satisfactorily solved by
degrees, and rather by the combined labours of many than by
the researches of any single enquirer.

These and similar considerations have induced me to lay the
following remarks before the public. Although I have deemed
it best for the present to withhold the special data derived
from a series of experiments (some of which were attended
with much labour,) as they are insufficient for the establish-
ment of a mathematical basis of this subject, and my object is

ENDOSMOSIS. 87

principally to draw attention to the leading phenomena which
occur in the mixing of fluids, and in their permeation through
organised substances, and, above all, to point out the mutual
connection existing between these phenomena, I shall rejoice
if these remarks should lead others to turn their attention
to an enquiry which is of so much importance to physiology,
and if the veil may thus be gradually removed, which now rests
upon this subject, as well as upon many of the most important
vital processes*.

If two fluids, whose constituents chemically attract each
other, come in contact, they will combine to form one mixed
fluid, of which each smallest particle will exhibit a like
property. If, for instance, we mix a fluid consisting of 20
parts of salt and 80 parts of water, with 100 parts of water,
a fluid will be formed from the mixture, the smallest distin-
guishable particle of which will contain 9 parts of water to
every 1 part of salt.

Precisely the same thing occurs on adding to the fluid some
solid body, that is soluble in it, as salt for instance. When
the solution is completed a fluid will be formed, the smallest
particle of which will exhibit the same character, consisting
of a definite quantity of the original fluid with a definite
quantity of the dissolved body.

* For the direction of such as are desirous of entering further into this
subject, I subjoin a list of the most important works regarding it:

Magnus, in Poggendorf 's Annalen. Bd. 10.

JV. W. Fischer do. „ 11.

Poisson do. „ 11.

Dutrochet, Me'moires pour 6ervir a l'Histoire Anat. et Physiolog. des Ve'ge'-
taux et des Animaux. Paris, 1837. T. i. p. 1 — 99. (Dutrochet himself regards
what he has written in this treatise upon Endosmosis as alone entitled to atten-
tion, considering all his former labours on the same subject, which are not
incorporated in this paper, as invalid. — Avaut-propos, p. xxxi.)

Kiirschner, Article, Aufsaugung, Absorption. Wagner's Handwbrterbuch
der Physiologic

E. Brucke, de Diffusione Uumorum per septa mortua et viva. Dissertat.
Berolini, 1842.

Poiseuille, Comptes Rendus, 1844. II., p. 994., &c.

C. Matteuci et A. Cima, Annales de Chimie et de Physique. Janvier,
1845.

Porrety Annals of Philosophy, 181C.

ENDOSMOSIS.

If we suppose the two commixing fluids to be two sepa-
rated masses, a, b, (fig. 1,) then the constituents of each will
attract those of the other. At the same time, however, the
constituents of a will attract those of b, and conversely, and
this latter species of attraction is stronger than that existing
between the constituents of each body among themselves.
Thus, atoms will pass from a to b, and again from b to a,
until both masses have become chemically equal; that is, until
they have assumed a like composition.

If, in this perfect admixture of the two fluids, as much
has passed from a to b as from b to a, both masses will retain
their original volume after their commixture; but if, on the
contrary, one mass takes up more than it gives off, the two
masses will change their original volumes in such a manner
that the one will have increased in proportion as the other has
diminished (fig. 2). In those cases, where the two commixing
fluids are contained in one common vessel, the difference of
size occurring in the two masses from their admixture is not
perceptible, for the mass a (fig. 3,) will, in accordance with
the laws of hydrostatics, assume the form «', when all want
of uniformity in the level of the fluid will disappear. There
are, however, cases in which the difference of volume in a
and b, arising from their commixture, may be perceived, and
where it admits of being measured; of these we will speak
presently.

Fig. 2.

Fig. 3.

Fig. 1.

Let us now consider the commixing fluids, not as two simple
masses, but as consisting each of a very large number of
particles, which are in each fluid perfectly similar to one
another. The manner in which these particles are arranged

ENDOSMOSIS.

exercises an essential influence upon the nature of the mix-
ture, that is, upon the time requisite for the perfect admixture
of the whole. Let a, in fig. 4, represent a particle of a
solution of salt, while the divisions from b to e are particles of
water. As each particle of water must take up a definite quan-
tity of salt in its perfect admixture, it is necessary that, in
the above advanced illustration, every saline particle coming
to e shall first pass through the particles from b to d: to
traverse this course, a certain time is, however, requisite, the
length of which will depend upon the length of the course
to be traversed.

In fig. 5, on the contrary, wThere the position of the par-
ticles is different, all the particles of water being in immediate
contact with the particles containing salt, the particle e will
take up salt much more rapidly than in fig. 4.

Fig. 4.

Fig. 5.

It happens, from this and similar relations, that equal
quantities of fluids of the same kind, which mix together,
require a very different period of time for their perfect ad-
mixture under different relations.

If, for instance, two fluids of different specific weights, which
will mix together, as water and a solution of salt,
spirits of wine and water, &c, be carefully brought
together in such a manner that the heavier fluid
shall be lowermost, the particles of both fluids being
as little changed from their position as possible, the
same case will occur as in fig. 4, and the perfect
admixture will be very slowly effected. We may
regard both masses as a column (fig. 6), the upper
half of which, as from a to c, being composed of
particles of spirits of wine, and the lower, from
d to /, of particles of water. As long as the

Fig. 6.

90 ENDOSMOSIS.

particles remain at rest in their position, each atom of spirits
of wine, passing from the particle c to f, must go through
d and e, and, in like manner, every atom of water, going from
d to a, must pass through c and b. The higher, therefore, the
columns are, the longer will be the period of time necessary
for the perfect admixture of the whole mass; but the case
is different when not only the atoms in the two fluids, but
whole portions of the mass, as from a to f, move; if, for
instance, we stir them. Then each particle of water will
come in immediate contact with a particle of spirits of wine,
as in fig. 5. The perfect admixture will follow much more
rapidly, and the height of the fluid column will not exercise
any appreciable influence upon the period of time required for
the perfect admixture.

The following are some of the principal causes that call
forth changes of place of whole portions of matter (and not
only of atoms), or, as we generally express it, currents, in the
commixing fluids.

Mechanical forces acting from without, as stirring, shaking,
&c.

The gravity, or unequal specific weight, of the commingling
fluids. If, for instance, we dissolve a piece of sugar in water,
taking special care to avoid every motion, it will make a great
difference in the period of time required for the complete
diffusion of the body in the fluid, whether we put it at the
bottom of the vessel, or hold it suspended on the surface of
the water. In the first case the particles of water saturated
with sugar are below, and remain there owing to their greater
specific weight; hence no currents will be formed, and the per-
fect admixture will only follow very slowly on the progression
of the atoms. If, on the contrary, the sugar is above, the parts
of the water saturated by it will have a tendency to sink, owing
to their greater specific weight; visible currents will then
arise in the fluid, and the mixture will be much more rapidly
completed.

Unequal heat, boiling, &c., act in like manner in calling
forth currents that accelerate the mixture.

As in every mixture of fluids, owing to one cause or
another, currents unavoidably arise, the amount of which can-

ENDOSMOSIS. 91

not be accurately determined, it is scarcely possible strictly to
define the difference which certain conditions occasion in the
perfect admixture of two commingling fluids. Every case
presented to us is a combined one, which at certain times
approximates more to the condition of absolute rest, and then
again at others more to that of absolute motion in the particles
of the fluid. It is, however, of importance to theory accurately
and strictly to distinguish these two kinds of motion, viz.,
the progression of invisible atoms, and that of the smallest visible
particles. If, as is generally done, we term the former currents,
we must not apply the same appellation to the latter.

In the cases we have been considering, the two fluids were
miscible, that is, they formed, in consequence of being brought
in contact with each other, one fluid, the smallest appreciable
particle of which had a perfectly homogeneous character after
the perfect admixture had taken place.

There are, however, cases in which two fluids withdraw
certain constituents from each other on being brought in
contact, without themselves mixing together.

Thus, for instance, oil (olein) and water do not mix; and
it is only after long-continued stirring that they form an
emulsion, that is, a fluid in which the particles of oil and water
are plainly to be distinguished from one another. If now
we rub oil and common salt together, and bring this mixture
in contact with water, the latter, whose particles have a greater
chemical attraction for the salt than have the particles of oil,
will withdraw the salt from the last-named particles, and after
the perfect admixture is completed we shall have, besides
the particles of oil, particles of water, each of which will hold
a certain quantity of salt in solution. On mixing equal
quantities of ether and water two fluid layers will be formed,
which will not mix further, and of which the upper and lighter
consists of ether and water in the proportion of 36 to 1, and
the lower and heavier one of water and ether in the proportion
of 9 to 1 . On separating the two fluids, and dissolving chloride
of mercury in the one, the other fluid will, on the two fluids
being again mixed, withdraw a certain portion of the salt from

92 ENDOSMOSIS.

the solution. (Briicke.) In both cases, therefore, the one fluid
abstracts from the other, without, as a necessary consequence,
yielding anything in return; the one increases, and the other
diminishes, in proportion to the quantity taken up.

If, in this process, we have regard to the quantity taken
up in a given time by a fluid under definite relations, and to
the quantity given off by the other fluid, we shall easily
perceive that the quantitative amount that has passed from
one to the other in a definite time is (in equal masses) greater
in proportion to the surface at which the two fluids are in
contact, and that it makes a considerable difference here
whether the change is effected during a state of rest of the
fluids, or while they are in motion (currents, § 3).

Relation of such Fluids as come in contact u'ith Permeable
Substances, or are separated from each other by Permeable
Partitions,

Hitherto we have considered both fluids as in immediate
contact: we will now investigate the cases in which a fluid
penetrates into a permeable substance, or where two fluids are
separated by a permeable partition. Every substance may
be regarded as a permeable partition, that admits of being
penetrated by one of two fluids, or merely by certain consti-
tuents of either.

Experience teaches us that many permeable substances
exercise upon the fluids penetrating into them a certain in-
fluence, which is different in different cases, and it requires
therefore a special experiment for each case, to determine its
nature and amount, although it admits of being brought under
certain general points of view. This influence is partly mecha-
nical and partly chemical.

6.
The mechanical influence exercised by a permeable sub-
stance upon the fluid penetrating into it consists in the fluid
being attracted by the substance, with a force that opposes a

EXDOSMOSIS. 93

certain resistance to other mechanical forces acting upon the
fluid, as, for instance, gravity, hydrostatic pressure, &c.

This force is generally designated by the term capillarity,
or capillary attraction. Of the numerous examples occurring
of this mechanical capillary force, I will here only adduce the
well-known phenomenon of water rising in a strip of blotting-
paper in opposition to the action of gravity. This mechanical
capillary force varies very considerably in different cases, de-
pending (1) upon the nature (of the material) of the permeable
substance and of the fluid; thus, for instance, water is attracted
by all substances which it is able to moisten, whilst between
glass and mercury no such attraction exists; and (2) upon the
arrangement of the permeable substance, its thickness, and the
size of its interstices (pores).

The many different kinds of filters furnish examples of the
various manners in which this mechanical capillary force acts
in individual cases. Thus, if a fluid be poured into an open
funnel, it will run through much more rapidly than when it
has first to pass through a filter. Its slower escape, in the
latter instance, does not only depend upon the size of the
openings being contracted by the solid substance intervening
between the capillary spaces of the filter, but likewise upon
the capillary attraction in the separate capillary spaces of the
vessel, which offers a certain resistance to the hydrostatic
pressure of the fluid, and thus in part removes it. In many
filters this resistance is very inconsiderable, and the fluid will
consequently pass rapidly through, while, in others, again, it
is much stronger, and the fluid then filtrates slowly. The force
of capillary attraction is, in many cases, so great that it is
able to equipoise a considerable opposing force. If, for in-
stance, a glass tube be so tightly closed by a bladder that no
fluid can penetrate between the glass and the animal membrane,
and if it be then filled with water, the capillary attraction of
the bladder will retain the water with so much tenacity, that
the pressure of a column of water several feet in height within
the tube will not be able, before a period of several hours have
elapsed, to force any appreciable quantity of water through
the bladder, and many days, or even weeks, will intervene
before a column of water, several inches in height, and whose

94 ENDOSMOSIS.

base covers the surface of the bladder, will be able to penetrate
through the membrane. But even in the case of thick animal
membranes, the resistance opposed by capillary attraction to
hydrostatic pressure and other moving forces, is not, generally
speaking, absolute, probably because almost every membrane
between its many small pores, in which capillary attraction is
very considerable, contains also several of larger size, through
which fluids may be forced, even under a very inconsiderable
amount of pressure.

The amount of this mechanical capillary force may be ascer-
tained for different permeable substances by means of experi-
ments, that at least approximate to a correct estimate; for this
purpose we use glass tubes of equal diameter, which are closed
with these different substances after being filled with a fluid;
by this means the quantities permeating in an equal period of
time by equal hydrostatic pressure may be accurately deter-
mined. Here there are, however, many relations to be re-
garded, which detract from the accuracy of the results obtained
in most experiments of the kind. Many substances, as, for
instance, the thinner animal membranes, are expanded by the
pressure of the column of fluid resting upon them, and thus
their efficient surface is increased, while, at the same time,
their texture is changed, their thickness diminished, and their
pores enlarged. When the experiments are prolonged for any
length of time, the substances frequently undergo changes from
putrefaction, &c. ; and owing to these causes it is scarcely pos-
sible to draw any valuable conclusions from a few experiments
of this nature, it being requisite to have many, even several
hundred experiments, from which to derive approximatively
correct mean numbers, as guiding-points for the comparison of
the mechanical capillary force of different substances with each
other, or with different fluids. It is, therefore, to be desired
that experiments of this nature were made on many different
substances, and in large numbers.

Many substances that are permeable by fluids appear to
exercise no further influence upon the penetrating fluid than
the above-named mechanical capillary attraction. We find in

ENDOSMOSIS. 95

this case, on comparing the results yielded by experiments on
different permeable substances, either differences only in the
time necessary for an equal quantity of fluid of equal superficies
and with equal pressure to permeate a substance, or differences
in equal periods of time in the quantity of the permeated fluid.
Such permeable substances do not act upon the quality of the
fluid to change it; thus, as a general rule, we find that in ordi-
nary filters, the permeated fluid has the same character as before
it was filtered, while only such particles are impeded in their
passage as are mechanically prevented by their size from passing
through the pores of the filter. The matter is not, however,
always so simple. Thus many permeable substances yield a
passage exclusively, or in a large proportion, only to certain
constituents of a compound fluid and not to others, and conse-
quently the permeated fluid acquires a different composition
from what it had when poured on the filter.

If, for instance, we pour an emulsion, composed of equal
parts of water and oil upon a thick filter having minute
pores, and that has been moistened by water, the latter will
pass freely through, but the oil more sparingly, and the per-
meated fluid will contain more water than oil. We may, in
this manner, deprive milk, by frequent filtration, of a large
portion of its butter globules.

If putrid water be filtered through animal charcoal, many
parts will be retained by the charcoal, and the filtered water
will be alike devoid of odour and taste.

If bile mixed with alcohol be filtered through animal char-
coal, the bile-pigment will be retained and the fluid appear
colourless after filtration.

If aqueous spirits of wine be inclosed in a bladder, propor-
tionally more water than alcohol will pass through the walls of
the membrane, and evaporate, whilst the mixture remaining in
the bladder will contain a relatively larger quantity of alcohol,
and a smaller one of water.

If, on the contrary, we inclose aqueous spirits of wine in a
caoutchouc bag, more of the spirit than of the water will escape
through the walls, while the fluid remaining in the bag will
contain proportionally more water and less spirit.

It would lead us too far, were we to adduce more examples

96 ENDOSMOSIS.

of this kind; those we have given suffice, however, to show that
in many cases the fluid that has passed through a permeable
substance has a different chemical composition from the one
possessed by the fluid from which it was derived.

I will not here make any attempt to investigate the primary
causes of this phenomenon. The result may be explained by
assuming that the permeable substance in some cases chemically
attracts and retains, while in others it repels certain kinds
of matter, preventing their entrance into its interstices. Both
cases yield so far the same result, that there is either no appear-
ance, or but a partial one, in the permeated fluid of either kind
of eliminated matter.

The simplest method of testing this chemical action of
different substances on various fluids consists in suffering the
latter to filter, in the manner described in the previous section,
through different membranes, and examining if, and to how
great an extent, the permeated fluid differs chemically from
what it was previous to filtration.

We will consider somewhat more attentively, according to
these general grounds, a few cases in which fluids come in con-
tact with permeable partitions.

One of the simplest cases is, that in which two like fluids
are separated from each other by a permeable partition. If the
hydrostatic pressure be equal on both sides no change will
occur. If, on the other hand, the pressure be stronger on the
one fluid, a certain quantity of this latter will pass over to the
other fluid. This quantity will correspond to the excess of
pressure on the one side, allowing for the resistance opposed to
the pressure by the mechanical capillary force. Here, how-
ever, we presuppose that the partition exercises no specific
chemical action on the fluid, but suffers it to pass unchanged.
The final result of the process consists, therefore, simply in
this, that the one fluid gives off a certain quantity to the other
without receiving anything in return; that is, the volume of
the one fluid increases as that of the other correspondingly
diminishes.

If, on the contrary, a specific chemical action of the

ENDOSMOSIS. 97

partition also come into play, that is, if the permeated fluid
have a different property from that on either side of it, the two
fluids will be chemically different, and the case then becomes
more involved, ranging itself in a class with others which we
shall subsequently consider.

Another tolerably simple case presents itself where a fluid
is on one side of a partition, and on the other a substance
attracting this fluid without being dissolved in it.

If, for instance, we fill a clay cell, or a glass tube secured
by some animal membrane, &c., with blotting paper, dried
sponge, or with dried animal or vegetable matter, fibrine,
coagulated albumen, resin, mucus, &c, and put it into a
vessel filled with water, the water will penetrate through the
partition and pass from its inner surface to the substance
chosen for the experiment, causing it to swell.

Here, too, the permeation is limited to one side; that is to
say, parts of the fluid pass through the partition without the
fluid itself receiving anything in return. The quantity of fluid
which passes through the partition in a given time depends upon
the superficies of the latter, and upon the greater or lesser
energy with which the substance in the interior attracts the
fluid without the partition, and likewise upon the quantity of
that substance.

The quality of the permeated fluid depends, (1) upon the
chemical influence of the partition on the given fluid. The
inner surface of the partition will only present to the substance
the fluid capable of passing through it: (2) upon the chemical
power of attraction exercised by the substance itself upon the
individual constituents of the fluid presented to it by the
partition.

We may also consider these cases as if a fluid were imbibed
by a permeable substance consisting of layers of different
natures (as a partition and inner substance,) and may thus
avail ourselves of them, in order, by a careful chemical investi-
gation of the fluid imbibed, (which may be obtained here in a
larger quantity,) to decide upon the chemical capillary action
which different substances exert upon compound fluids.

98 ENDOSMOSIS.

10.

The cases are more involved where two fluids of different
character, capable of being mixed, are separated by a permeable
substance. Here, owing to the mutual chemical attraction
exercised by the constituents of the one fluid on those of the
other, both fluids will strive to become chemically equal,
(see 1,) and this perfect admixture will always succeed as the
final result of their being brought in contact, although in a
shorter or longer period of time and attended by different
accessory circumstances, whilst the volumes of the two fluids
may either remain unchanged, or the volume of the one in-
crease as that of the other diminishes.

The only truly practically important point to be traced out
in the arrangement of all experiments on this subject may be
embraced in the following simple question: Which constituents of
the fluid A (as well ivith respect to quality as quantity,} pass in a
given time to B, and which pass from B to A? When we have
ascertained this, we know not only what is the chemical com-
position of both fluids after their reciprocal action has continued
some time, but likewise the amount of their volumes, by which
we learn whether the one fluid has taken from or given to the
other, or whether both have retained their volumes unchanged.
Conversely, however, this question may be easily answered by
a quantitative chemical investigation of the two fluids with
respect to their absolute amounts, and it is much to be re-
gretted that the experiments hitherto made upon the so-called
Endosmosis have been directed almost exclusively to the in-
vestigation of the changes occurring in the volumes of both
fluids, although this really only forms one single point of the
whole process, while little regard has been paid to the quanti-
tative changes occurring in fluids.

The above question can only be answered by an experiment
for each individual case, and as every single experiment, for
the reasons already given, yields results which, in spite of the
greatest caution, are not precisely alike in all cases, the answer
must be sought in many experiments, rather than in a single one.
In this respect there is still a great hiatus to be filled up, and
individual experiments, and even individual series of experi-

EXDOSMOSIS. 99

merits, afford such unsatisfactory results, that I omit giving the
data derived from a course of observations I pursued with respect
to the mixture of solutions of salt and water, and of various
salts in water.

Certain general considerations may, however, be combined,
serving partially to explain the results of many experiments,
and partially to enable us to determine beforehand some few at
least with an approximation to the truth.

The cases appertaining to the chemical mode of action of
the partition may be divided into three classes.

1. The partition may exercise no chemical action upon the
fluids: both permeating the substance freely, and Fi„ 7#
then mixing within it. The chemical composition of
the fluid within the substance of the partition,
(fig. 7) will then be the simple result of the reci-
procal chemical action of the two original fluids
(A and B, fig. 7). The partition takes no other
part in the mixture than by means of the mecha-
nical capillary force it exercises: considered from
this point of view, the mixture is effected pre-
cisely as if both fluids were in one common vessel
(see 1 to 3).

2. The partition may only suffer the constituents of one of
the two fluids to pass through it. The fluid within the partition
(C, fig. 7,) will then have the same composition as one of the
fluids (c = A or B). The result is similar to what was spoken
of at 4, where salt [is brought in contact with oil and water, or
as at 9. Only such constituents as are able to penetrate the
partition pass from A to B, but not conversely from B to A.

3. The partition may admit of the passage of the constituents
of both fluids, but in unequal proportions. (7, therefore, will
have a peculiar composition, different to that which it would
have possessed, if A and B were directly mixed.

The further process is the same in all three cases, and con-
sists in C becoming identical with B as well as with A, in ac-
cordance with the laws that obtain concerning the mixture of
fluids in common vessels, and this continues until all parts of
A, B, and C are in a state of perfect chemical and mecha-
nical equilibrium.

112

100 ENDOSMOSIS.

Let us consider somewhat more attentively a few cases of

this kind.

The most simple are those that have been spoken of under
the head 2. C, (fig. 7) = A, is attracted by 5; a portion of
the former will therefore pass over to B, without, however, any
part of B conversely passing to C. What is lost by C is again
supplied by A forcing itself in, and this newly acquired portion
of A passes again to B, until finally the whole of the fluid A
will be conveyed to B. The latter, therefore, has been
augmented by the whole mass A, and has the same composition
as if A and B had been immediately mixed, provided always that
the force with which the partition retains a portion of A in C
be not greater than the force with which this portion is
attracted by B. We have not many opportunities of observ-
ing pure cases of this kind in nature: almost all permeable
partitions of this species contain large pores through which
some portion of B may penetrate, so that instead of all the fluid
passing to B} a quantity of a mixture of A and B will be found
in A. Amongst cases of this kind, we may mention the one in
which water and spirits of wine are separated by an animal mem-
brane or by a caoutchouc plate, or that in which there is water
on one side of the partition, and gum, mucus, or some other
insoluble animal or vegetable substance on the other.

In the cases belonging to the first and third class, C is a
mixture of A and B, and will therefore give off constituents
from A to B as well as from B to A. It now depends partly
upon the nature of C, and partly upon the amount of attraction
reciprocally exercised by the separate constituents of A and B
on each other, whether more will pass from A to B, or from B
to A, or whether equal quantities will go to both sides ; the
result will, therefore, differ very much in different cases, and the
confirmation of this law of transition for different partitions and
different fluids can only be arrived at by practical experiments.

A few laws of this nature may be derived from the
results yielded by the numerous experiments of Dutrochet.
Thus, for instance, we know, that when concentrated solutions
of salts, gums, albumen, and other substances, mix by means
of an animal membrane with water, or with diluted aqueous
solutions of their own nature, more of the water will pass to

ENDOSMOSIS. 101

the salt-solution, and generally from the diluted to the con-
centrated fluids, than conversely, and consequently as the
water diminishes in bulk, the solution will increase. The
increase of volume of the concentrated fluid is, however, more
considerable in proportion to the difference in the degree of
concentration of the two fluids. When distilled water and
a very concentrated solution of salt mix through an animal
membrane, more water will pass to the salt, than if the water
and a weaker solution of salt be separated by a membrane of
like nature.

The contrary is the case with most acids when not too con-
centrated: on their mixing with water through an animal
membrane, more of the acid will pass to the water than
conversely; the volume of the acid consequently diminishes.

The quantities of matter passing in an equal time from A to
B and from B to A, depend (other relations being the same) upon
the extent of the superficies of the partition; the larger this
is, the more will pass through it, and conversely.

The motion or rest, of the two fluids, exercises also an essen-
tial influence upon the quantity of that which passes in equal
periods of time. This quantity (other relations being the same)
is greatest when both fluids are in continued motion (see 3).

The nature of the membrane is of the greatest moment:
all influences affecting its chemical and mechanical capillary
forces, diminish for the same reason the relations of the
mixture of the fluids that are separated by it. Most animal
membranes are composed of different layers, which not unfre-
quently possess different capillary forces; it is not, therefore,
always a matter of indifference as far as the result is concerned,
whether one superficies of the partition be turned towards the
one or other of the fluids, and of this we have evidence in the
experiments of Matteucci and Cima. These circumstances all
combine to make the phenomena exhibited in the mixture of
substances by means of permeable partitions very much in-
volved, rendering it extremely difficult to draw any previous
conclusions as to what may occur in a certain case.

Let us now endeavour to apply a few of the results obtained
to the processes in the human organism.

Many phenomena in the human body depend exclusively

102 ENDOSMOSIS.

upon the permeation of fluids through organic partitions, while
in other phenomena this process plays a more or less important
part. In endeavouring in the following pages prominently to
bring forward a few examples appertaining to this class, my
principal aim has been directed to the furtherance of this
object, which is of the greatest importance to physiology and
pathology, urgently recommending the pursuit to others, as a
labour of too exhausting a nature for the cultivation of one
alone.

In the first place, it is clear, that whatever the body takes
up from without as a means of nutriment, in the widest sense
of the word, can only reach the interior of the body by passing
through animal membranes in the condition of a fluid. I do
not speak here of the gases, that entering the lungs and the
external skin, pass thence into the blood-vessels and the inner
parts of the body, but will limit myself especially to the consi-
deration of those means of nutriment which enter the stomach
in the form of food and drink, and are then dissolved as far as
they were not already fluid and resorbed. How are we to
regard this resorption?

The food we partake of passes first through the mouth,
the fauces, and the oesophagus. All these parts are covered
by a thick epithelium, which does not readily yield a passage
to fluids. In the stomach, however, whose walls permit fluids
to pass more readily through them, an energetic reaction takes
place between its fluid contents and the blood flowing through
the gastric vessels, and this process is continued throughout the
whole length of the intestinal canal. The contents of the stomach
are usually much more aqueous than the blood: the various kinds
of drink, and most articles of fluid food, are naturally so; and
concentrated food becomes so by being mixed with the aqueous
saliva and gastric juice. As is usually the case where a con-
centrated fluid comes in contact with one that is more aqueous,
there is no doubt a larger quantity of the thinner fluid passes
through the membranous partition to the concentrated, than
conversely. Thus, by degrees, a larger portion of the contents
of the stomach and of the intestinal canal (both water and
the substance it holds in solution,) is conveyed into the
vascular system than, on the other hand, passes from the

ENDOSMOSIS. 103

blood to the fluid of the digestive canal, as has been directly
shown by the experiments of Poiseuille. We are still deficient
in exact researches respecting the substances that pass from the
blood-vessels to the contents of the intestinal canal, but they
are probably salts with a small quantity of extractive matters
and some protein compounds, which latter are then converted
into mucus. What is usually denominated gastric and intes-
tinal mucus is doubtlessly the equivalent yielded in the act
of digestion by the blood to the contents of the stomach
and the intestinal canal, but it will, of course, be understood
that gastric and intestinal mucus may be secreted by the blood,
in the manner we shall subsequently consider under the head
of secretions, without that fluid receiving anything in return.
It is, however, arranged in the most wonderful manner that
almost the whole contents of the stomach and intestinal canal
may gradually permeate into the interior of the organism,
without any appreciable quantity of matter on that account
passing as an equivalent from the blood to the digestive canal.
In the first place the acid character of the gastric juice seems
to be of great importance. Thus, we learn from the experi-
ments of Dutrochet that acid fluids, especially in mixing
with another fluid by means of an animal membrane, yield
more than they receive from the fluid, and thus the acid
character of the contents of the stomach seems to be precisely
the means by which resorption is promoted in a simple physical
manner. It would be, however, highly desirable to institute
a series of carefully conducted experiments on the probable
influence of the acid in the gastric juice on the resorption
of the chyme. This latter exhibits its acid property not only
in the stomach but also at the entrance of the small intestines,
and retains it through the great portion of the passage through
it. We consequently see that this aid to the resorption of
the chyme is not limited to the stomach alone.

The more two fluids separated by an animal membrane
differ in their degree of concentration, the more proportionally
will the concentrated take up from the thinner fluid. The
passage of the nutriment into the blood is, therefore, essentially
promoted by the commingling of other watery juices, as, for in-
stance, the bile and the pancreatic fluid, to which may be added,

104 ENDOSMOSIS.

the gratification of the desire to drink that is generally felt soon
after eating, and thus contributes to the dilution of the contents
of the stomach. The bile contains in itself much more water than
the blood; consisting, on an average, of about 10 per cent, of
solid constituents, while the blood contains more than 20 per cent.
The bile, however, becomes yet more aqueous in the intestinal
canal, owing to a portion of its solid constituents (the bilate of
soda) being decomposed by the acid of the chyme, and separated
in a partially modified condition as an insoluble substance,
dyslysin, &c. It contributes, therefore, essentially to the
dilution of the chyme, facilitating its resorption. In like
manner the pancreatic fluid acts, which, although we know
but little of its composition, is, at all events, much more
aqueous than the blood (containing about 8 per cent, of solid
constituents). Another means, acting with the same object,
is the circulation of the blood, the consequence of which is,
that the portion of blood become more dilute by the resorp-
tion of aqueous particles from the chyme, is constantly removed
and replaced by more concentrated blood. Owing to all these
relations, an augmented difference is brought about in the con-
centration of the chyme and of the blood, and the largest part of
the food is thus conveyed from the intestinal canal directly into
the blood-vessels, without an equally large quantity of the
contents of the latter being transferred to the intestinal canal.
Numerous experiments require, however, still to be carefully
conducted regarding this point, with special reference to the
question, whether all constituents of the chyme pass alike
readily into the blood, or whether there is a difference in this
respect.

In the meantime, notwithstanding these arrangements, and
especially after partaking of solid or very rich food, a portion
of the chyme will often remain in so concentrated a condition
in the intestinal canal that no part of it can pass into the
blood-vessels through an interchange of their contents. Special
provision is made for these cases by means of the chyle-vessels,
or lacteals. Whatever may be conjectured concerning the origin
of these vessels, so much is certain, that they arise from cavities
in the intestinal villi which do not communicate with the
intestinal cavity by open mouths, but are separated from it

ENDOSMOSIS. 105

by a membranous partition, — mucous membrane with epi-
thelium. The passage of the contents of the intestinal canal
can, therefore, only be effected by a penetration through this
partition. Here, again, we find different arrangements tending
to promote the passage. In the first place, the chyle-vessels
after fasting are filled with a fluid, lymph, which, although
less concentrated than the blood, is more so than the chyle,
and, like the blood, draws more from it than it yields. This,
however, is only a secondary agency, and the chyle- vessels
stand far below the blood-vessels in the active exercise of this
function. A second, and far more efficient agent, which
transfers the fluid part of the contents of the intestines into
the chyle-vessels, and then conveys it further on its course,
is of a mechanical nature. By the peristaltic action of the
intestinal canal (the contraction of its muscular coats) so con-
siderable a mechanical pressure is exercised upon its contents
that a portion of them (in a fluid condition, of course,) is forced
through the walls of the intestines, and conveyed into the
chyle-vessels to be again propelled onward. As the valves
of the chyle-vessels prevent the return of the chyle on the
cessation of the peristaltic motion of' the intestinal canal, the
portion of the chyme which, owing to its concentration, could
not pass directly by diffusion into the blood, is, in this manner,
circuitously conveyed into the vascular system. It follows,
therefore, that, on the occurrence of deficient or abnormal
contraction of the intestinal canal, that the part of digestion
which depends upon resorption through the chyle-vessels must
suffer a disturbance.

The chyle-vessels further aid in the process of digestion
by serving for the resorption of fat. Fat does not mix with
water, and as it can either not at all, or only with difficulty,
pass through membranes moistened by aqueous solutions, it
cannot penetrate in more than a very small quantity from the
chyme into the blood-vessels. Fat may, however, penetrate
from the intestinal canal into the chyle-vessels, as has been
shown by experiments made on the chyle of animals which
have been killed soon after partaking of fatty food. The
permeation of fat through the intestinal wralls into the chyle-
vessels is no doubt effected in the same way that oil penetrates

108 ENDOSMOSIS.

through a filter moistened with water; thus, in some parts
of the filter, the water is displaced by the long contact with
particles of oil, so that the filter becomes saturated in these
parts, and thus forms, as it were, bridges for the passage of the
succeeding particles of oil, which are driven over in the same
manner as the aqueous parts of the chyme are forced by mecha-
nical pressure into the chyle-vessels. Fat, therefore, is not
dissolved by the action of a digestive fluid, but the solution is
effected by the heat of the body, and fatty mixtures whose
fusion-point is above 40° C, are either not at all digested,
that is resorbed, or only gradually dissolved by the agency of
more fluid fat subsequently taken into the system by food.
In the resorption of fatty food, the surface of the intestinal
canal divides itself, as it were, into two parts, of which the
one resorbs aqueous chyle, and the other fat. The digestion
of fatty food is more slowly effected than that of aqueous
substances, owing to the time required by the fat to force
the water out of the walls of the separate villi. The resorption
of aqueous fluids is rendered more difficult, when fat has been
partaken of on an empty stomach, as the fat deposits itself
upon the surface of the "intestinal canal, covering it with a
coating of oil, which hinders the entrance of the water. This
explains the reason of the inconvenience experienced on par-
taking copiously of water after eating fatty food, and, in like
manner, why the effect of intoxicating drinks, as ale, for in-
stance, may be retarded for some time by taking a few spoons-
ful of oil on an empty stomach.

A physiological consequence derived from the views pre-
viously given with reference to the resorption of alcohol, is
worthy of remark. If alcohol be diffused with watery fluids
through animal membranes, only a small portion of the spirit
will pass to the water, while a great deal of the latter will
go over to the spirit. This evidently indicates that spirits of
wine, when taken into the stomach in a highly concentrated
condition, can only reach the blood slowly, and in a diluted
state. Thus, the spirit cannot possibly retain its usual action
of coagulating albumen, after it has entered the blood of the
living organism.

In certain cases, the resorption of food and drinks in the

ENDOSMOSIS. 107

intestinal canal is impeded by the action of medicines. This
in the case with salts, for instance, which almost all have a
purgative action, and arises from the fact, that concentrated
saline solutions take more from the fluid of the blood in pass-
ing through animal membranes, than they yield to it; this
again has been directly shown by Poiseuille, although, as in the
case of other points, accurately conducted experiments and
observations are still required to determine what are the sub-
stances that pass from the blood into saline solutions, and in what
relations the exchanges occur. Opium, as is well known, checks
diarrhoea; and this phenomenon, according to Poiseuille's investi-
gations into the subject, admits of being explained by the changes
effected in the capillary forces of the coats of the intestines; thus
on the addition of opium to a saline solution, which is thoroughly
diffused with the serum of the blood by means of an animal
membrane, contrary to what is the case before the opium
is added, more of the saline solution is attracted than is
given off; but here again we must regard further chemical
investigations as highly desirable. Probably, at some future
period, when experiments have been more fully pursued regard-
ing this point, many other actions of medicines upon their
resorption in the intestinal canal will have to be referred to the
above-named phenomena of diffusion.

It is only such constituents of the contents of the intestinal
canal as are dissolved which can be resorbed; all undissolved
or insoluble matter passes off with the excrements. This seems
to be so clear, that I should not have deemed it necessary to
speak of it, were it not, that in recent times some physiologists
have maintained the view, that even undissolved substances
might penetrate, when reduced to the condition of the finest
particles, into the (non-existing) open mouths of the chyle-
vessels.

In the occasional instances where the external skin has
reorbed liquids by means of baths or frictions, the case is
precisely the same as in the intestinal canal, only that here,
owing to the greater thickness of the epidermis, the action is
less energetic, so that resorption is principally limited to those
parts, where the epithelium is thinnest; as, for instance, in the
numerous glands of the skin.

108 ENDOSMOSIS.

We will now consider the processes, in which fluids pass by
diffusion from the interior of the organism — processes which are
usually designated by the term secretion. I will here again
limit myself to liquids, as the laws according to which elastic
fluids are diffused through liquids, as for instance, in the process
of respiration, are still less known.

One of the simplest of these processes is the transpiration of
the skin, and the secretion of sweat.

The former evidently arises from the skin, that is to say, the
cutis and the under layers of the epidermis, (the so-called rete
malpighii) being saturated with a fluid, originating from the
blood, and of which a certain quantity, composed of water,
with gases and other volatile constituents, evaporates in the air,
whilst the non-volatile constituents remain. In the formation
of sweat, we have an opportunity of collecting this fluid in a
larger quantity, and of examining it to compare its composition
with that of the blood, from which it is evidently derived.
There is, however, this great difference between the two, that
sweat is very much more dilute than blood, containing besides
the water only a small quantity of salts and extractive matter.
We must consider somewhat in the following manner the many
agents which contribute towards its formation. Its primary
source is the blood, which everywhere in the vascular system is
exposed, owing to the contraction of the heart, to a greater
hydrostatic pressure than the fluid in the cellular tissue, &c.
Thus, a certain quantity of the liquor sanguinis passes constantly
out by means of the walls of the vessels, this quantity correspond-
ing to the excess of hydrostatic pressure in the vessels over that
of the external fluids, minus the mechanical capillary force of
the walls of the vessels. The chemical composition of this fluid
depends upon the chemical capillary force of the walls of the
vessels, that is to say, it is changed into a modified form of
liquor sanguinis. Whilst this fluid presses to the upper sur-
face of the skin into the lumina of the perspiratory glands, it
must permeate through other organic partitions, the tissue of
the cutis, the walls of the perspiratory glands and their
epithelium. These, however, act upon it in such a manner as
to change it; that is to say, they suffer some of its constituents
to pass more easily than others, owing to which causes the com-

ENDOSMOSIS. 109

position of sweat differs considerably from that of the liquor
sanguinis, from which it is derived. We will endeavour to
follow these relations more thoroughly in several individual
cases. In the first place, it is evidently of great consequence
to know how the fluid is composed which escapes from the
blood through the vascular walls, and which, owing to its
being distributed in every direction where the blood-vessels are
ramified, I will designate by the name of the general nutrient
fluid. It will be seen at once that its character is dependent
upon several conditions: 1. On the character of the blood.
2. On the character of the vascular walls, that is, upon their
mechanical and chemical capillary forces; it is, therefore, pro-
bable that it may differ in the veins and in the capillary vessels ;
and it is further possible, that different capillary vessels, even
in their normal condition, admit of the passage of a different
fluid through their walls. 3. Its character doubtlessly depends
upon the relations of pressure in the vascular system, so that
a vascular wall that is very much thickened and contracted, will
admit of the passage of a different fluid from what is trans-
mitted through a very relaxed wall. But as the relaxation and
contraction of the vascular walls depend upon the nervous
system, we perceive the important influence exercised by the
latter upon the secretion. As yet we unfortunately know but
little that can be relied upon regarding the influence of all
these relations on the character of the permeated fluid, and a
wide field is thus opened to us in which great and lasting
services may be rendered to science by the earnest pursuit of a
series of carefully conducted experiments. From a comparative
consideration of the composition of different normal and patho-
logical fluids, which may be regarded as more or less changed
nutrient fluid, as lymph, serum of pus, inflammatory exu-
dations, &c, we learn that the general nutrient fluid forced
by the hydrostatic pressure of the blood from the capillary
vessels, contains essentially the same constituents as the liquor
sanguinis, although in a different proportion. It generally con-
tains more water, somewhat more of extractive matters and
salts, but on the other hand, less of the protein compounds
(albumen and fibrin) than the liquor sanguinis. The fluid
which passes through the walls of the veins (serous dropsy),

110 ENDOSMOSIS.

appears to differ from the general nutrient fluid by the absence
of fibrin. It may also, as experience testifies, suffer consider-
able alterations in its character, either owing to dilatation or
contraction of the vessels, or from changes occurring in the
blood, but as yet our knowledge is very deficient on this
head.

Let us now turn from the consideration of the general
nutrient fluid to that of the formation of sweat. Sweat is
ultimately derived from this fluid, which is far more dilute
than the liquor sanguinis, containing more salts and extractive
matters, and less of the protein compounds, than the latter. Before
this fluid reaches the perspiratory glands it experiences a second
filtration through the substance of the cutis, and the walls
and the epithelium of those glands. The latter retains certain
constituents of the nutrient fluid (as the protein compounds,
converting them to their own epigenesis and nutrition,) and a
portion of the salts; in return they probably give off to the
permeating fluid a small quantity of matter produced by their
chemical process of development; as, for instance, extractive
matters. The sweat consequently differs very considerably
in its composition from the liquor sanguinis, containing,
besides water, only a trifling quantity of salts and extractive
matters. Sweat may assume a different composition owing
to pathological causes, partly from the changes thus effected
in the capillary relations of the blood-vessels (when, conse-
quently, an abnormal nutrient fluid will be secreted), and
partly owing to the changes occurring in the capillary relations
of the skin and the epidermis. There exists a certain similarity
between the conditions of the formation of sweat and those of
the different pathological fluids which we observe to make
their way through the skin, as the contents of blisters produced
by burns, cantharides, pustules, bullae, &c. On removing the
epidermis, a portion of this fluid is converted into pus, owing
to the fact that the protein substances which, in the normal
condition, contributed to the regeneration of the epidermis,
are now spontaneously developed into pus-cells.

The lachrymal secretion is precisely similar to the formation
of sweat. Here, too, the secretion, which doubtlessly origi-
nates from the blood, or the general nutrient fluid, is so

EXDOSMOSIS. Ill

modified in its passage through the walls of the lachrymal
glands and their epithelium, as to consist almost exclusively
of water, with a small quantity of salts and extractive matters.

Besides the mere discharge of fluids from the blood by
diffusion, a new agent is added to most secretions. Thus a
portion of the permeating fluid is modified by an organo-
chemical process, and mixed in this altered form with the
secretion. This is most plainly shown in the secretions of
the reproductive organs, — in the testicles and ovaries, — where
this portion forms the principal mass of the secretions. Thus,
for instance, in the testicles by the fluid passing into the secret-
ing canal, and doubtlessly from its protein compounds, organised
structures, the seminal filaments, or spermatozoa, are deve-
loped, which, together with the remaining fluid, constitute the
secretion.

A very general product of this kind, occurring in almost
all secreting canals, is mucus. It is formed upon all mucous
membranes, that is to say, upon all very vascular surfaces
covered with a thin epithelium. Its origin may probably be
explained on the hypothesis that the more scanty cells of the
epithelium do not, like the numerous layers of the epidermis,
consume the whole protein contents of the secretion in their
own nutrition, but leave a portion which, combined with
salts and alkaline bases, passes onwaids in a modified form;
composing a mucous investment to the secreting canal. The
common law, as it may be termed, according to which the
protein substances disappear that had passed during the normal
secretion from the general nutrient fluid to the secreting canal
(being either converted into mucus, or applied to purposes
of nutrition), further explains the fact, which at first sight
appears so paradoxical, that no secretion (with the exception,
perhaps, of the pancreatic fluid,) contains, in its normal con-
dition, coagulable albumen, whilst both the blood and the
nutrient fluid are so rich in this substance.

Milk alone affords an exception to this rule. It contains
no albumen, but casein, which, as Scherer showed, is only a
modification of albumen produced by combination with alkaline
bases. The mammary gland exhibits so many peculiarities in
the character of its secretion, that we will devote some little

112 ENDOSMOSIS.

time to its consideration. In the first place, its secretion is
highly concentrated, (containing, on an average, 15 per cent, of
solid constituents,) and it thus approximates more to the cha-
racter of the blood, forming a strong contrast to the highly
dilute secretions of sweat and tears. The walls of this gland
admit of the passage of the nutrient fluid in a highly concen-
trated condition, but exercise such an influence upon it as to
convert the albumen into casein, that is, by combining it with
alkalies. This, doubtlessly, is first effected by the action of the
cells.

Another peculiarity in this gland consists in its taking up
fat and extracting it from the blood. This action likewise
seems to originate in the glandular cells, where fat is attracted,
and after being dissolved within the cells is liberated, and then
mixes with the milk in the form of globules. Although not
impossible, it seems very improbable to me that fat should be
first formed in the mammary gland from other substances.

Other glands, as the sebaceous glands of the skin, viz.,
those in the axilla?, the glands of the generative organs, and the
ceruminous glands, likewise possess the property of attracting
fat from the blood or the nutrient fluid. If once such a gland
be filled with fat the latter substance will, owing to physico-
chemical grounds, exclude the passage of aqueous fluids, and
the gland will then only take up fat. The fatty cellular tis-
sue in the parenchyma of the body appears to have the same
property.

The salivary glands stand at a medial point between the
perspiratory and mucous glands. Their secretion is very
aqueous, containing scarcely 1 per cent, of solid constituents,
and besides water exhibits only a small quantity of salts and
extractive matters, the protein compounds of the permeating
nutrient fluid being converted into mucus.

It still remains to be proved whether the salivary matter
contained in this secretion had existed in the blood, and been
attracted in a perfectly developed form in the glandular cells, or
Avhether it is formed within the glands by the chemical action
of the cells.

The glands of the stomach exhibit many peculiarities: the
product of their secretion so far coincides with the saliva that

KNDOSMOS1S. 113

it contains watery fluid with salts and extractive matters, while
the remains of dissolved glandular cells testify to the presence
of protein consumed and changed by organization. But the
gastric juice exhibits the peculiarity of containing a free acid.
The manner in which this acid is conveyed to it is so much the
more difficult to understand, that we do not even well know its
chemical character, for as this acid was formerly regarded as
hydrochloric acid, it has been supposed by more recent inves-
tigations to be lactic acid, combined with a little phosphoric
acid, which latter, however, simply owes its origin to the
action of the lactic acid upon the alkaline phosphates simul-
taneously present. By what means is the presence of this free
acid brought about? The answer to this question is not easy,
and many conjectures may be hazarded on the subject. The
acid owes, without doubt, its origin to the blood, and the pro-
cess of its formation may be owing to the decomposition of
certain salts of the blood, (alkaline lactates or chlorides) by the
cells of the gastric glands which retain the acid, whilst the
alkali returns by diffusion into the blood. As the cells gra-
dually become broken up and dissolved, the acid is liberated and
mixes with the gastric juice.

The opposite to this occurs in the liver, whose secretion
contains soda combined with a weak organic acid, bilic acid.
The processes in the secretion of the bile are not clearly un-
derstood, principally because we do not know whether the biliary
matter is contained as such in the blood, or whether it is first
formed in the liver. Either may possibly be the case, and we
may thus with tolerably equal grounds of justification oppose
to each other two different views concerning the origin of the
bile. The constituents of the bile may be contained as such in
the blood. This appears certain with respect to some of them,
(as, for instance, the bile-pigment,) which are attracted by
and enter the hepatic cells, and by their solution these consti-
tuents are liberated, and with the fluid which passes from the
liquor sanguinis by diffusion into the biliary ducts form bile.
Or we may suppose that the fluid entering the biliary canals of
the liver is first converted into bile by the chemical activity of
the hepatic cells, when, probably, many products of these meta-
morphoses are returned to the blood by diffusion in conse-

114 ENDOSMOSIS.

quence of the reciprocal action exercised upon each other by
the bile and the blood of the hepatic capillaries.

It still remains for us to consider the urinary secretion. The
urine consists in the first place of a very aqueous fluid, com-
posed similarly to sweat, of water principally, with some few
salts and extractive matters, which passes by diffusion from the
blood into the urinary canals. It contains, however, additionally,
specific substances, as urea, uric acid, hippuric acid, and their
salts, and colouring matter, and these, at least with reference
to the quantities in which they occur, are peculiar to the urine.
These substances are certainly present in the blood, but the
quantity in which we find them here, renders the assumption
probable that owing to a peculiar power of attraction inherent,
no doubt, in the glandular cells of the urinary canals, they are
attracted by these from the blood into the cells, and thus passed
into the urine.

According to the explanations given by Bowman regarding
the histology of the kidneys, we may represent the matter to
ourselves somewhat in the following manner.

In the malpighian bodies, that is, in the vascular loops which
penetrate immediately into the entrance of the urinary canals,
an aqueous fluid, holding salts in solution, forces itself through
the walls of the renal vessels into these tubes. In its further
course through these canals, the peculiar constituents of the
urine, as uric acid, hippuric acid, colouring matter, urea, &c, are
taken up in large quantities by the cells from the blood, and
either pass into the urine by the disintegration of these cells,
(by which they are liberated) or are taken up from the glandu-
lar cells of the urinary canals by means of the above-mentioned
watery fluid originating in the malpighian bodies.

The blood-producing glands, as the spleen, the supra-renal
capsules, thymus, and thyroid, also manifest peculiar relations.
Substances occur in them, which passing by diffusion from the
blood, undergo many modifications in the glands themselves, as,
for instance, by cellular formation, and after being to a certain
extent modified, are again returned to the blood by diffusion.

We may thus, in a great measure, deduce the peculiarities
of almost all the secretions from physical and chemical relations,
and there are, therefore, grounds for hoping, that at no very

EXDOSMOSJS. 115

remote period, when organic chemistry shall have made further
advances, we may learn to know many of these processes in
their normal and pathological relations.

The two widely diffused processes of the absorption of the
nutriment and of secretion are not, however, the only ones in
the organism in which diffusion of fluids plays a prominent
part. They are merely the two visible and extreme members
of a long series of processes, which are connected together
like the links of a chain. Everywhere in the interior of the
body similar phenomena are continually occurring, which,
however, owing to their multifarious forms, and the constant
change of the conditions on which they depend, do not admit
of being pursued in detail, but must be simply sketched in
their rough outlines. All portions of the body are saturated
with this general nutrient fluid, of which we have already
spoken. The first condition requisite to its secretion is an
excess of hydrostatic pressure in the blood-vessels, by which
a portion of the liquor sanguinis modified by the mechanical
and chemical forces of the vascular walls, is forced out of the
vessels. If, on the one hand, the composition of the blood,
and, on the other, that of the nutrient fluid, remain unchanged,
while the mechanical and chemical capillary forces of the
intervening vascular walls are likewise the same as before,
both fluids will soon be brought into a condition of equilibrium
and cease to act further upon each other.

These relations are nevertheless subject to continual fluc-
tuations in every part of the body. In the first place, the
nutrient fluid is constantly changed by the tissues with which
it is brought in contact. These are incessantly depriving it
of some of its parts, at the expense of which they increase,
decomposing others, and contributing new ones — the products
of their own chemical metamorphoses. Thus, the nutrient
fluid is ever being chemically changed, and hence follows
naturally a continual new chemical admixture with the fluid
contents of the blood-vessels. Another agent ever at work in
altering the relations of equilibrium of these fluids is of a
mechanical nature. The hydrostatic relation of equilibrium
existing between the liquor sanguinis and the nutrient fluid
is dependent, on the one hand, on the pressure of the blood in

116 ENDOSMOSIS.

the vessels, and, on the other, on the pressure to which the
general nutrient fluid is subjected. As the pressure of the
blood in the vessels undergoes constant change from the in-
fluence of the vaso-motor nerves, occasioning dilatations and
contractions of the vessels, and then again from the alternating
contractile action of the heart, the hydrostatic pressure upon
the nutrient fluid is likewise of an alternating character. It
is continually changed by the local contraction or relaxation of
the muscles and other contractile tissues, by temporary relax-
ations and contractions of the lymphatic vessels, which either
give a free passage to the nutrient fluid (lymph) flowing into
them by diffusion, or momentarily retain it, and thus increase
the hydrostatic pressure. Another agent of change is afforded
by the alternating property of the blood. This fluid, parting
first with its watery constituents in the organs of secretion,
becomes, in consequence, more concentrated, and then again,
taking up water in the intestinal canal, is rendered more dilute;
the character of its chemical composition being thus liable to
greater or lesser changes at every moment in every part of the
body. As now all these agents, that are at one time constant
and lasting, at another periodically variable, are continually
changing in their degree of intensity, a source is opened to the
motions of diffusion in the fluids of the body, which, sometimes
inclined in one direction, sometimes in another, are never at
rest, and manifest their presence by keeping up that constant
molecular progression of the fluids which is essentially necessary
to the stability of life as the basis and auxiliary of the meta-
morphosis of matter, but which not unfrequently becomes the
cause of disease.

To those already considered we might add other motions
of diffusion, which without being in immediate connection with
the circulation of the blood and the lymph, occur between the
various cells, whilst the fluid contained within each individual
cell constantly modified by chemical metamorphoses, endeavours,
by changes of its mechanical and hydrostatical relations, to
bring itself into a state of equilibrium with the parts surround-
ing it, in which it is, however, continually disturbed by the
intervention of new causes.

These considerations might be carried out much more in

ENDOSMOSIS. 117

detail, and would admit of being exemplified by many illustra-
tions, did I not deem it superfluous to enlarge further upon the
subject. My main object was simply to draw attention to those
important phenomena of the diffusion of fluids in the body,
which together with the molecular chemical processes constitute
the great department of the metamorphosis of matter, and to
encourage others to* institute a numerous series of experiments,
and collect materials in larger quantities than we have as yet
possessed them, in order to enable us to give an exact and
mathematical basis to our labours in a region of darkness, that
has hitherto been solely occupied by vague hypothesis.

118

IV.

REPORT.

UPON ISOMORPHISM.

By PROFESSOK OTTO, of Brunswick.

Gay-Lussac was the first who made the observation that a
crystal of potash-alum on being pnt into a solution of am-
monia-alum, continues to increase without changing rts form,
and that we may therefore cover this crystal with alternate
layers of the two kinds of alum without injuring its regu-
larity or its original form. Beudant subsequently found that
a mixed solution of sulphate of copper and sulphate of u-on
would give crystals, having the form of the latter but till
containfng a considerable quantity of copper Mitscherlieh
was, however, the first who by means of h* investigations
into the crystalline forms of chemical compounds, observed
an identity of form in a sufficient number of cases, to deter-
mine that this was the consequence of a similarity m composi-
tion. He termed this similarity in crystalline forms isomor-
phism (from Unlike, and poph form), and the bodies which
assume one and the same form, he named isomorphousboA™.
The law from which this is derived is as follows :-/m» «»
equal number of atoms, similarly combined, a crystal of like Jorm
is produced. The crystalline form is dependent upon the
number and the relative position of the atoms, and not upon
their chemical nature. „ .

We may easily perceive, that if this law admit of general
application, the crystalline form must give an extremely im-

ISOMORPHISM. 119

portant aid in the judgment of the atomic constitution of chemi-
cal compounds. For if equal numbers and like arrangements of
the atoms produce like crystalline forms, two compounds of
the same crystalline form must have a like atomic constitution.

The law is not confirmed in its full extent, and it has also
hitherto been impossible to follow it through the whole series of
elements, since the cases of comparison are circumscribed by
the disinclination shown by certain groups of elements to enter
into combinations, whose composition is perfectly similar to
that of the combinations of other groups.

In the year 1839, Kopp discovered the connexion of iso-
morphism and the equivalent volume, (atom-volume,) and
showed definitively in 1840, (Annal. der Chem. und Pharm.,
bd. 36, s. 1,) that isomorphous bodies have the same equivalent
volume, or, in other words, that, in isomorphous bodies, the
specific gravities stand in the same relations as the equivalent
weights, — or that equivalent quantities of isomorphous bodies Jill
an equally large space.

Before we proceed to the special consideration of isomorph-
ous bodies, to learn under what limitations we must admit the
validity of Mitscherlich's law, and how far the proposition of
Kopp is confirmed by experience, it seems to me to be neces-
sary, or at any rate advisable, to characterise somewhat more
empirically those bodies which we name isomorphous, and at
the same time investigate how far they may be distinguished in
crystallization from those which are not isomorphous with them.
This will be a further elucidation of the previous observations
of Gay-Lussac and Beudant.

If we add chloride of sodium, or chloride of potassium, to an
aqueous solution of nitrate of potash, and evaporate the solu-
tion till crystallization occurs, the nitrate of potash separates
and the chloride of sodium remains in the mother-liquid. The
solid mass of the crystals of nitrate of potash contains no chloride
of sodium; the chloride of sodium, which is very commonly found
mixed with nitrate of potash, arises partly from the mother-liquid
drying on the crystals, and partly because the latter, if of
considerable size, are internally hollow, and retain the mother-
liquid in their cavity. By repeated crystallization the whole of
the chloride of sodium may be removed, and the crystals left

120 ISOMORPHISM.

perfectly pure. The same is the case with respect to a solution
of carbonate of soda, to which a little sulphate of soda has
been added. It yields on evaporation crystals of carbonate of
soda, which contain no sulphate of soda in their solid mass,
but are often rendered impure by this salt in the mother-liquid
adhering to or occurring between the crystals; the latter may,
however, be obtained perfectly pure by repeated crystallizations.
The same thing may be observed in a solution of sulphate of
copper to which persulphate of iron has been added; on evapo-
ration it yields crystals of sulphate of copper, which can only
contain a portion of the salt of iron by the adhesion of the
mother-liquid, in which this salt remains. The copper salt
may, by repeated crystallizations, be obtained perfectly free
from the adherent iron salt.

We may easily perceive that the separation of salts by
frequent crystallization, depends upon the incapacity of many
salts to enter into the crystalline mass of other salts ; this
operation is, therefore, very frequently had recourse to in the
purification of salts. The less easily soluble of two salts, or,
where the solubility differs but little, the one that is present in
the larger quantity, will first crystallize ; and the mother-liquid,
which then contains the remaining salt in excess, will, on further
evaporation, yield crystals of the latter.

It is, however, quite different in other cases. Common alum,
the well-known hydrated double salt of sulphate of potash and
sulphate of alumina, readily separates in colourless crystals;
the so-called chrome alum, an analogous hydrated double salt,
composed of sulphate of potash and sulphate of the oxide of
chromium, separates just as easily in decidedly violet crystals.
On dissolving common alum and chrome alum together in
water, the crystals of the former do not separate from those
of the latter, but we obtain crystals that exhibit throughout
their whole mass a mixture of common alum and chrome
alum, and are of a lighter or darker violet colour in pro-
portion as the former or the latter preponderates in the solu-
tion. Crystals of every shade of colour may thus be obtained,
and no separation of the two salts can be effected by repeated
crystallization. A solution of iodide of potassium, to which a
little chloride of potassium has been added, does not on evapo-

ISOMORPHISM. 121

ration yield crystals of pure iodide of potassium, and leave the
chloride of potassium in the mother-liquid, but it gives crystals
containing iodide of potassium and chloride of potassium, which
are rendered impure by the adhesion or enclosure of the mother-
liquid with its chloride of potassium, and we are unable to obtain
by crystallization an iodide of potassium perfectly free from
chloride of potassium. It has already been mentioned, that a
solution of sulphate of copper mixed with persulphate of iron,
will yield only crystals of sulphate of copper on evaporation,
whilst the iron salt remains in the mother-liquid. If, however,
instead of persulphate of iron we add sulphate of the protoxide,
the two salts will not remain separate, crystals appearing which
contain, throughout their whole mass, iron as well as copper.
However frequently we may repeat the crystallization, we shall
always obtain crystals of this kind; a separation of the iron
protosalt from the copper salt cannot, therefore, be effected by
recrystallization. If we change the protoxide of iron in the
solution into sesquioxide of iron, as, for instance by the
action of chlorine, we shall obtain by evaporation crystals of
sidphate of copper, while the iron salt now remains in the
mother-liquid. It is only in this manner, that by recrystalliza-
tion we can remove the iron from sulphate of copper containing it.

Salts, and bodies in general, which enter jointly in this
manner into a crystalline mass, and can occur together in crys-
tals in the most different proportions, have been termed iso-
morphous, because the capacity of crystallizing in common is to
be explained by similarity of form. Common alum crystallizes
in octahedra; chrome alum appears in the same form; if both,
therefore, occur in common in a fluid, particles of the one may
combine in any and in alternate relations with those of the
other to form a crystal. This is likewise the case with chloride of
potassium and iodide of potassium; both salts are individually
separated in the same form by crystallization, and from mixed
solutions cubic crystals are produced, consisting of alternating
particles of chloride of potassium and iodide of potassium, as
both have an equal capacity for the process of crystallization.

Kopp has shown, as has been already mentioned, that on
the occurrence of such a combined crystallization, equivalent
quantities or weights of bodies must fill an equally, or at

122 ISOMORPHISM.

least nearly equally large space. The equivalent of common
alum is 5931, its specific gravity, according to Kopp, is 1*724;
the equivalent volume is consequently 1724" 3440* The
equivalent of chrome alum is 6246, the specific gravity is 1*848 ;
and the equivalent volume is consequently -p|^ = 3380. We
thus see that the two equivalent volumes are not absolutely of
equal size. This cannot surprise us, as neither the equivalent
weights nor the specific gravities, on which the calculation of
the equivalent volumes are based, are absolutely whole num-
bers. Where we find a close approximation to an agreement,
the want of perfect accordance may generally be ascribed to
the inaccurate determination of the specific gravity. But even
in differences of the equivalent volumes, which, from their
greater amount, cannot be referred to the want of exactness in
question, there is either a similarity, or at any rate so great a
resemblance of the forms exhibited, as to show us that we are
still dealing with isomorphism. If, for instance, the crystals of
compounds having nearly the same equivalent volume belong
to the regular system, the influence exercised by the deviation
in the equivalent volume will extend, in like manner, in all
dimensions, and in all three axes of the crystal, and, conse-
quently, the form must remain wholly unchanged as the angular
dimensions continue the same. But if the crystals belong to
other systems, the influence spoken of will be extended in a
different way upon the different axes of the crystal, a change
of the angular relations taking place, and the form not, there-
fore, remaining absolutely the same. In this case, however,
we also term the bodies isomorphous ; at any rate, the sugges-
tion made of replacing the word isomorphous by homoimor-
phous, (similarly formed,) has not been generally adopted.

Mitscherlich has ascertained, as already mentioned, that a
like chemical constitution corresponds with the similarity of
form of isomorphous bodies.

Common alum is ... KS + Al A1S3 + 24H

Chrome alum is ...KS + CrCrS3 + 24H

The constitution, like the form, is the same. If we consider
these formulae, we see that the chrome alum represents an alum

ISOMORPHISM. 123

in which sulphate of alumina is replaced by sulphate of the
oxide of chromium. As this substitution occurs without any
change of form being induced, we are accustomed to consider
sulphate of the oxide of chromium as isomorphous with sulphate
of alumina. In these two salts, however, there is the same
acid, consequently the oxide of chromium and alumina are also
isomorphous with each other; or we may say that chrome alum
is an alum in which the alumina is replaced by the oxide of
chromium without any change of form. As now the oxide of
chromium and alumina have one constituent in common, namely,
oxygen, chromium and aluminum must also be isomorphous;
or we may finally say, that chrome alum represents an alum in
which there is chromium in the place of the aluminum of com-
mon alum. Chloride of potassium and iodide of potassium are
isomorphous, both containing the same metal, potassium; con-
sequently, chlorine and iodine are isomorphous.

In accordance with the view, that isomorphous combina-
tions proceed from the union of isomorphous constituents, which,
as may easily be perceived, leads us to the groups of isomor-
phous elements, we define isomorphous bodies as those which
may replace each other in chemical combinations, of course
where the weights are equivalent, without any change of
form being induced. We say that the oxide of chromium
may replace alumina, in combination, without a change of
form, protoxide of iron may replace protoxide of manga-
nese, chromium may replace aluminum, and iron may replace
manganese. The formulae which are applied to combinations
in which isomorphous bodies occur, correspond with this view,
which, as we shall see below, is not perhaps throughout
correct.

The following is the formula for that spathic iron, which,
besides protoxide of iron, also contains protoxide of manga-
nese and magnesia.

Fe I

Mn [C. This formula does not designate a combination of 1 equiv.

Mgl

of carbonic acid, 1 equiv. of protoxide of iron, 1 equiv. of pro-
toxide of manganese, and 1 equiv. of magnesia, but a combina-
tion in which there is added to 1 equiv. of carbonic acid 1 equiv.

124 ISOMORPHISM.

made up of protoxide of iron, protoxide of manganese, and mag-
nesia together. The amount of oxygen contained in these three
different bases is, when taken together, 1 equiv., and we should

[Fe )

therefore, perhaps, be more correct if we were to writei ]\|n \Q,

(Mg)
The formula for the mixt alum composed of alumina alum
and chrome alum would, according to this, be as follows: —

. ... Al Al ...
K S { S3 + 24H, &c.

CrCr J

As the upright bracket occupies a considerable amount
of room in printing, it would be better perhaps to place the
symbols of isomorphous bodies alongside, instead of above one
another, in order to be able to connect them by means of a
horizontal bracket, as for instance, thus:

Fe Mn Mg C; K S, Al Al Cr Cr S3 + 24H.

I wonder that those critics, who suffer nothing to escape
their attacks, have not animadverted upon the customary
mode of writing the formula for the combinations in which
isomorphous bodies are contained. They might possibly main-
tain that these formulae are quite unwarrantable, especially
where the word atom is used for equivalent. We designate
by the formula Fe O, C 02, a combination which we suppose
consists of 1 at. of protoxide of iron and 1 at. of carbonic acid ;

FeO |
the formula Mn 0 > C 02 designates, according to the strict

MgOJ
meaning, a combination of 1 at. of carbonic acid with alter-
nating fractions of 1 at. of protoxide of iron, protoxide of
manganese and magnesia, which is evidently at variance with
the indivisibility of atoms.

It is, however, clear that this formula is only an abbre-
(mFeO, C02
viation of < nMn O, C 02 in which m, n, and o are alternating

UMg O, C 02
quantities of weights. The above-mentioned spathic iron

ISOMORPHISM. 125

has been formed from a fluid, in which the three isomorphous
salts, carbonate of protoxide of iron, carbonate of protoxide
of manganese, and carbonate of magnesia, are contained, and
crystals are thus formed, in which alternating and various
quantities of the individual salts enter into combination. If
we suppose 1000 atoms of the above-named carbonate in an
individual crystal of spathic iron, 700 at. may be carbonate
of protoxide of iron, 200 at. carbonate of protoxide of man-
ganese, and 100 at. carbonate of magnesia; but at the same
time every other kind of relation may occur. On this account,
therefore, as already mentioned, we shall most correctly define
isomorphous bodies to be such as, in consequence of a like
constitution, and on account of like equivalent volumes, possess
like crystalline forms, and may also crystallize in common in
the same form. If we assume the numbers to be large enough,
we shall be able to reckon how many equivalents of the one
isomorphous salt and how many of the other occur in a mixed
combination. If the relation be a simple one it is customary
to express it in the formula. The magnesian spar of Zillerthal,

fiCa ••
Jena, and some places, has the following formula { , . C

(aMg

or ^CaiMgC, which is intended to show that this com-
bination is not to be regarded as Ca C + Mg C, that is to
say, as a double salt of carbonate of lime and carbonate of
magnesia, but as a mixture of isomorphous salts in the relation
indicated. In the same manner, the magnesian spar of Halle
in the Tyrol, of Saberg, and many other places where it is

found, has the formula 4 Ca fMg C, containing 3 at. of car-
bonate of magnesia to 1 at. of carbonate of lime.

We shall now turn to the special consideration of the most
important and interesting isomorphous groups, in which an
opportunity will often occur of specially developing what has
already been given in general outline.

Titanic acid and binoxide of tin are isomorphous, for the
former crystallizes as rutile, the latter as tin-stone, in the

126 ISOMOKPHISM.

same forms of the Second System of crystallization*. The
binoxide of tin is Sn 02, titanic acid Ti 02, and, according
to Kopp, both combinations have the same equivalent volume,
(Annal. der Chem. und Pharm., Bd. 36, s. 5;) the isomorphism
is, therefore, founded here upon like number and grouping of
the atoms (Mitscherlich's law), and upon like size of the chemi-
cal atoms of the compounds (Kopp's law). As may be seen by
consulting the tables of the equivalent volumes (atom-volumes)
of the elements, titanium and tin have not the same equivalent
volume in a free condition, for that of titanium is 56, while
that of tin is 101; it is, therefore, impossible for 56 c. c. of
titanium to replace 101 c. c. of tin without a change of form
occurring, and hence tin and titanium cannot possibly be
isomorphous in an isolated condition, and hence it follows,
consequently, that isomorphism of the constituents is not to
be inferred, or, at any rate, not absolutely, from the isomorphism
of analogous combinations, — or that the isomorphism of com-
binations is either not at all, or not always, based upon the
isomorphism of the constituents. The isomorphism of binoxide

* The Systems of Crystallization will be referred to by their numbers, which,
with their synonymes, are as follows : —

1. The First System, or Regular System, characterized by three axes which
are similar and at right angles to each other. The regular system of Weiss.
The octahedral system of Miller.

2. The Second System, characterized by three perpendicular axes, but of
which two only are similar. The two-and-one-axis system (4-membered) of
Weiss. The pyramidal system of Miller.

3. The Third System, characterized by four axes, of which three similar
are placed in the same plane and cut each other at angles of 60°; the fourth
axis is different, and stands perpendicular to the plane of the other three. The
three-and-one-axis system (6-membered) of Weiss. The rhombohedral system
of Miller.

4. The Fourth System, characterized by three axes which are unlike, but
perpendicular to each other other. The one-and-one-axis system (2-and-2-
membered) of Weiss. The prismatic system of Miller.

5. The Fifth System, characterized by three unlike axes ; two of which
are oblique to each other, but the third is perpendicular to the plane of the
other two. The two-and-one-membered system of Weiss. The oblique pris-
matic system of Miller.

5. The Sixth System, characterized by three unlike axes which are all
oblique to each other. The one-and-one-membered system of Weiss. The
doubly oblique prismatic system of Miller. (See the Principles of Crystallography,
by V. Regnault.) — Ed.

ISOMORPHISM. 127

of tin and titanic acid can, therefore, only occur where, in one
or other of the two combinations, either the oxygen or the
metal is contained with some other than the original equivalent
volume.

According to the table (page 14), titanic acid contains oxygen
with the equivalent volume 32, whilst the binoxide of tin con-
tains oxygen with the equivalent volume 16, if we assume the
metal to be contained in it with its original equivalent volume.
The equivalent volume of titanium is expanded 11 times in
titanic acid, if, as in the case of binoxide of tin, we assume the
oxygen to be contained with the equivalent volume 16.

Alumina (corundum), sesquioxide of iron (ologistic iron),
and oxide of chromium, are isomorphous; they crystallize in
the Third System. The composition of these oxides is per-
fectly analogous, corresponding to the formula R2 03 ; their
equivalent volumes are of equal size (Kopp, page 5), or only
present such differences as may be explained on the want of
perfect identity of form. For instance, the same angle measures
85° 58' in ologistic iron, and 86° 6' in corundum. (See above,
and also further on.) Here, too, the isomorphism of the combi-
nations is not based upon the isomorphism of the constituents,
for the equivalent volume of iron is 45, while that of chromium
is 64, (page 4;) the isomorphism is therefore dependent upon
the oxygen entering into sesquioxide of iron with the equi-
valent volume 32, and into oxide of chromium with that of
16; the equivalent volume of sesquioxide of iron is, therefore,
2x45 + 3x32=186, that of oxide of chromium is 2x64 + 3
x 16 = 176. With ologistic iron, ilmenite (titanic iron) is
isomorphous. According to Mosander, the latter may be
considered as a titanate of the protoxide of iron (Fe O, Ti 0.2);
whence sesquioxide of iron and titanate of protoxide of iron,
although two combinations of totally different constitution,
but of like number of atoms, are isomorphous, and we might
conclude that isomorphism requires no similarity of constitu-
tion, but only equality in the number of the atoms. Ilme-
nite may, however, also be regarded as sesquioxide of iron,
in which 1 equivalent of iron is replaced by 1 equivalent of
titanium. If, then, we write the formula for sesquioxide of
iron as Fe + Fe + 30, and that for ilmenite as Fe + Ti + 30,

128 ISOMORPHISM.

the isomorphism is based upon similarity of constitution*. The
equivalent volume of ilmenite is likewise nearly equal to that
of sesquioxide of iron. In the same manner, however, as the
equivalent volume of chromium deviates from that of iron,
so also does that of titanium; for that of iron is 45, and that
of titanium 56, and consequently titanium and iron are as
little isomorphous in an isolated condition as chromium and
iron. The equivalent volume of sesquioxide of iron is, as stated
above, 186, that of the combination, Fe Ti 03, amounts, in like
manner, to 45+56 + 96 = 197, consequently so nearly equal to
that of sesquioxide of iron that isomorphism may occur.

Kopp is of opinion that bodies which are not themselves
isomorphous with others, may together yield isomorphous com-
binations, without any change occurring in their equivalent
volume; that is, when they form the preponderating quantity,
and where the deviation of the equivalent volume does not
exercise any disturbing influence upon the isomorphism.

Oxide of antimony, Sb 03, and arsenious acid, As 03, which,
are dimorphous, being both capable of crystallising in the First
and Fourth Systems, are isomorphous in both forms. The
isomorphism of these two combinations has long remained un-
detected, owing to their dimorphism, notwithstanding that it
was conjectured from the similarity of their composition and
of their chemical relation. Previous to their dimorphism being
discovered, there was only the regular form known for arsenious
acid, while, in the case of the oxide of antimony, the form of
the Fourth system was known, contrary to what was the case
with the arsenious acid, and thus it was naturally believed that
arsenious acid crystallized in the First System, and that oxide
of antimony, which is so similar to the former in chemical
relations, crystallized quite differently, namely, in the Fourth
System. Their dimorphism has led to the discovery of their
isomorphism. The equivalent volume of the two combinations
is equally large, that of the oxide of antimony, is, for instance,
^63 =344, that of arsenious acid 1%377'5 =334; the equi-

* A sesquioxide of titanium, Ti203, has lately been formed, which esta-
blishes a different conclusion, that ilmenite is a mixture of two isomorphous
oxides, Ti.,0.. and Fe, O.,.— Ed.

ISOMORPHISM. 129

valent volume of antimony is, however, 240, that of arsenic
is 160, so that the two metals cannot be isomorphous in an
isolated condition, provided dimorphism, and different allotropic
conditions with which we are as yet unacquainted, do not
occur.

It is worthy of remark that the equivalent volume of
antimony is exactly 11 times that of arsenic.

If, therefore, we suppose the antimony to be contained
in the oxide of antimony, and the oxygen, with the respective
equivalent volumes of 240 and 32 without condensation or
expansion (page 14), we may assume that the arsenic is
contained in the arsenious acid with I! times its original

equivalent volume.

As arsenious acid and the oxide of antimony, so the cor-
responding sulphur compounds of arsenic and antimony,
orpiment, As S3, and common antimony ore, Sb S3, are like-
wise isomorphous, as both occur in forms of the Fourth System.
The equivalent volume of both is equally great. The combi-
nations of both sulphides with sulphide of silver, the light
red silver ore, 3AgS+AsS3, and the dark red silver ore,
3AgS + SbS3, crystallize in the same rhombohedra; in like
manner, arsenic-fahlerze and antimony-fahlerze are isomor-
phous, and have the same equivalent volume.

Arsenide of nickel (Kupfer-nickel), Ni2 As, and antimo-
nide of nickel, Ni2 Sb, two combinations in which arsenic
and antimony are to be regarded as negative constituents,
are likewise isomorphous in forms of the Third System.
Sulphide of nickel (Haarkies), Ni S, or Ni2 S2, has the same
form as these, a combination of different chemical constitu-
tion, and consisting of 2 equivalents of nickel and 2 equiv.
of sulphur. Whilst arsenide and antimonide of nickel con-
tain 1 equiv. of arsenic and antimony to 2 equiv. of nickel.
We shall become acquainted with other examples, in which
similarity of form occurs with great difference of the gene-
rally assumed chemical constitution. Here we must either
recognise an isomorphism without reference to similarity
of form, or we must grant that the law of Mitscherlich is
not generally applicable, or we must finally endeavour to
harmonise the constitution with this law. If we hold to the

130 ISOMORPHISM.

equivalents of arsenic, antimony, and sulphur, and speak of
isomorphism between arsenide, antimonide, and sulphide of
nickel, we must assume that 2 equiv. of sulphur in the sulphide
of nickel are isomorphous with 1 equiv. of antimony and
arsenic in the antimonide of nickel, and arsenide of nickel.
If, however, we divide the equivalent of arsenic and antimony
into two atoms, as the specific heat of these metals demands,
and as the volume-theory requires according to Berzelius, we
obtain for arsenide of nickel and antimonide of nickel the
formuke Ni As and Ni Sb, which perfectly correspond with
the formula for sulphide of nickel, Ni S. A division of the
equivalents of arsenic and antimony into two atoms may
indeed be based upon the isomorphism of arsenide of nickel,
antimonide of nickel, and sulphide of nickel.

If we compute the equivalent volumes of the three com-
binations according to this last formula, we find that arsenide
of nickel and sulphide of nickel have nearly the same equi-
valent volume, but that the equivalent volume of arsenide
of nickel is nearly 1-L times greater. Kopp's law, conse-
quently, does not allow of the isomorphism of the three com-
binations.

Iron pyrites, Fe S2, and arsenical or speiss-cobalt, Co As,
crystallize in the same hemiedric forms of the First System,
and what has already been said applies also to these, for iron and
cobalt have the same equivalent volume, and are, therefore,
certainly isomorphous. We must, consequently, either assume
that S2 is isomorphous with As, or change the formula Co As
into Co As2, in order to establish the similarity of constitution.
Here, too, isomorphism cannot be explained by similarity of
the equivalent volumes, for the equivalent volume of cobalt
blende is almost exactly 1 -j times greater than that of iron
pyrites. A similar form to that of the above-named com-
pounds is presented by cobalt blende, Co S2 + Co As2, nickel
blende, Ni S2 + Ni As2, and even by nickel-antimony blende,
Ni S2 + Ni Sb2. Cobalt blende and nickel blende have the
same equivalent volume ; the equivalent volume of the nickel-
antimony blende is, however, only about |- as large, namely,
the same as that of speiss-cobalt. This is, as has been stated,
Co As2; if, therefore, we write the formula for nickel-antimony

ISOMORPHISM. 131

blende thus, Ni^S2-|Sb2, both have a like form, a like equi-
valent volume, and a like constitution. As cobalt blende,
CoS2+CoAs2, is isomorphous with sulphide of iron, Fe S2,
occurring as pyrites, we might assume that arsenical pyrites,
FeS2 + FeAs2, would be isomorphous with cobalt blende.
Such, however, is not the case, but arsenical pyrites is isomor-
phous with that second form of the bisulphide of iron, with
white pyrites (Sperkies,) having an equivalent volume approxi-
mating to that of the latter. (See Rammelsberg, Lehrbuch der
Stochiometrie, page 241.)

Sulphide of lead, Pb S, and selenide of lead, Pb Se, are
isomorphous, crystallizing in forms of the First System.
With the same constitution they have also the same equivalent
volume. It is probable that sulphur and selenium have also
the same equivalent volume in an isolated condition, for as
selenium can only with great difficulty be obtained free from
porosities, the specific gravity is probably found somewhat
too low. The same, perhaps, holds good for tellurium, which,
in a chemical point of view, stands together with sulphur and
selenium.

Many other metallic sulphides, composed according to the
formula R S, appear in forms of the regular system, as, for in-
stance, sulphide of zinc, Zn S, zinc blende and sulphide of silver,
Ag S. Does isomorphism occur here amongst themselves and
with sulphide of lead? The form is the same, and the constitut-
ion we also assume to be the same, why, therefore, should we
not call it isomorphism? The equivalent volume of the sulphide
of silver is nearly the same as that of sulphide of lead, that
of sulphide of zinc differs considerably from both, but this
difference of the equivalent volume does not exercise the same
influence upon crystals appertaining to the regular systems,
as to those belonging to the other systems. On the other
hand, we must observe that lead, silver, and zinc are not
approximated by any other cases of isomorphism, and that even
if silver and lead show some resemblance in a chemical cha-
racter, zinc differs very much from them in this respect.
We have already mentioned that the substitution in different
crystals for each other is a main characteristic of isomorphous

K 2

132 ISOMORPHISM.

bodies, but I am not aware that sulphide of zinc and sulphide
of lead, or sulphide of zinc and sulphide of silver, can enter
indifferently into the same crystal. It is likewise uncertain
whether the silver obtained from galena exists as sulphide of
silver in the mass of its crystals, or whether it originates from
other commingled combinations containing sulphide of silver.
We must also observe here, that the occurrence of forms
appertaining to the regular system is not sufficient to esta-
blish isomorphism, since bodies, which are evidently of [different
composition, and of which there can be no idea of isomor-
phism— if we do not adhere to the mere verbal meaning of
the term — occur in such forms; chloride of sodium crystal-
lizes, like galena, in cubes ; diamond, magnetic iron, and alum
crystallize in octahedrons, phosphorus and garnet in the garnet
dodekahedron.

Even in other systems, although more rarely, we meet with
the same form combined with a very decidedly different con-
stitution, so that even here, with identity of form, we must
not speak of isomorphism without some limitation.

The natural disulphide of copper, Cu2 S, (copper blende)
crystallizes in a form of the Fourth System. On its being
fused, or on fusing large masses of copper and sulphur, we
obtain octahedral, consequently regular crystals of disulphide
of copper. (G. Rose, Mitscherlich.) Disulphide of copper,
Cu2 S, is consequently dimorphous. As the natural sulphide
of silver, Ag S, likewise occurs in forms of the regular sys-
tem, and as, according to Becquerel, artificial sulphide of
silver is also obtained in octahedrons, we may say that Cu2 S
and Ag S are isomorphous. The circumstance that silver and
copper blende combined, Cu2 S + Ag S, occurs in the form
of copper blende, is in favour of the isomorphism of both sul-
phides.

It is, however, most certainly proved, according to H. Rose,
that Cu2 S and Ag S replace each other in alternating rela-
tions in Polybasite. We must now either admit that under
some circumstances Cu2 is isomorphous with Ag in combi-
nation, or we must divide the equivalent of silver into 2 atoms,
as is required by the specific heat of the silver. If we carry
out this division, copper blende and silver blende will have a

ISOMORPHISM. 133

like atomic constitution, the former is Cu2 S, the latter Ag2 S.
The equivalent volumes of both combinations differ very much
in other respects from each other. We shall subsequently see
what are the results from the division of the equivalent of
silver into 2 atoms.

Isomorphism may be best observed in salts, owing to their
peculiar tendency to form crystals. We shall very much faci-
litate our view of combinations in these classes if we place
together those cases of isomorphism in which the halogen (salt-
radical), or the acid is different, and then those in which the
metal or the base are different.

The salts of tungstic acid are isomorphous with the analo-
gous molybdates. Thus, molybdate of lead (yellow lead earth)
and tungstate of lead occur in similar forms of the Second
System. The constitution of these salts is the same, for
instance, the formula of tungstic acid is Wo 03, that of
molybdic acid Mo 03.

Sulphates, seleniates, chromates, and manganates of the
same base, form well-marked groups of isomorphous salts.
The analogous salts of these acids, which all contain 3
atoms of oxygen to 1 atom of the positive constituent, are
isomorphous, when free from water, or when they contain
an equal number of atoms of water of crystallization. Thus,
neutral sulphate of potash and chromate of potash, which are
both anhydrous, crystallize in the same form of the Fourth
System.

The same applies to anhydrous and hydrated sulphate and
seleniate of soda, and to anhydrous sulphate and seleniate of
silver. The acid sulphate of potash (K O, S 03 + H O, S 03)
has the same form as the corresponding seleniate. Isomorphism
extends also to the double salts; thus, for instance, the ammo-
niacal sulphate, seleniate, and chromate of silver, have the
same form. (Ag O, S 03 + 2 N H3, &c.)

The equivalent volume of sulphate of potash is ^qq =409,
the equivalent volume of chromate of potash is 122.177 =451;
the two salts have therefore only an approximatively equal
equivalent volume. As the angular relations of their crystals
are not exactly equal but merely approximatively so, the
salts, strictly speaking, are not isomorphous, but homoemor-

134

ISOMORPHISM.

phous.

I'io*. 1 exhibits a prism frequently met with in these
salts. According to Mitscherlich {Pogg.
Annal, Bd. 18, s. 168), the inclination of
two of its faces is as follows :

Fig. 1.

In sulphate of potash
„ chroraate ,,

g : 2-g

112° 22'
1110 10'

so that the greater equivalent volume cor-
responds to the smaller angle of inclination.
In seleniate of potash this angle amounts
to 111° 48' 5", in sulphate of ammonia to
111° 15'.

Johnston has observed that chromate of lead is dimorphous,
and isomorphous in its least frequent forms with molybdate
of lead. This establishes an affinity between molybdic, chro-
mic, sulphuric, and other analogous acids.

The salts of phosphoric and arsenic acids constitute another
well defined isomorphous group. For every arseniate there
is a phosphate, corresponding, with respect to its composition,
with the former, and having the same form. Thus, ordinary
arseniate and phosphate of soda exhibit a like crystalline form,
and have also a similar composition (Graham's Chemistry);
each of the salts contains 2 equivalents of soda and 1 equiv.
of water as bases, with 1 equiv. of acid and 24 equiv. of water
of crystallization. With a different quantity of water of crys-
tallization, namely, with 14 equiv., the form of this salt is
quite different, but it is again the same in both. The so-called
acid salts of these acids, which, for 1 equiv. of soda and
2 equiv. of water as bases, contain 1 eq. of acid and 2 of
water of crystallization, have the same form. Phosphate and
arseniate of lead, which are likewise isomorphous, replace each
other in alternating and undetermined relations in the native
arseniates and phosphates of lead.

We do not know whether phosphoric and arsenic acids are
themselves isomorphous; both are known as anhydrous in an
amorphous condition only. Phosphorus and arsenic, however, are
not isomorphous; the former crystallizes regularly, the latter
according to the Third System, and the equivalent volumes
are different. It is possible, however, that the isomorphism

ISOMORPHISM. 135

here is concealed by dimorphism, and that the phosphorus
possesses a different form in different allotropic conditions.
Chlorine, bromine, and iodine have the same equivalent
volume, and their analogous combinations are isomorphous.
Thus the chlorides are isomorphous with the iodides and bro-
mides, the chlorides with the iodates and bromates, and the
perchlorates with the periodates. All these combinations cor-
respond exactly with respect to their composition. Iodic acid
and bromic acid contain, like chloric acid, 5 equiv. of oxygen to
1 equiv. of the halogen ; periodic acid contains, like perchloric
acid, 7 equiv. of oxygen to 1 equiv. of halogen. Cyanogen
and fluorine appear also to belong to this group ; cyanogen has
the same equivalent volume as chlorine, &c. (p. 4.) The
cyanides and fluorides have the same form as the chlorides, and
in the case of chloride and fluoride of calcium replace each

other in alternating relations : ~|a ~1 + 3 Ca O, P 05. The

Ca C1J

permanganates are isomorphous with the perchlorates; thus
permanganate of potash and perchlorate of potash may be sub-
stituted for each other in crystals like other isomorphous bodiee.
Perchloric acid, however, contains 1 equiv. of chlorine to 7
equiv. of oxygen, whilst 2 equiv. of manganese occur together
with 7 equiv. of oxygen in permanganic acid, and we must
therefore assume, that 2 equiv. of manganese can replace 1
equiv. of chlorine, without producing any change of form, or,
as we are accustomed to say, that 2 equiv. of manganese in the
permanganates are isomorphous with 1 equiv. of chlorine in the
perchlorates. The isomorphism of the perchlorate and perman-
ganate of potash is regarded by Berzelius as the most decided
reason for the division into two parts of the equivalent of chlo-
rine, and consequently also of bromine, iodine, &c, a division
which, as we know, is also required by the volume theory, and
which corresponds best with the specific heat of chlorine. Ber-
zelius expresses himself decidedly, that it is undoubted that
permanganic acid consists of 2 atoms of manganese and 7 atoms
of oxygen: as now perchlorate of potash is isomorphous with
permanganate of potash, it follows that perchloric acid must
consist of 2 atoms of chlorine and 7 atoms of oxygen analogous
to permanganic acid. In the same manner hypochlorous acid

136 ISOMORPHISM.

consists of 2 atoms of the radical and 1 atom of oxygen. Hence
it further follows, continues Berzelius, that if hypochlorous acid
be formed of 2 vols, of chlorine gas and 1 vol. of oxygen gas,
these two gases must, with the same volume, contain an equal
number of atoms; while he further derives the division of the
hydrogen equivalents into 2 atoms from the substitution of
hydrogen by means of chlorine, volume for volume, in organic
combinations.

If we divide the equivalent of chlorine into 2 atoms, we
can explain the isomorphism existing between perchlorate of
potash and permanganate of potash, by the similarity of the che-
mical constitution of the salts, that is, by their consisting of an
equal number of similarly grouped atoms, corresponding to the
law established by Mitscherlich (KO, Mn2 07 and KO, VC12 07.)
We have, however, already seen many cases, and will have occa-
sion to observe many more, to which this law cannot be applied
in its full force. I fully admit the value which isomorphism
has for the determination of constitution within certain groups
of combinations, but if we are to admit that 1 atom of silver
in sulphide of silver (Ag S) has the same value in an isomo-
phous relation as 2 atoms of copper in disulphide of copper
(Cu2 S), when we see that 1 atom of magnesia, protoxide of
iron, &c, may be replaced without any change of form by 3
atoms of water, (see further on,) and that ammonium, a group
of 5 equivalents (H4 N), may take the place of 1 equivalent of
potassium, it is not difficult to assume, that 2 atoms of manga-
nese and 1 atom of chlorine may mutually replace each other in
the above-named combination without any change in form.
Berzelius, in the last edition of his Manual, has set down the
atomic weight of ammonium as equal to its equivalent weight
(VH8 VN2), whilst previously the equivalent of ammonium was
made to consist of 2 atoms (VH4 VN). It would certainly be
very striking to halve the equivalent of chlorine on account of
its isomorphous relations, and not carry out this division in the
equivalent of potassium, which is as urgently demanded, if 1
atom of ammonium be set down as = VH4 VN. It may further
be urged against the division of the equivalent of chlorine into
2 atoms, that dichloride of copper (Cu2 CI) is isomorphous
with dinoxide of copper (Cu2 O), but as both combinations

ISOMORPHISM.

137

occur in forms appertaining to the regular system, we lay little
stress upon this isomorphism.

Isomorphous groups of salts, whose acids are the same, but
their bases different, are still more frequent than the above-
enumerated isomorphous groups of salts, with like base and dif-
ferent acids. A very remarkable group of this kind is formed
by those carbonates which are comprised in mineralogy under
the name of spars: —

Iron-spar....

Talc-spar..
Zinc-spar

Fe O, C 02
Mg O, C 02
Zn O, C 02

Fig. 2.

Calc-spar Ca O, Co2

Somite tca^O ^C°2

Manganese-spar MnO, Co2

The fundamental form of all these
salts is a rhomboid, (fig. 2,) and their
bases replace each other in the most
various relations in the crystals. Thus
spathic iron almost always contains
carbonate of protoxide of manganese
and carbonate of magnesia; manganese
spar contains carbonate of the protoxide
of iron, carbonate of magnesia, and car-
bonate of lime ; the zinc spar, carbonate of the protoxide of iron,
&c. As, however, a perfectly exact agreement of form does
not occur in sulphate, chromate, and seleniate of potash, so
there are likewise differences that deserve notice, with regard to
the angular relations of the principal rhomboid in these different
salts; the angle of inclination at the extreme edges of this
rhomboid varying from 105° 5' to 107° 40'. The following
table shows that this difference of form stands in the closest
connection with the difference of the equivalent volume: —

Formula.
CaC

Equiv. Spec. Grav. Equiv.Vol.
625 272 230 105° 5'

**.fg..
MnC

FeC

MgC
ZnC

579

2-88

201

106° 15'

720

359

200

106° 51'

625

3-83

189

107° 0'

533

300

178

107° 25'

781

4-44

176

107° 40'

138 ISOMORPHISM.

We see that the size of the equivalent volume corresponds
to the smallness of the solid angle. Kopp has given a formula
for this relation of the equivalent volumes to the solid angle,
(Annah der Chemie und Pharm.9 Bd. 36, s. 15,) and Rammels-
berg draws attention to the fact, that in dolomite, which may-
be looked upon as a combination of -$ an equivalent of carbo-
nate of lime, (calc-spar,) and 2" an equivalent of carbonate of
magnesia (talc-spar), and its equivalent volume, therefore,
the arithmetical mean of the equivalent volumes of these
two constituents, the size of the angle in question is also the
exact arithmetical mean of their size in calc-spar and talc-spar,
namely, 10515^107125 '=106o15.

If the diminution of the angle of the terminal edges of the
rhombohedron be really dependent upon the augmentation of
the equivalent volume, this angle will not only become smaller,
if one constituent is replaced in the crystals by another of
larger equivalent volume, as, for instance, carbonate of mag-
nesia by carbonate of lime, but it will also become smaller if
an increase of the equivalent volume be produced by any
other means. This may happen by the crystals being heated,
since by the action of heat the density is diminished, and
the equivalent volume is increased. Mitscherlich has long
since shown that heat exercised the influence in question
upon the angle of this crystal. Crystals belonging to the
regular system are equally expanded in all dimensions, the an-
gular relations remain, therefore, the same, notwithstanding
that the equivalent volume is increased in size. The same
must, consequently, also be the case when, in crystals of this
kind, constituents which are not of absolutely equal equivalent
volume replace each other.

The salts of baryta, strontia, and lead of the same acids
form well-marked isomorphous groups. Sulphate of baryta
(heavy spar), sulphate of strontia (celestine), sulphate of lead,
crystallize in the same form of the Fourth System. Car-
bonate of baryta (witherite), carbonate of strontia (stron-
tianite), and carbonate of lead (white lead ore), have likewise
the same crystalline form, belonging also to the Fourth
System, but differing in each. The anhydrous nitrates of

ISOMORPHISM. 139

these bases, and the bromates with 1 equiv. of water, are like-
wise isomorphous.

The group of the carbonates of these bases is allied, in an in-
teresting manner, with the preceding group, namely by dimor-
phism of some of the members. Carbonate of lime, as arra-
gonite, has, for instance, exactly the same form as the carbo-
nates of baryta,- strontia, and lead, and is, consequently,
isomorphous with them ; on this account we often find a portion
of the lime replaced in arragonite by strontia, or even by
oxide of lead. On the other hand, it appears that carbonate of
lead can occur in the other form of carbonate of lime, as calc-
spar, and consequently that it may also be dimorphous, for we
find in plumbocalcite, which has the form of calc-spar, carbo-
nate of lead in alternating relations with carbonate of lime.
We should never, therefore, have suspected the isomorphism
of these two carbonates, had we only learnt to know them in
their ordinary forms. This is a highly important fact, for the
deficiency of isomorphism between other bodies may also be
occasioned by their being dimorphous, when the two forms
have not yet been fully ascertained. According to Dufrenoy,
the form possessed by carbonate of protoxide of iron as jun-
kerite is also that of arragonite; this salt is therefore likewise
dimorphous.

Lime is connected with oxide of lead, baryta, and strontia,
by the salts of other acids. Thus, apatite, Ca C1 + 3Ca O, P 05,
and native phosphate of lead, PbCl + 3PbO, P 05, have the
same crystalline form ; and in the latter, a portion of the
phosphate of lead is often replaced by phosphate of lime, as
also occasionally a part of the chloride of lead by fluoride of
calcium. It has already been mentioned that arsenic acid
occurs in the place of phosphoric acid. The hyposulphates
of strontia, oxide of lead and lime, crystallize with 4 equiv.
of water, according to the Third System, and are likewise
isomorphous among themselves. (Heeren.)

What has been said of the relation existing between form
and equivalent volume in the groups of spars, also applies to
the groups of salts of baryta, strontia, and lead, and to those
salts of lime which are thus isomorphous by dimorphism. The
simplest characteristic form, to which the other forms may

140

ISOMORPHISM.

Rfil

Angles.

118° 30'

61° 30'

106° 50'

117° 16'

62° 44'

108° 12'

117° 14'

62° 46'

108° 13'

116° 16'

63° 44'

108° 27'

be referred, is in the case of the carbonates
of baryta, strontia, oxide of lead and lime, such
as we have represented in fig. 3.

The angles of the crystals are not, however,
exactly the same in these four salts; the solid
angles a and b, and the angle of the ends, c c,
are different in these, and the equivalent volumes
are also different, as the following table will
show.

Formula. Equiv. Spec. Grav. Equiv. Vol.

BaC 1230 4 30 286

SrC 921 360 256

PbC 1669 6-47 258

CaC 625 293 213

We here clearly perceive how the equivalent volumes most
accurately agree in the salts which are most perfectly isomor-
phous. The differences in the angles and in the equivalent
volumes are really so small in carbonate of strontian (stron-
tianite) and carbonate of lead (white lead ore) that they may
be esteemed mere errors of observation. Carbonate of baryta
(witherite) and carbonate of lime, as arragonite, differ con-
siderably from these two salts with respect to their equiva-
lent volumes and angular relations; and an increase of the
equivalent volume corresponds to an increased angle a, and
smaller angles b and c. If we increase the equivalent volume
of one of these salts, (if we heat it,) a change in the angle
corresponding to this increase of the equivalent volume occurs,
the angle a becomes more obtuse, and the angles b and c become
less so. (Kopp.)

In sulphate of lead the angle a amounts to 101° 32', in
sulphate of baryta (heavy spar) to 101° 42', and in sulphate
of strontian (celestine) 104° 48'; here, therefore, there is no
perfect isomorphism.

Interesting results present themselves when we seek to refer
the isomorphism of the salts of baryta, strontia, and lead, to
the isomorphism of the oxides or of the metals. If we allow
the specific gravities 4*73 and 3*93, as given by Karsten, to

ISOMORPHISM. 141

stand for baryta and strontia, these oxides will not have the
same equivalent volume, for that of baryta amounts to 202,
and that of strontia to 164; they cannot, therefore, replace
each other in salts without a change of form, as like volume is
necessary. But the metals of these oxides, barium and
strontium, have apparently, according to Kopp, the same
equivalent volume, and we might hence conclude, that in the
salts of baryta and strontia, those earths and the acids are not
proximate constituents, but, as the binary theory of the salts
demands, barium and strontium with a compound radical; and
consequently it would be more correct to write Ba S 04 than
Ba O, S 03. The conversion of the one salt into the other is not
then effected by an exchange of baryta and strontia, but of
barium and strontium. If we include lead in this considera-
tion, we find that its equivalent volume does not agree with
those of barium and strontium, whilst the equivalent volume
of oxide of lead approaches very nearly to that of strontia,
being 156. With respect to salts of lead, we must therefore
give the preference to the generally prevalent view of the
constitution of salts, whilst we assume two different constitu-
tions for the salts of strontia. (Kopp, Annalen der Chemie u.
Phar. Bd. 36, s. 18.) Similar results are yielded with respect to
the bases and metals of other salts, and it likewise is worthy
of notice, that although chloride of barium and chloride of
strontium have the same equivalent volume, that of chloride
of lead differs from both.

Potash (K O) and oxide of ammonium (Am O, that is,
H4 N O), form with the same acids, salts which are isomor-
phous; the same is the case with potassium and ammonium
(Am, that is, H4 N or VH8VN2), with salt-radicals. Thus the
sulphates, chromates, nitrates, &c, of potash and oxide of
ammonium have the same form, and in many double salts a
substitution of the oxide of potassium for oxide of ammonium
occurs without any change of form. (See below.) The
isomorphism of the potassium and ammonium compounds
shows in the clearest manner, that similarity of form does not
invariably indicate a like molecular constitution of bodies, for a
group of 5 at. of hydrogen, and 1 at. nitrogen (10 at. vol.) has in
these compounds the same crystallographic value as a single

142 ISOMORPHISM.

atom of potassium. It is very probable, that salts of soda and
sodium belong to this group, but that the isomorphism is con-
cealed by dimorphism, except in a few cases. Thus we find
soda in soda-alum, in the place of the potash in potash-alum;
the crystalline form of both alums is, like their composition,
the same, and both crystallize in octahedra. Frankenheim has
shown that microscopic crystals of nitrate of potash possess the
form of nitrate of soda. It is also conjectured that soda may
take the place of potash and lime in the mineral chabasite
without any change of form. Anhydrous sulphate of soda,
which cannot yet be obtained in the form of sulphate of potash,
is exactly isomorphous with sulphate of silver, and the two
salts have nearly the same equivalent volume, which differs
from the equivalent volume of sulphate of potash and oxide of
ammonium. Thus, throughout, the corresponding potassium
and ammonium compounds exhibit a great correspondence in
their equivalent volume as demanded by isomorphism, but the
corresponding sodium compounds are far from being included
under the same head, their equivalent volume always approach-
ing nearly to that of the corresponding silver compound. The
difference of the equivalent volumes of the analogous but non-
isomorphous potassium and sodium compounds is so great —
the equivalent volumes standing for the most part nearly in the
relation of 4 : 3 — that in the case of dimorphism establishing
the isomorphism, very considerable changes must take place in
the density in order that the equivalent volumes may become
equal. (See Kopp.)

The isomorphism of sulphate of silver and sulphate of soda
is corroborative of the present atomic weight of silver, whilst,
as we have already seen, the isomorphism between Cu2 S and
Ag S may indicate the propriety of a division of the atom of
silver. If, owing to the isomorphism of the latter, we would
carry out the division of the silver equivalent into two atoms,
in which the oxide of silver would be changed from Ag O
to Ag2 O, and the dinoxide of silver from Ag2 O to Ag4 O,
the division would lead to the halving of the equivalents of
the sodium, owing to the isomorphism of the sulphate of
soda with sulphate of silver. And these salts would then
have the formulae, Ag2 O, S 03, and Na2 O, S 03, and since

ISOMORPHISM. 143

there probably exists an isomorphism of sodium compounds
with the potassium and ammonium compounds, the equivalents
of the potassium and ammonium would necessarily be halved.
At the time Berzelius gave the formula VH4 VN to ammonium,
and oxide of ammonium contained, according to him, 2 at. of
ammonium and 1 at. of oxygen, a division of the equivalent
of potassium might also have been based upon this. Now,
Berzelius has certainly more correctly made the atomic weight
of ammonium correspond with the atomic weight of potassium,
and consequently 1 at. ammonium is, according to him, — 1 eq.
of ammonium, namely, VH8 VN2.

It has already been stated that similarity of form is not
invariably connected with similarity of composition, but that
perfect correspondence in form, that is, actual isomorphism, may
also take place between bodies, which exhibit a complete dif-
ference with respect to their constitution. Carbon, chloride of
sodium, arsenious acid, and alum may all crystallize in octa-
hedra, although they have not even a remote resemblance in
composition.

Even if we do not admit that the ordinary occurrence of
the simple form of the cube and the regular octahedron,
proves the existence of isomorphism, except in cases where we
might expect to meet with it, in consequence of well-grounded
similarity of composition, as in the different kinds of alum, the
isomorphous groups we have been considering, present us with
examples of identity of form, without indicating any well-
grounded resemblance in the composition, against which the
objection in question cannot be urged. Thus the dimorphous
bisulphate of potash, K O, S 03 + H O, S 03, is in one form
isomorphous with sulphur of the Fifth System, and in the
other form with felspar. Nitrate of soda, Na O, N 05, has
the form of calc-spar, Ca O, C 02. The crystals of nitrate of
potash, K O, N 05, known under the name of saltpetre, have
the form of arragonite, consequently that of the other form of
carbonate of lime, and microscopic crystals of this salt, as
already observed, occur in the form of nitrate of soda, conse-
quently in that of calc-spar. Finally, anhydrous sulphate of
soda, (and oxide of silver, see above,) Na O, S 03, are not
isomorphous with manganate of baryta, Ba O, Mn 03, but

144 ISOMOKPHISM.

with permanganate of baryta, Bn O, Mn2 07. In the two
first-named bodies the bisulphate of potash and sulphur, the
absence of all analogy is obvious. There is not either any
resemblance in the composition of nitrate of potash and carbo-
nate of lime, or between permanganate of baryta and sulphate of
soda, if we abide by the common views regarding the constitu-
tion of these salts.

Attempts have been recently made, from various quarters,
to deduce the isomorphism of these salts from the law esta-
blished by Mitscherlich ; but these attempts must for the pre-
sent be regarded as mere speculations. Thus Schaffgotsch
recommends the division of the equivalents of silver, sodium,
and potassium into two atoms, and then avails himself of it to
explain the isomorphism of the nitrates of potash and soda with
carbonate of lime. If, for instance, we exhibit the composition
of nitrate of potash, in accordance with the binary theory of
the salt, by the formula K, VN2 06, it will be after the halving
the equivalent of the potassium, K2, VN2 06, or K, VN 03. As,
according to this theory, carbonate of lime has the formula
Ca, C 03, the analogy of the composition of nitrate of potash
and of carbonate of lime is made apparent, K+vN03 and
Ca + C 03. The same is the case with nitrate of soda. We per-
ceive by this, that the division of the equivalent of the nitrogen
into 2 at., as required by isomorphism, may be supported.

The dimorphism of carbonate of lime and nitrate of potash
may possibly, according to Schaffgotsch, be grounded upon the
fact, that calc-spar and the nitrate of potash crystallized in this
form may be composed according to the formulas Ca2 + C2 06
and K2-fvN2 06, while arragonite and common nitrate of pot-
ash have the formulae Ca+C 03 and K+VN 03. Dimorphism
would, therefore, be a consequence of polymerism.

In opposition to this Clark proposes doubling the atomic
weights of sodium and silver, in order to explain the isomor-
phism between sulphate of soda, sulphate of silver, and per-
manganate of baryta. If we carry out this doubling, we have
as follows : —

Sulphate of soda Na + 2S -j- 80

Sulphate of silver Ag + 2S +80

Permanganate of baryta Ba + 2Mn + 80

-\d

ISOMORPHISM. 145

or according to the binary theory of salts,

Sulphate of soda Na + 2S 04

Sulphate of silver Ag +2S04

Permanganate of baryta Ba +2Mn04

or finally,

Sulphate of soda Na02 + 2S03

Sulphate of silver Ag02 + 2S03

Permanganate of baryta Ba02 + 2Mn03

Isomorphism is, consequently, deduced from a like number,
and a like arrangement of the atoms.

"We will now after these remarks turn to the consideration
of new groups of isomorphous salts. The ordinary green vitriol
is a sulphate of the protoxide of iron with 7 equiv. of water,
Fe O, SO3 + 7 HO; with the same quantity of water, the
sulphate of the protoxide of cobalt crystallizes at an ordinary
temperature, while sulphate of protoxide of manganese crystal-
lizes below 6° C. The crystalline form of these three salts is
the same, of the Fifth System : they are isomorphous. Sulphate
of magnesia and sulphate of zinc likewise crystallize at the
ordinary temperature with 7 equiv. of water, (the magnesian
salt, Mg O, S 03H-7 H 0, the zinc salt, Zn 0, S 03 + 7 H O,)
as also sulphate of nickel below 5° C. These three salts are
isomorphous with each other, and have an approximatively
equal equivalent volume, but they are not isomorphous with
the three first-named, as their crystals belong to the Fourth
System. The difference of form of these two groups of salts,
with similar constitution, may possibly be dependant upon
dimorphism; thus it may be, that under certain circumstances
the three last salts crystallize in the Fifth System or three first
in the Fourth System. It is also understood, that the nickel salt
separates from its solution above 30° C. with 7 equiv. of water,
in crystals of the Fifth System, whence it would appear, that
the salt is trimorphous. But from a solution above 30° C. the
manganese and zinc salts crystallize with 6 equiv. of water,
in the Fifth System, and of the same form. The sulphate of
magnesia crystallizes between 20° and 30°, with 4 equiv. of
water, and the sulphate of iron also at 80°, both in the Fourth
System, and isomorphous. (Regnault.) Sulphate of copper with
5 equiv. of water, Cu O, S 03 + 5 HO, crystallizes in forms of

146 ISOMORPHISM.

the Fifth System; the sulphate of manganese crystallizes with
the same quantity of water, and in the same form, under 20°.
It is very probable, if not certain, that the sulphates of all the
above-named bases, that is, of magnesia, the protoxide of iron,
manganese, cobalt, and nickel, zinc, and copper, with the same
quantity of water, are isomorphous; but that the isomorphism
is in some cases concealed by dimorphism, and in others it has
not been made known from want of acquaintance with the
circumstances in which the salts of these bases may exist with
particular proportions of water. Sulphate of iron crystallizes,
as has already been stated, at the usual temperature with
7 equiv. of water, in the Fifth System ; and sulphate of copper,
with 5 equivo of water, in the Sixth System. It is probable
that, under circumstances with which we are at present unac-
quainted, the former may separate with the proportion of water
and the form of the latter, and the latter may separate with
the water and the form of the first. This appears to be proved
by a mixed solution of the sulphate of iron and copper yield-
ing a mixed crystal with the water and form of the sulphate of
iron. The same holds good with reference to a solution of the
sulphate of iron mixed with the sulphates of cobalt and zinc. In
the process of smelting copper in the Mansfeld district in Ger-
many, a salt is obtained by washing the copper ore, with 7
equiv. of water, in large violet crystals, of the form of sulphate
of iron, in which the bases are oxides of copper, iron, cobalt,
manganese, zinc, and magnesia. (See Rammelsberg, Lehrbuch
der Stochiometrie^)

It is interesting and worthy of remark, that light blue
crystals are yielded by a mixed solution of sulphate of zinc
(ZnO, S 03 + 7 H O,) and sulphate of copper, (Cu O, S 03 +
5 H O,) which have not the form of either of these salts, but
that proper to sulphate of iron, (Fe O, S03 + 7 HO,) and

which contain like it 7 equiv. of water] ^>SOo + 7 HO.

[Zn OJ 6

If a crystal of sulphate of iron is placed in such a mixed solu-
tion, it is enlarged by the matter of the double salt, as if by its
own, exhibiting a bluish green nucleus, and a light blue ex-
ternal layer; and thus we may obtain it formed of alternating
coloured layers by applying alternately a pure solution of the

ISOMORPHISM. 147

sulphate of iron and of the mixed solution. Instead of sulphate
of zinc we may make use of sulphate of nickel or sulphate of
magnesia, with the same results. (Mitscherlich.)

The fact that sulphates of these bases, of like constitution,
are isomorphous, is confirmed by the circumstance, that the
double salts they form with sulphate of potash and oxide of
ammonium, are accurately isomorphous, and have precisely the
same composition. All contain 1 equiv. of the alkaline sul-
phate and 6 equiv. of water; for instance, —

MgO,S03

K 0, S 03

6H0

Ca 0, S 03

K 0, S 03

6H0

Zn 0, S 03

K 0, S 03

6H0

Ni 0, S 03

K 0, S 03

6H0

MgO,SOs

KO, S03

6H0

&c.

We observe, that the above-named bases include all those
which, combining with carbonic acid, yield the interesting isomor-
phous group of the spars, with the exception of lime. Here
the question naturally arises, is sulphate of lime connected by
isomorphism with the sulphates of these bases? The question
may be answered in the affirmative. In a sulphate of iron of
like form with hydrated sulphate of lime (gypsum), we have a
like constitutiou, (that is, Fe O, S 03 + 2 H O,) with the
sparing solubility of sulphate of lime.

As the bases, which form isomorphous salts with carbonic
acid, also behave in like manner with reference to sulphuric
acid, the conjecture necessarily forces itself upon us, that all
the bases in question, will yield analogous and isomorphous
salts, with all acids, a conjecture which is generally confirmed
by experience. Thus, hydrated chloride of calcium and chlo-
ride of magnesium have a similar constitution, and probably a
like form; so also, we have hydrated nitrates of magnesia, man-
ganese, cobalt, copper, and zinc, of similar composition; and
have reason to suppose, that magnesia, lime, protoxide of iron,
protoxide of manganese, frequently replace each other, in the
case of the silicates, as for instance in augites.

If we compare the isomorphism of the sulphates which
have just been considered, with reference to the equivalent
volume of the bases and metals which they contain, we shall
arrive at results similar to those that have already been ob-

148 ISOMORPHISM.

tained in the salts of baryta, strontia, and lead. The equiv.
vols, of Fe, Mn, Co, Ni, Cu, are equal, namely 44; an equiv.
of manganese may, therefore, directly replace 1 equiv. of iron,
without any change of form occurring, since each fills the
same space. So also with Co, Ni, and Cu. The equiv. vol. of
zinc is, however, 58, that of cadmium 81, that of magnesium
84, (spec, gravity = 1*87); a substitution of any one of these
metals for another, equivalent for equivalent, is, therefore, only
possible without any change of form occurring, provided con-
densation takes place; moreover, the equivalent volumes of the
basic oxides, R O of these metals, have not all like equivalent
volumes ; at least, not so far as the data of which we are now in
possession allow of our judging, but it is a remarkable fact,
that the equivalent vols, of Mg O and Cu O are almost iden-
tical, so that, consequently, although Cu and Mg cannot replace
each other, Cu O and Mg O may do so without any change of
form occurring. The equiv. vol. of oxide of zinc is larger
than that of the last two oxides, but anhydrous sulphate of
zinc and sulphate of magnesia, have nearly the same equiv. vol.
(Kopp, op. cit., p. 10.)

Graham is of opinion, that the water stands in an isomorphous
relation with the bases of these groups, equivalent for equivalent;
the isomorphism is certainly not proved, but the resemblance,
which in a chemical point of view, water exhibits towards these
bases, indicates that such is the case. Thus, according to that
chemist, nitrate of copper, hydrated nitric acid of spec. grav.
1*42 and sub-nitrate of copper, are expressed by the formulas

CuO, N05 + 3HO
HO, N05 + 3HO
CuO, N05 + 3(CuO, HO)

Crystallized oxalic acid, or oxalate of oxide of hydrogen, cor-
responds with oxalate of magnesia; hydrated sulphuric acid,
HO, S 03 + H O, corresponds with sulphate of magnesia,
Mg O, S 03 + H O, and this similarity, in a chemical point of
view, is regarded by Graham as a main objection to the division
of the equivalent of hydrogen into 2 atoms.

Very recently Scheerer has brought water into an isomor-
phous relation with the same bases in a totally different

ISOMOKPHISM.

149

manner; he shows that 3 atoms of H O, have frequently, in a
crystallographical and also a chemical relation, the same value
as 1 at. MgO, FeO, MnO, and probably also as CaO, NiO,
Zn O; and that in like manner, 2 at. of H O correspond to 1 atom
CuO. (PoggendorfFs Annalen, 1846, Bd. 68, s. 319, &c.)
The examination of two minerals, cordierite of Krageroe and
aspasiolite, led to the discovery of this species of isormorphism,
named by Scheerer, polymeric isomorphism. The analysis gave
the mean as follows: —

Silica

Alumina ....

Magnesia ....

Lime

Protoxide of Iron....

Protoxide of Manganese

Water

Cordierite. Aspasiolite.

50*44 50*40

32-95 32-38

12-76 8-01

1*12 trace

0-96 234

trace ... trace

102 6'73

99-25

99-86

If we consider the iron in cordierite as sesquioxide of iron, a
view favoured by its deficiency of colour, the relation of the
oxygen of the silica (Si 03) to the oxygen of the bases R2 03
and R O is as follows : —

Si03
26-20

R203
15-64

RO
5 26

which very nearly corresponds to the formula 3R O, 2 Si 03 +
3(R203, Si03); consequently, the relations of oxygen would
be 26-20, 1572, 5*24.

The green colour of aspasiolite indicates the presence of
protoxide of iron, which has been confirmed by quantitative
analysis. The relation of the oxygen is consequently thus
computed: —

Si03

2618

R,03
15-12

PvO

363

HO
5-98

All attempts to deduce a formula from this in the usual
manner, fail, and we obtain improbable expressions showing no
harmony with the formula of cordierite, to which, however,
aspasiolite stands in the closest relation. The two minerals
possess the same crystalline forms; rhombic prisms of 120°,
with combinations of O P' <x> P oo and oo Poo; the most com-

150 ISOMORPHISM.

plete transitions from the one mineral to the other present
themselves in the same specimen, and there are crystals partly
consisting of aspasiolite and partially of cordierite. The nucleus
especially is formed of the latter, and distributes itself, as may
be discerned with certainty, by its difference in hardness and
colour.

On comparing the composition of aspasiolite with the
cordierite of Krageroe, we find that the silica and alumina
stand nearly in the same relation in both minerals, and that it
is mainly owing to the diminished quantity of magnesia attri-
butable to the large quantity of water, that the mixture of the
former differs from that of the latter. It naturally, therefore,
occurs to the mind, that both minerals, cordierite and aspasiolite,
are isomorphous, and that this isomorphism is owing to a certain
quantity of water being able to replace a certain quantity
of magnesia. We find by computation, that in such a case,
three equivalents of water replace one equivalent of magnesia.
If, in this relation, we were to substitute magnesia* for the
water contained in the aspasiolite, we should obtain the follow-
ing relation of oxygen for this mineral:

Si03 R203 RO

26-18 15-12 5-63

while that found for the cordierite of Krageroe, is 26*10 : 15-64
: 5*26, and these numbers approximate much more if we reckon
a part of the iron of the aspasiolite as sesquioxide, for in-
stance, about the half, as

Si 03 R2 03 RO

26-18 15-52 537

It consequently follows, that the similarity in the crystalline

* According to the equivalency of 3 at. of H 0 to 1 equiv. of Mg O, and con-
sequently of 3 X 112-55 (= 3375) to 258*35, there is for every part by weight
of water aag88?'fg5 = 0*766 parts by weight of magnesia. For the 6*73 per cent,
therefore, of water contained in the aspasiolite, we must take into account
6 73 X 0-766 = 5-15 per cent, of magnesia. For the computations of the rela-
tions of the oxygen we need, of course, only divide the quantity of oxygen con-
tained in the water by Z, and add the quotient to the oxygen of the magnesia
from which we find, in the case before us, that the whole amount of the oxygen
in R O = 3-63 + |x 5-98 = 563.— (Scheerer.)

ISOMORPHISM. 151

form of cordierite and aspasiolite may be explained by the
assumption, that 3 equiv. (atoms) of water may isomorphously
replace 1 equiv. (1 at.) of magnesia. The manner in which
this substitution is effected has been shown by Scheerer in the
interesting work alluded to, by reference to a large series of
minerals and compounds. If we reckon from the analysis of
the different forms of serpentines, the quantity of oxygen in
the bases, R O, and that of the water, H O, and add to them in
the manner indicated, (that is, by taking from the amount of
the oxygen of the water |3HO == R O) the sum will in all
cases be equal to the oxygen contained in the silica, and hence
we have for all forms of serpentine the simple formula of
(RO), 3Si03. The insertion of the RO means, that in this
portion, a larger or smaller part of the magnesia, and its

isomorphous base (Fe, Mn, &c.,) is replaced by water in the rela-
tion indicated. This formula is, however, that for olivine (3R O,
Si 03) with this difference only, that in the latter no magnesia
is replaced by water. Serpentine may therefore be regarded as
hydrated olivine, and hence we have an explanation of the
reason that crystallized serpentine (of Snarum) has the same
crystalline form as olivine.

The doubly refracting datolite, for which Rammelsberg has
given the formula 2 (3Ca O, Si 03, + 3B 03, Si 03 + 3 H O,)
in which boracic acid plays the part of a base, has, according

to Scheerer's view, the formula 3(Ca Si + Ca B) + (R) Si. The
combinations of magnesia with carbonic acid and water (mag-
nesia alba, &c.) come to have in the same way very simple for-
mulae. In the magnesian sulphates, containing 7 equiv. of
water, 1 equiv. always occurs as water of crystallization, while
6 equiv. ( = 2R,) are combined as bases with the sulphuric acid.
Graham observed that these 6 equiv. of water escape with con-
siderably more facility than the seventh equivalent. He called
the latter constitutional water, and the former water of crystal-
lization ; but, according to Scheerer's mode of observation, we
must invert these terms. As I shall have occasion to revert to
Scheerer's admirable treatise, I must be permitted to remark,
that it would not have lost in any way, if the author had re-
frained from extending polymeric isomorphism (although only
provisionally,) to magnesian sulphates and similar compounds.

152 ISOMORPHISM.

The isomorphous group of the alums, which has already been
mentioned at page 122, ranges itself with both the preceding
isomorphous groups. The general formula for alums is R O,

S 03 + R2 03, 3S 03 + 24H O, or shorter R S, RR S3 + 24 H;
R2 03 may be alumina, sesquioxide of iron, sesquioxide of
chromium, sesquioxide of manganese, (common alum, iron
alum, chrome alum, manganese alum); and R O may be potash,
oxide of ammonium, or soda, (potash alum, ammonia alum,
soda alum). Common alum is, for the most part, potash-and-
alumina alum, and the formula is consequently as follows : —

KS + A1A1S3 + 24H0.

Ammonia and iron alum is Am S + FeFe S3 + 24H O.

Potash and chrome alum is KS + Cr Cr S3 + 24H 0.

All these alums occur in octahedra, and their equivalent
volumes are all equal. In this case we also see that isomor-
phism of the constituents is not the necessary result of isomor-
phism of the combinations. The sesquioxides of iron and chro-
mium and alumina have the same equivalent volume, and may,
therefore, directly replace each other, equivalent for equivalent,
without the forms being deranged, but the metals, chromium
and aluminum, have a different equivalent volume from iron.

Reciprocal substitution of these atoms amongst themselves,
without any change of form, occurs as frequently in the sesqui-
oxides as among the protoxides belonging to the magnesian
group; and it is most clearly indicated in those spinell-like
minerals that crystallize in the regular system, in which the
sesquioxides are contained as negative constituents. The type
of this group of minerals is as follows: —

Spinell

.... MgAlAl

Pleonast

.... M^FeAlAl

Gahuite

.... Zn~MnFe A1A1

Magneticum ....

.... FeFeFe

Chrome iron-stone ....

.... ivMg cTai

Franklinite

.... MnFeZnFeMn

ISOMORPHISM. 153

As the concluding isomorphous group, we may mention the
one which is composed of the double chlorides, from the union
of chloride of potassium, or chloride of ammonium, with the
chlorides of platinum, palladium, iridium, and osmium; to
which belong chloride of potassium and platinum, K CI, Pt Cl2;
chloride of ammonium and platinum, Am CI, Pt Cl2; chlo-
ride of potassium and palladium, K CI, Pd Cl^ &c. These
double chlorides crystallize together in octahedra, and have a
similar constitution. As the equivalent volumes of these four
metals are equal, we may derive the isomorphism of the double
chlorides from the isomorphism of the metals.

Chemists have always been very desirous of finding some
simple physical characteristic, by which to ascertain the atomic
weights ; and identity of volume in a gaseous condition, identity
of specific heat, and similarity in the crystalline form, have in
turn been esteemed as certain means of recognition.

At the proper places we have already spoken at large upon
the subject of the two first-named characteristics, and stated
that, although they must not be left wholly out of the ques-
tion, they nevertheless cannot serve as generally available aids
in determining the atomic weight.

The question as to whether isomorphism may serve, uncon-
ditionally, as a means of determining the atomic weight, and
whether, consequently, the differences of the atomic weights
which it enforces, must, in all cases, be directly taken into
account, will be answered by the consideration of the isomor-
phous groups. It cannot be denied, that the indications of
isomorphism agree better than those of specific heat, and better
than the weight of the volumes with those views of the consti-
tution of bodies, which have been derived from purely chemi-
cal considerations; and they are most invaluable in establishing
an analogy of composition in a group of bodies; whilst they
afford us an exact characteristic, that may be explained by
numbers, in the place of the common and ill-determined re-
semblance between compounds, which has been observed by
chemists more from a practised tact than from rule, and which
has hitherto been their only guide in classification. If, there-
fore, we admit that isomorphism is a certain test of the re-
semblance of atomic constitution, within the limits of a group

154 ISOMORPHISM.

of elements aud their compounds, we must, nevertheless, ques-
tion whether the relation of the atoms to the crystalline form
is, without exception, the same throughout the whole series
of the elements, or whether all agree exactly in this or any
other physical property.

It is even possible that the crystalline form and isomorphism
may not be the consequence of the atomic constitution, or that
they are not immediately and necessarily connected with it, but,
on the contrary, they may arise from a secondary property of
bodies, in which a single atom may accidentally resemble a
compound body. This seems to be proved by sulphur in one of
its two forms being isomorphous with bisulphate of potash,
and that a single atom of potassium is isomorphous throughout
a long series of combinations with the group of 5 atoms com-
posing ammonium. The last circumstance, as well as the iso-
morphism of the cyanogen compounds with the chlorine com-
pounds, (Hg Cy and Hg CI,) shows, at any rate, clearly, that
the compound radicals have the same value in an isomorphous
relation as the simple radicals, — the elements. It must also be
remarked, that Will has exhibited a quinine-and-iron alum;
and that, according to Ortigosa, there probably exists an alum,
which, in the place of R O, contains the alkaloid conine,
(C„ HI6 N.)

The dimorphism of simple as well as compound bodies also
indicates the subordinate character of the crystalline form.
Is it probable that sulphur and carbonate of lime could by
merely a trifling difference of temperature be brought to yield
totally different crystals, if form were the result of an un-
changeable atomic constitution? The crystalline form is
perhaps dependent upon a property of bodies, with which
we are as yet unacquainted, and which may stand in a frequent
and general, although not iu variable, relation to their atomic
condition. If this be granted, it will not be at variance with
the doctrine of isomorphism, if one atom of a certain group
of elements possess the same crystallographic value as two or
more atoms of another group. (Graham.)

Truly isomorphous bodies resemble each other in many
other properties besides form. The corresponding arseniates
and phosphates agree in taste, in the degree of force with

ISOMORPHISM. 155

which they retain water of crystallization and in various
other properties. Seleniate and sulphate of soda, isomorphous
compounds, are both efflorescent salts, and their resemblance,
with regard to their solubility in water, goes so far, that both
show the unusual deviation from the law of increase of solu-
bility at higher temperatures, of being more easily soluble in
water at 92° F., than in water heated to the boiling-point. It
certainly appears that isomorphism is always accompanied by
a resemblance in many other properties, and it seems as if it
were a feature indicative of the closest relationship between
two bodies.

It will subsequently be shown that bodies act more readily
upon each other as solvents, and admit of being mixed in a
fluid form, in proportion to the similarity existing in their
composition. Attraction towards another possessing a similar
character with themselves, is probably the reason that the
particles of isomorphous bodies blend so readily together, and
on this may depend the difficulty of separating them, when
they are once dissolved in a common solvent. Perman-
ganate and perchlorate of potash, isomorphous salts, may
certainly crystallize apart from the same solution, owing
to their solubility being so different; and potash alum may,
by crystallization, be freed from the greatest part of the iron
alum contained in it, because the latter is more easily dissolved,
and remains in the mother-liquid; but most isomorphous salts,
for instance, sulphate of protoxide of iron and sulphate of
copper crystallize when they are dissolved together, yielding
not distinct, but homogeneous, crystals, which are a mixture
of both salts in indefinite proportions. This blending of
isomorphous salts is very frequently found in minerals, as we
have already seen; it was wholly inexplicable, and seemed to
be entirely at variance with the theory of equivalents until
the capacity of isomorphous bodies to crystallize in common
was recognised as a natural law. If isomorphism have only
a limited value in the determination of the atomic numbers of
bodies, it follows, as has already been stated, that the chemist
is especially led to adhere to the equivalents. I again repeat,
the more the investigator divides the equivalents in order
to bring them in accordance with the physical properties of

156 ISOMORPHISM.

bodies, the more will these atoms yield in chemistry to che-
mical atoms, that is, to the equivalent atoms.

The relative equivalents of bodies resembling each other
chemically admit, generally speaking, of being ascertained with
the greatest certainty; not so, however, with respect to the
relative equivalents of elements of heterogeneous chemical
character. There can be no doubt as to the relative quantities
required in the reciprocal substitution of chlorine, iodine,
bromine, oxygen, sulphur; nor as to the same quantities in
hydrogen, potassium, magnesium, calcium, barium, iron, &c. ;
nor the quantities by weight of antimony and arsenic, or even
perhaps of phosphorus and nitrogen, which may replace each
other. But what, for instance, is the quantity of nitrogen or
phosphorus, or arsenic, necessary to take the place of 100 parts
of oxygen? This is another question. Have the elements
a dhTerent equivalent in different combinations? are they in-
fluenced by allotropism?

Graham, who always deduces isomorphism of combinations
from isomorphism of the constituents, which naturally leads us
at last to the isomorphous elements, has divided these elements
into isomorphous groups. As in the formation of these groups,
the isomorphism of the combinations must necessarily serve as
a guiding point throughout these combinations, they must be
of interest even to those chemists, who do not admit without
some reservation that isomorphism of the combinations results
from isomorphism of the constituents, embracing as they do
such elements as are connected together by the isomorphism
of their compounds. This mode of grouping cannot therefore
be passed by without some short notice.

Classification of Elements, (Graham).

The extent to which the isomorphous relations of bodies
have been traced, will appear on reviewing the groups or
natural families in which the elements may be arranged, and
observing the links by which the different groups themselves
are connected; these classes not being abruptly separated, but
shading into each other in their characters, like the classes
created by the naturalist for the objects of the organic
world.

ISOMORPHISM. 157

I. Sulphur Class. — This class comprises four elementary-
bodies: oxygen, sulphur, selenium, tellurium. The three last
of these elements exhibit the closest parallelism in their own
properties, in the range of their affinities for other bodies, and
in the properties of their analogous compounds. They all form
gases with one atom of hydrogen, and powerful acids with
three atoms of oxygen, of which the salts, the sulphates,
seleniates, and tellurates are isomorphous: and the same rela-
tion undoubtedly holds in all the corresponding compounds of
these elements.

Oxygen has not yet been connected with this group by a
certain isomorphism of any of its compounds ; but a close cor-
respondence between it and sulphur appears, in their compounds
with one class of metals being alkaline bases of similar pro-
perties, forming the two great classes of oxygen and sulphur
bases, such as oxide of potassium and sulphide of potassium;
and in their compounds with another class of elements being
similar acids, giving rise to the great classes of oxygen and
sulphur acids, such as arsenious and sulpharsenious acids. They
farther agree in the analogy of their compounds with hydrogen,
particularly of binoxide of hydrogen and bisulphide of hydro-
gen, both of which bleach, and are remarkable for their
instability; and in the analogy of the oxide, sulphide, and
telluride of ethyl, and of alcohol and mercaptan, which last is
an alcohol with its oxygen replaced by sulphur. This class is
connected with the next by manganese, of which manganic acid
is isormorphous with sulphuric acid, and consequently man-
ganese with sulphur.

II. Magnesian Class. — This class comprises magnesium,
calcium, manganese, iron, cobalt, nickel, zinc, cadmium, copper,
hydrogen, chromium, aluminum, glucinum, vanadium, zirco-
nium, yttrium, thorinum. The protoxides of this class, includ-
ing water, form analogous salts with acids. A hydrated acid,
such as crystallized oxalic acid or the oxalate of water, corres-
ponding with the oxalate of magnesia in the number of atoms
of water with which it crystallizes, and the force with which
the same number of atoms is retained at high temperature- ;
hydrated sulphuric acid (HO, S 03 4- HO) with the sulphate
of magnesia (MgO, S 03 + HO). The isomorphism of the

158 ISOMORPHISM.

salts of magnesia, zinc, cadmium, and the protoxides of man-
ganese, iron, nickel, and cobalt, is perfect. Water (H O) and
oxide of zinc (Zn O) have both been observed in thin regular
six-sided prisms; but the isomorphism of these crystals has not
yet been established by the measurement of the angles. Oxide
of hydrogen has not, therefore, been shown to be isomorphous
with these oxides, although it greatly resembles oxide of
copper in its chemical relations. Lime is not so closely related
as the other protoxides of this group, being allied to the follow-
ing class. But its carbonate, both anhydrous and hydrated, its
nitrate, and the chloride of calcium, assimilate with the
corresponding compounds of the group; while to its sulphate
or gypsum, Ca O, S 03 + 2H O, one parallel and isomorphous
compound, at least, can be adduced, a sulphate of iron, Fe O,
S 03 + 2H O (Mitscherlich), which is also sparingly soluble
in water, like gypsum. Glucina is isomorphous with lime
from the isomorphism of the minerals euclase and zoisite.
(Brooke.)

The salts of the sesquioxide of chromium, of alumina, and
glucina, are isomorphous with those of sesquioxide of iron
(Fe2 03), with which these oxides correspond in composition ;
and the salts of manganic and chromic acids are isomorphous,
and agree with the sulphates. The vanadiates are believed
to be isomorphous with the chromates. Zirconium is placed
in this class, because its fluoride is isomorphous with that
of aluminum and that of iron, and its oxide appears to have
the same constitution as alumina; and yttrium and thorium,
solely because their oxides, supposed to be protoxides, are
classed among the earths.

III. Barium Class. — Barium, strontium, lead. The salts
of their protoxides, baryta, strontia, and oxide of lead, are
strictly isomorphous, and one of them at least, oxide of lead,
is dimorphous, and assumes the form of lime, and the preceding
class in the mineral plumbocalcite, a carbonate of lead and
lime. (Johnston.) But certain carbonates of the second class
are dimorphous, and enter into the present class, as the car-
bonate of lime in arragonite, carbonate of iron in junckerite,
and carbonate of magnesia procured by evaporating its solution
in carbonic acid water to dryness by the water-bath (G. Rose),

ISOMORPHISM. 159

which have all the common form of carbonate of strontia.
Indeed, these two classes are very closely related.

IV. Potassium Class. — The fourth class consists of potas-
sium, ammonium, sodium, silver. The term ammonium is
applied to a hypothetical compound of one atom of nitrogen
and four of hydrogen (X H4), which is certainly, therefore,
not an elementary body, and probably not even a metal, but
which is conveniently assimilated in name to potassium, as
these two bodies occupy the same place in the two great classes
of potash and ammonia salts, between which there is the most
complete isomorphism. Potassium and ammonium themselves
are, therefore, isomorphous. The sulphates of soda and silver
are similiform, and hence also the metals sodium and silver;
but their isomorphism with the preceding pair is not so clearly
established. Soda replaces potash in soda alum, but the form
of the crystal is the common regular octohedron; nitrate of
potash has also been observed in microscopic crystals, having
the rhomboidal form of nitrate of soda*, which is better evidence
of isomorphism, although not beyond cavil, as the crystals were
not measured. There are also grounds for believing that
potash replaces soda in equivalent quantities in the mineral
chabasite, without change of form. The probable conclusion
is, that potash and soda are isomorphous, but that this relation
is concealed by dimorphism, except in a very few of their
salts.

This class is connected in an interesting way with the other
classes through the second. The subsulphide of copper and
the sulphide of silver appear to be isomorphous, although two
atoms of copper are combined in the one sulphide, and one
atom of silver in the other* with one atom of sulphur; their
formulae being —

Cu2 S and Ag S.

Are then two atoms of copper isomorphous with one atom of
silver? In the present state of our knowledge of isomorphism,
it appears necessary to admit that they are.

* Frankenheim, in Poggendorf's Annalen, vol. xl., page 447. See also a
paper by Professor Johnston on the received equivalents of potash, soda, and
silver; Phil. Mag., third series, vol. xii., p. 324.

160 ISOMORPHISM.

The fourth class will thus stand apart from the second,
which is represented by copper, and also from the other classes
connected with the second, in so far as one atom of the present
class is equivalent to two atoms of the other classes in the
production of the same crystalline form. This discrepancy
may be at once removed by halving the atomic weight of
silver, and thus making both sulphides to contain two atoms
of metal to one of sulphur. But the division of the equivalents
of sodium, potassium, and ammonium, which would follow that
of silver, and the consideration of potash and soda as suboxides,
are assumptions not to be lightly entertained.

It was inferred by M. Mosander, that lime with an atom
of water is isomorphous with potash and soda, because Ca O
+ H O appears to replace K O or Na O in mesotype, chabasite,
and other minerals of the zeolite family. The isomorphism of
natrolite and scolezite is so explained: Na O, A1203, 2Si03,
2HO with CaO, Al2 03, 2Si03, 3HO. On the other hand,
it is strongly argued by M. T. Scheerer, that one equivalent
of magnesia is isomorphous with three equivalents of water,
from the equality of the forms of cordierite and a new mineral
aspasiolite, the first containing Mg O, and the second 3H O
in its place; and from a review of a considerable number of
alumino-magnesian minerals. One equivalent of oxide of
copper, however, is supposed to be replaced by two equivalents
of water.

Y. Chlorine Class. — Chlorine, iodine, bromine, fluorine.
These four elements form a well-defined natural family. The
three first are isomorphous throughout their whole combinations
— chlorides with bromides and iodides, chlorates with bromates
and iodates, perchlorates with periodates, &c. ; and such fluorides
also as can be compared with chlorides appear to affect the same
forms. The fluoride of calcium of apatite, Ca F, 3(3 Ca 0,P 05),
is also replaced by the chloride of calcium. It is connected
with the second class through perchloric acid; the perchlorates
being strictly isomorphous with the permanganates. But the
formulae of these two acids are —

CI 07 and Mn2 07,
one atom of chlorine replacing two atoms of manganese. Or,

ISOMORPHISM. 161

this class has the same isomorphous relation as the preceding
class to the others: and such I shall assume to be its true
relation. Although halving the atomic weight of chlorine,
which would give two atoms of chlorine to perchloric acid,
is not an improbable supposition, still it would lead to the
same strange conclusion as follows the division of the equi-
valent of sodium, — namely, that chlorine enters into its other
compounds, as well as into permanganic acid, always in the
proportion of two atoms; for that element is never known to
combine in a less proportion than is expressed by its presently
received equivalent. Cyanogen (C2 X), although a compound
body, has some claim to enter this class, as the cyanides have
the same form as the chlorides.

VI. Phosphorus Class. — Xitrogen, phosphorus, arsenic, anti-
mony, and bismuth; also composing a well-marked natural
group, of which nitrogen and bismuth are the two extremes,
and of which the analogous compounds exhibit isomorphism.
These five elements all form gaseous compounds with three
atoms of hydrogen; namely, ammonia, phosphuretted hydrogen,
arsenietted hydrogen, &c. The hydriodates of ammonia and
of phosphuretted hydrogen are not, however, isomorphous.
Arsenious acid and the oxide of antimony, both of which
contain three atoms of oxygen to one of metal, are doubly
isomorphous. Arsenious acid also is capable of replacing oxide
of antimony in tartrate of antimony and potash or tartar emetic,
without change of form ; and arsenic often substitutes antimony
in its native sulphide. The native sulphide of bismuth (Bi S3)
is also isomorphous with the sulphide of antimony (Sb S3).
Nitrous acid (X 03), which should correspond with arsenious
acid and oxide of antimony, likewise acts occasionally as a
base, as in the crystalline compound with sulphuric acid of
the leaden chambers. The complete isomorphism of the arse-
niates and phosphates has already been noticed. But phos-
phoric acid forms two other classes of salts, the pyrophos-
phates and metaphosphates, to which arsenic acid supplies no
parallels.

This class of elements is connected with the others by
means of the following links: — Bisulphide of iron is usually
cubic, or of the regular system; but it is dimorphous, and, in

162 ISOMOKPHISM.

sperkise, it passes into another system, and has the form of
arsenide of iron; Fe S2, or rather Fe2 S4, being isomorphous
with Fe2 As S2. Again, bisulphide of iron, in the pentagonal-
dodecahedron of the regular system, is isomorphous with cobalt-
glance, Fe2 S4 with Co2 As S2 : so that one equivalent of arsenic
appears to be isomorphous with 2S. This is also supported by
the isomorphism of the sulphide of cadmium and sulphide of
nickel (Cd S and Ni S, or Cd2 S2 and Ni2 S2), with the arsenide
of nickel (Ni2 As). Tellurium has also been observed in the
same form as metallic arsenic and antimony. The phosphorus
class approximates also to the chlorine class; nitrogen and
chlorine both forming a powerful acid with five equivalents
of oxygen, nitric acid, and chloric acid; but of the many
nitrates and chlorates which can be compared, no two have
proved isomorphous. Nor do the metaphosphates appear at
all like the nitrates, although their formulas correspond.

Nitrogen, it must be admitted, is but loosely attached to
this class. It is greatly more negative than the other members
of the class, approaching oxygen in that character, with which,
indeed, nitrogen might be grouped, N being equivalent to 20.
For while phosphuretted hydrogen is the hydride of phospho-
rus, or has hydrogen for its negative and phosphorus for its
positive constituent, ammonia is undoubtedly the nitride of
hydrogen, or has nitrogen for its negative and hydrogen for its
positive constituent. The one should be written PH3, and the
other H3 N — a difference in constitution which separates these
bodies very widely. An important consequence of classing
nitrogen with oxygen is, that, in the respective series of com-
pounds of these elements, cyanogen becomes the analogue of
carbonic oxide, C2 N being equivalent to CO, or, rather
C202.

VII. Tin Class. — Tin, titanium. Connected by the iso-
morphism of titanic acid (Ti 02) in rutile with peroxide of tin
(Sn 02) in tin-stone. Titanium is connected with iron and the
second class. Ilmenite and other varieties of titanic iron which
have the crystalline form of the sesquioxide of that metal,—
namely, that of specular iron, and also of corundum (alumina),
—are mixtures of a sesquioxide of titanium (Ti2 03) with ses-
quioxide of iron (H. Rose).

ISOMORPHISM. 163

VIII. Gold Class. — Gold, which is isomorphous with silver
in the metallic state. Gold will thus be connected, through
silver, with sodium and the fourth class.

IX. Platinum Class. — Platinum, iridium, osmium. From
the isomorphism of their double chlorides. The double bichlo-
ride of tin and chloride of potassium crystallizes in regular
octahedrons, like the double bichloride of platinum and potas-
sium, and other double chlorides of this group ; which, although
not alone sufficient to establish an isomorphous relation between
this class and the seventh, yet favours its existence (Dr. Clark).
The alloy of osmium and iridium (Ir Os) is isomorphous with
the sulphide of cadmium (Cd S) and sulphide of nickel (Ni S)
(Breithaupt).

X. Tungsten Class. — Tungsten, molybdenum, tantalum,
niobium, and pelopium. From the isomorphism of the tung-
states and molybdates, the salts of tungstic and molybdic acids,
W03 and M0O3. Tantalic acid is isomorphous with tungstic
acid : tantalite (FeO, Ta03) with wolfram (FeO, W03). So
are molybdic and chromic acids ; the tungstate of lime, tung-
state of lead, molybdate of lead, and chromate of lead (in the
least usual of its two forms), being all of the same form. This
establishes a relation between molybdic, chromic, sulphuric,
and other analogous acids*. Niobium and pelopium are in-
troduced into this class as they replace tantalum in the tanta-
lites of Bavaria.

XI. Carbon Class. — Carbon, boron, silicium. These ele-
ments are placed together, from a general resemblance which
they exhibit without any precise relation. They are not known
to be isomorphous among themselves, or with any other element.
They are non-metallic, and form weak acids with oxygen, — the
carbonic, consisting of two of oxygen and one of carbon, and
the boric and silicic acids, which are generally viewed as com-
posed of three of oxygen to one of boron and silicium. Silicic
acid may, perhaps, replace alumina in some minerals, but this
is uncertain.

Of the elements which have not been classed, no isomor-
phous relations are known. They are mercury, which in some

Johnston, Phil. Mag. 3d series, vol. xii. p. 387.

M 2

164 ISOMORPHISM.

of its chemical properties is analogous to silver, and in others
to copper, cerium, didymium, lanthanum, lithium, rhodium,
ruthenium, palladium, and uranium. Ruthenium, however, is
believed to be isomorphous with rhodium, from the correspon-
dence in composition of their double chlorides. Didymium and
lanthanum are also probably isomorphous with cerium, as they
appear to replace that metal in cerite.

165

PHYSICAL INVESTIGATIONS ON DYEING.

By M. Chevreul.

ON THE INFLUENCE THAT TWO COLOURS MAY

EXERCISE UPON EACH OTHER WHEN

SEEN SIMULTANEOUSLY.

Introduction.

The investigations I have pursued on the subject of dyeing,
considered in the most general and comprehensive point of
view, may be classed in three different series.

The first Series embraces, under the head of physical
enquiries :

1. All that has relation to the principle of the simulta-
neous contrast of colours: this principle is so infinitely varied,
that in spite of my wish to concentrate my labours within
the department of chemistry, as applicable to dyeing properly
so called, I have been unable to refrain from directing my
attention to the establishment of a theory that might serve as
a guide in those arts and manufactures, whose object it is to
challenge attention by the assortment and arrangement of
colours. This will explain how my first article on this subject,
published in 1828, and printed in the 11th volume of the
Memoires de TAcademie, led me into a course of investigations
of such extent that they occupy, together with my previous
observations, an octavo volume of 721 pages, which appeared
in 1839.

166 PHYSICAL INVESTIGATIONS ON DYEING.

2. A theory of the optical effects of articles of silk.

3. The exposition of a mode of defining and naming
colours according to a rational and experimental method.

The Second Series comprises investigations which I term
physico-chemical, from their depending on the principle of the
mixture of colours, which come within the department of
physics, and being at the same time connected with chemical
actions in all those cases, in which the principle has to be
applied to the fixation of several coloured substances on stuffs
of different kinds, by means of the process of dyeing.

The Third Series includes my chemical investigations, pro-
perly so called, on dyeing. Six memoirs have already appeared
in the Recueil des Memoires de VAcademie, and are as follows.

First Memoir. An introduction and general considerations
on dyeing. Vol. XV., p. 383, Memoires de V Academic

Second Memoir. On the proportion of water that various
stuffs absorb in atmospheres of 65, 15, 80, and 100 of Saus-
sure's hygrometer. Vol. XV., p. 409. Introduction to the
third, fourth, fifth, and sixth Memoirs, Vol. XVI., p. 41.

Third Memoir. On the action of pure water on stuffs dyed
with various colouring matters. Vol. XVI., p. 47.

Fourth Memoir. The changes effected by light, atmospheric
agents, and hydrogen gas on curcuma or turmeric, sumach,
carthamus, archil, saxon blue (sulfo-indigotic acid), indigo, and
prussian blue, when impressed on cotton, silk, and woollen
stuffs. Vol. XVI., p. 53.

Fifth Memoir. On the changes effected by heat and atmo-
spheric influences on curcuma or turmeric, sumach, carthamus,
archil, saxon blue (or sulfo-indigotic acid), indigo, and prussian
blue, and other colouring matters impressed on cotton, silk,
and woollen stuffs. Vol. XVI, p. 181.

Sixth Memoir. On the several changes of colour expe-
rienced by prussian blue when impressed upon different stuffs.
Vol. XIX, p. 491.

An appendix to this Memoir, containing some general con-
siderations and inductions relative to the nature of living
organized beings.

Seventh Memoir. On the composition of wool; on the
theory of clearing wool of the yolk, and some properties de-

PHYSICAL INVESTIGATIONS ON DYEING. 167

rived from its composition, that may influence the arts and
manufactures to which it is applied. Read before the Academy,
April 20, 1840.

Eighth Memoir. Considerations on the theory of dyeing,
and applications of this theory to the perfection of several
practical processes in general, and to that of dyeing with indigo
in the blue vat, in particular. Read before the Academy,
November 23, 1846.

1. The frequent opportunities I have had of observing the
very various and strongly contrasting colours, that are required
in the royal manufactories of tapestry have enabled me to make
some observations, which I trust will not be wholly useless
to those whose object it is, by means of an assortment of
differently coloured objects, to produce the best possible effect
to please the eye. The phenomena of which I shall treat in
this paper fall under the head of what natural philosophers
term accidental colours, in accordance with the term applied
to them by Buffon, who was the first to treat of the subject
with any degree of detail; before, however, I enter more fully
into the question, I would direct attention to some of the
principles of optics that have the most intimate connection
with the subject of which I am treating.

2. A ray of solar light is composed of an indeterminate
number of differently coloured rays; since, on the one hand, it
is impossible to distinguish each in particular, and as, on the
other, they do not all differ equally from one another, they have
been divided into groups, to which have been applied the terms
red rays, orange rays, yellow rays, green rays, blue rays, indigo
rays, and violet rays. It must not, however, be supposed that
all the rays comprised in the same group, as, for instance, in
that of the red rays, are identical in colour. On the contrary,
they are generally considered as capable of differing more or
less among themselves, although we recognise the sensation
separately produced by each one as comprised in that which
we ascribe to red.

3. When light is reflected by an opaque white body, it
does not experience any modification in the proportion of the

168 PHYSICAL INVESTIGATIONS ON DYEING.

different coloured rays, which constitute white light. If this
body be not polished, each point of its surface must be considered
as radiating in every direction the light falling upon it; and if
the body be polished, a regular or specular reflection will be pro-
duced, but in this case, as in the former, there will be a certain
quantity of light reflected irregularly, or in every direction.

4. When the light is reflected by an opaque coloured body
there is always: (1) a reflection of white light, (2) a reflection
of coloured light, which is owing to the fact that the body
absorbs or extinguishes in its interior a certain number of
coloured rays, and reflects others. It is evident that the rays
reflected are of a different colour from those that have been
absorbed, and besides, that if they were combined with the
former, white light would be reproduced. It is this mutual
relation possessed by coloured rays of again forming white
light by their blending together that has led us to term certain
of them as complementary of others. Further, it is evident
that opaque coloured bodies reflect white and coloured light,
both regularly and irregularly, or only irregularly, according
as to whether or not they are polished.

5. It would be erroneous to suppose that a red or a yellow
body reflected, besides white light, only red or yellow rays;
each of these bodies reflects, besides, all kinds of coloured rays ;
but those rays, which make us believe it to be red or yellow,
being more numerous than the others, produce a greater effect;
these other rays exercise, however, an incontestible influence
in modifying the action of the red or yellow rays upon our
organs of vision, and this will explain the innumerable diffe-
rences of shade which we remark among different red and
different yellow bodies, &c.

6. I shall explain in my first paragraph:

1. The manner of observing the phenomena treated of
in this memoir:

2. The law of these phenomena, and the formula re-
presenting them:

3. The application of the law to a certain number of
these phenomena:

4. The effect of colours upon white, and reciprocally
of white upon colours :

PHYSICAL INVESTIGATIONS ON DYEING.

169

5. The effect of colours upon black, and reciprocally
of black upon colours :

6. The influence of the chemical nature of coloured
bodies upon the phenomena observed:

7. The reciprocal influence of different kinds of colours
belonging to the same group:

8. The interpretation of the phenomena on the hypo-
thesis that the red, yellow, and blue are simple colours,
and that the orange, green, indigo, and violet are compound
colours :

9. The connection between my observations on acci-
dental colours and those made by earlier observers :

10. The physiological cause to which the explanation of
accidental colours is referred.

I shall treat in a second paragraph of some applications of
the preceding observations.

Tie. 1.

Article I. — Manner of observing the Phenomena treated of in

this Memoir.

7. If we look simultaneously at two narrow zones of
different colour placed side
by side, the colours will be
more or less modified. The
following is a very simple
manner of convincing our-
selves of this proposition.

8. We take two bands,
o and o', (fig. 1) of the same
colour and identical, and two
other bands, P and p', of an-
other colour and identical.
They must be 0*o inch in
width, and 2*5 inches in
length. They may be formed
of some stuff, or of paper, or
of ribbon, of the width I have

170

PHYSICAL INVESTIGATIONS ON DYEING.

indicated. We then paste the band o' with gum to a card,
put o at the distance of J^th of an inch and the band p in
such a manner that it shall touch o; and finally paste on p',
at the distance of -^-th of an inch from p.

9. Now if we look at the card in a certain direction and
during some seconds, we shall almost always see four differently
coloured bands. It must be observed that o' and p' serve
as terms of comparisons to judge of the modifications expe-
rienced by o and p in their juxta-position.

10. I give seventeen observations by way of illustration :

Colours used in the
Experiment.

Red

Orange

Red

Yellow

Red

Blue

Red

Indigo

Red

Violet

Orange
Yellow

Orange
Green....

Orange
Indigo

Orange

Violet

Yellow
Green....

Yellow

Blue

Green....

Blue

Green...
Indigo

Green....

Violet

Blue

Indigo

Blue

Violet

Indigo

Violet

Modification.

inclining

to violet,
yellow.

violet, or less yellow.

green, or less red.

yellow,
green.

yellow,
blue.

55
5)

yellow,
indigo.

55
55

red.

bright green, or less red.

55
55

red.
blue.

55
33

yellow, or less brown,
blue, or brighter indigo.

55
33

yellow, or less brown,
indigo.

53
•5

brilliant orange,
blue.

55
35

orange,
indigo.

33
33

yellow,
indigo.

33
33

yellow,
violet.

35
35

yellow,
red.

55
35

green,
deep violet.

green,
red.

55
33

blue,
red.

PHYSICAL INVESTIGATIONS ON DYEING.

171

11. Before I proceed further, I would wish to draw atten-
tion to one of the most important observations recorded in this
memoir; namely, that the reciprocal modifications of colours are
not limited to the case where the modifying coloured zones are
contiguous to one another, for they may be observed even when the
zones are separated. The following experiment will show this
conclusively: take two stripes of the same blue paper, o, o',
(fig. 2) and two stripes of the same green paper, p, p'. The

Fi*. 2.

r^^

' t :>;i ,

blue and green must be of the same height of tone. The
stripes are to be 4 inches in length, and 0*8 inch in width.
Place them parallel to one another, in such a manner that o is
at 0*44 inch from P, o' at 0*28 inch from o, and finally p' at
0*28 inch from p. Standing then at six paces from the card,
you will see the colours modified: o will be of a less green
blue than o', and P will be of a green more yellow than p'.

I shall frequently have occasion to revert to this remarkable
fact.

172 PHYSICAL INVESTIGATIONS ON DYEING.

Article II. — Law of the preceding Phenomena, and the Formula

representing them.

12. After having satisfied myself that the above-mentioned
phenomena were constantly presented to my sight when not
fatigued, and that many persons accustomed to judge of colours
saw them as I did, I endeavoured to reduce them to some
sufficiently general expression to enable one to foresee the
effect that would be produced upon the organ of sight by the
juxtaposition of two given colours. All the phenomena that
I have observed seem to me to depend upon a very simple law,
which, taken in its most general signification, may be expressed
in these terms: in the case of the eye seeing at the same time two
colours which are in contact, they will appear as dissimilar as
possible.

13. From what I have said of the complementary colours,
it is evident that the colour of the stripe O (fig. 1) will differ
as much as possible from that of the stripe P, when the com-
plementary colour of P is added to the colour of o; in like
manner, the colour of P will differ in the greatest possible
degree from the colour of O, when the complementary colour
of the latter is added to the colour of P. Consequently in
order to know what the two colours o and P will be when in
juxtaposition, it will be sufficient to find the complementary
colour of p and add it to the colour o, and the complementary
colour of o and add it to p.

14. An analogous result would be obtained by taking the
colour p from o, and o from p.

15. Let us represent

The colour of the stripe o by a, more white by B.
The colour of the stripe P by a, more white by b'.
The complementary colour of a by c.
The complementary colour of a! by c.

Considered according to the first manner of seeing them (13),
the colours of the two stripes seen separately will be :

Colour of o=a + B.
Colour of P = a + b'.

PHYSICAL INVESTIGATIONS ON DYEING. 173

They will become by juxtaposition as follows;

Colour of O = a + B + c.
Colour of P = a' + b' + c.

According to the second method of seeing them (14) we

suppose

. . . [white = b.

b reduced to two portions = < . , ., / / , /\

r \+ white = {a + c ).

, , _ . f white = V

b reduced to two portions = I , , . , , N

[ 4- white = {a + c).

The colours of the two stripes seen separately will be :

Colour of o== a + b + a' + c
Colour of P == a + b ' + a + c;

They will become by juxtaposition as follows:

Colour of 0= a + b + c'
Colour of P= «' + b' + c.

The results are the same excepting that there is less white
in this case than in the other.

ARTICLE III. — Application of the Law to the seventeen Observa-
tions of Article I.

Orange and green.

16. Blue (the complementary colour to orange) on being
added to green makes the latter incline towards a bluish tint,
or renders it less yellow.

Red (the complementary colour to green) when added to
orange makes the latter incline towards a reddish tint, or makes
it less yellow.

Orange and indigo.

17. Blue (the complementary colour to orange) when
added to indigo makes it incline towards a blue tint, or renders
it less red.

Yellow inclining to orange (the complementary colour to
indigo) makes orange incline towards yellow, or makes it less
red.

174 PHYSICAL INVESTIGATIONS ON DYEING.

Orange and violet.

18. Blue (the complementary colour to orange) makes
violet incline towards indigo.

Yellow inclining towards green (the complementary colour
to violet) makes orange incline towards a yellow tone of
colour.

Green and indigo.

19. Red (the complementary colour to green) when added
to indigo renders it more violet, or more red.

Yellow inclining to orange (the complementary colour to
indigo) on being added to green causes it to incline towards
yellow.

Green and violet.

20. Red (the complementary colour to green) on being
added to violet gives it a redder tinge.

Yellow inclining to green (the complementary colour
to violet) on being added to green makes it incline to
yellow.

Orange and red.

21. Blue (the complementary colour to orange) on being-
added to red makes it incline towards violet.

Green (the complementary colour to red) makes orange
incline towards yellow.
Violet and red.

22. Yellow inclining to green (the complementary colour
to violet) on being added to red makes it incline to orange.

Green (the complementary colour to red) causes violet to
incline towards indigo.
Indigo and red.

23. Yellow inclining to orange (the complementary colour
to indigo) on being added to red, causes it to incline towards
orange.

Green (the complementary colour to red) makes indigo
incline to blue.

Orange and yellow.

24. Blue (the complementary colour to orange) makes
yellow incline to green.

Indigo inclining to violet (the complementary colour to
yellow) give a reddish tinge to orange.

PHYSICAL INVESTIGATIONS ON DYEING. 175

Green and yellow.

25. Red (the complementary colour to green) on being
added to yellow makes it incline to orange.

Indigo inclining to violet (the complementary colour to
yellow) makes green incline towards blue.
Green and blue.

26. Red (the complementary colour to green) on being
added to blue makes it incline towards indigo.

Orange (the complementary colour to blue) on being added
to green makes it incline towards yellow.
Violet and blue.

27. Yellow inclining to green (the complementary colour
to violet) makes blue incline to green.

Orange (the complementary colour to blue) on being added
to violet makes it incline to red.
Indigo and blue.

28. Yellow inclining to orange (the complementary colour
to indigo) on being added to blue makes it incline to green,

Orange (the complementary colour to blue) on being added
to indigo makes it incline to violet.
Red and yellow.

29. Green (the complementary colour to red) on being
added to yellow makes it incline to green.

Indigo inclining to violet (the complementary colour to
yellow) makes red incline to violet.
Red and blue.

30. Green (the complementary colour to red) on being
added to blue makes it incline towards green.

Orange (the complementary colour to blue) on being added
to red makes it incline towards orange.
Yellow and blue.

31. Indigo inclining to violet (the complementary colour
to an orange yellow) makes blue incline towards indigo.

Orange (the complementary colour to blue) makes yellow
incline towards orange.
Indigo and violet.

32. Yellow inclining to orange (the complementary colour
to indigo) on being added to violet makes it incline towards
red.

176 PHYSICAL INVESTIGATIONS ON DYEING.

Yellow inclining to green (the complementary colour to
violet) on being added to indigo makes it incline towards blue.

33. It is evident that, other things being the same, the
modification of colours in juxtaposition will be more marked
in proportion to the difference between the complementary
colours (c or c) added to each; for supposing that the com-
plementary colour c added to o be identical with it, as would
be the case with the complementary colour c added to the
colour p, the modifications of o and p would be confined to
a mere augmentation in the intensity of the colour. But do
we know at the present day any two coloured bodies capable
of presenting to our view two perfectly pure colours comple-
mentary to each other? Assuredly not. All that we observe
coloured by reflection transmit, as I have already remarked (5),
besides white light, a great number of differently coloured
rays. We are not, therefore, able at present to name a red
body and a green body, or an orange body and a blue body,
or a body of a yellow inclining to orange and an indigo-
coloured body, or, finally, a body of a yellow inclining to
green and a violet-coloured body, reflecting colours that are
perfectly pure and complementary to each other, so that their
juxtaposition shall merely occasion a simple augmentation of
intensity in colour. If, therefore, it be less easy in general to
verify the law of contrast with respect to red and green bodies,
or orange-coloured and blue bodies, 8fc3 than with reference to
those of which I have treated in the seventeen observations,
detailed at (10), we shall find that, in endeavouring to establish
this law for the first-named bodies, their colours will acquire
the most remarkable splendour, vivacity, and purity, and this
result, which is perfectly conformable to the law, will be easily
understood, since any object, for instance, of an orange colour
reflects blue rays in the same manner as an object of a blue
colour reflects orange rays (5). Thus, it is evident that when
you place a blue stripe in contact with one of an orange colour,
the colours of the two objects in juxtaposition will be mutually
purified and rendered more brilliant, whether this arise from
the first-named stripe imbibing blue from the vicinity of the
second, as that again receives orange from the vicinity of the
blue stripe (13), or whether we assume that the blue stripe

PHYSICAL INVESTIGATIONS ON DYEING. 177

destroys the effect of the blue rays of the second stripe, as
that destroys the effect of the orange rays coming from the
blue stripe (14). It may, however, happen that the blue
appears to incline to green or violet, and the orange to yellow
or red, that is to say, that the modification is not limited to
intensity of colour, but extends likewise to the physical com-
position; whatever the case may be, if the latter effect be pro-
duced, it will incontestibly be much more feeble than the
former, and furthermore, on looking a certain number of times
at the same coloured stripes, you will observe that the blue,
which at first appeared more green, will soon appear more violet,
and that the orange, which had at first seemed to be more
yellow, will soon appear to be more red, so that the phenomenon
of modification, which depends upon the physical composition
of the colour, will not be so constant as those which are treated
of in the seventeen preceding observations (10).

I will, however, detail the observations that I have made
upon bodies whose colours are nearly complementary to each
other.

Red and green.

34. Red (the complementary colour to green) on being
added to red increases its intensity.

Green (the complementary colour to red) on being added
to green increases its intensity.

Such is the theoretical result.

The practical result is generally in conformity with it, when
a green inclining to yellow is brought in juxtaposition with:

1. A red inclining to orange.

2. A crimson red.

3. An intermediate red.

In repeating these observations upon each of these assem-
blages of colour a certain number of times, different results
might be noticed, that is to say, in one case, the red may ap-
pear more orange, and the green more yellow, and in another,
the red may appear more violet, and the green more blue, and
here it may be observed, that this change may be attributed
sometimes to a difference in the intensity of the light shining
upon the colours, and sometimes, again, to fatigue of the
eyes.

178 PHYSICAL INVESTIGATIONS ON DYEING.

On bringing a green of a less yellow or more blue character
in juxtaposition with:

1. A red inclining to orange,

2. A crimson red,

3. An intermediate red,

the results will be the same as in the case of the first green,
with this difference, however, that in the blending of the blue,
green, and the crimson red, when observed a certain number of
times, the green and the red will appear almost constantly more
yellow than they are separately, a result which may easily be
conceived.

Orange and blue.

35. Blue, (the complementary colour to orange,) on being
added to blue, increases its intensity.

Orange, (the complementary colour to blue,) on being added
to orange,) increases its intensity.

On repeating these experiments with a deep blue and an
orange, which is not too red, the two colours will appear most
frequently to assume a reddish tinge, otherwise one might
observe the contrary.

Yellow, inclining to orange, and indigo.

36. Yellow, inclining to orange, (the complementary colour
to indigo,) adds intensity to a yellow inclining to orange.

Indigo (the complementary colour to yellow inclining to
orange,) adds intensity to indigo when added to it.

The results of observation and theory are almost invariably
in accordance with one another.

37. Yellow, inclining to green, (the complementary colour
to violet,) on being added to the same colour gives it intensity.

Violet (the complementary colour to yellow inclining to
green,) gives intensity to violet when added to it.

The result of the experiment is almost always in conformity
with the law.

Article IV.— Of the Effect of Colours upon White, and
reciprocally of White upon Colours.

38. If the law, which I have already established, be cor-
rect, it will be understood that white itself will be affected by
the presence of colours ; thus if you bring a colour o in juxta-

PHYSICAL INVESTIGATIONS ON DYEING. 179

position with white, the latter will appear slightly coloured by
the complementary colour to 0, although it must be owned that
the colour is too feeble to be determined with complete cer-
tainty. I have, therefore, contented myself with observing if
the colour complementary to the colour used in the experiment
corresponded with the tint as seen by my eyes upon the white
stripe opposed to the coloured one.
Red and white.

39. Green (the complementary colour to red,) blends with
the white.

Red appears more brilliant and deeper.
Orange and white.

40. Blue (the complementary colour to orange,) blends
with the white.

Orange appears more brilliant, and of a deeper colour.
Yellow inclining to green, and white.

41. Violet (the complementary colour to yellow inclining
to green,) blends with white.

Yellow appears more brilliant and deeper.
Green and white.

42. Red (the complementary colour to green,) blends with
white.

Green appears more brilliant and deeper.
Blue and white.

43. Orange (the complementary colour to blue,) blends
with white.

Blue appears more brilliant, deeper, and perhaps more green.
Indigo and white.

44. Yellow, inclining to orange, (the complementary
colour to indigo,) blends with white.

Indigo appears more brilliant and deeper.
Violet and white.

45. Yellow, inclining to green, (the complementary colour
to violet,) blends with white.

Violet appears more brilliant and deeper.
Black and white.
46. Black and white, which may, in some degree, be con-
sidered as complementary to each other become, conformably to
the law, more different from each other than when seen sepa-

180 PHYSICAL INVESTIGATIONS ON DYEING.

rately, and this is owing to the effect of the white light reflected
by the black, being more or less destroyed by the light of the
white band. By an analogous action white heightens the tone
of the colours with which it is brought in juxtaposition.

Article V. — Of the Effect of Colours upon Black, and
reciprocally of Black upon Colours.

47. The phenomena presented by black, when exposed to
the influence of colours, appears to me to be owing to the
colour with which it is brought in contact, acting, relatively
to the eye, upon the white light reflected by the black surface
in the same manner as if it were brought in juxtaposition with
a white surface. According to this, the black should be
tinged by the complementary tone of the colour touching it,
and as the tinge which it assumes is not weakened by so much
white light, as in the case where the colour is brought in con-
tact with white, it must to be so much the more striking. On
the other hand, as white heightens the tone of colours brought
in contact with it, black, on the contrary, tends to make them
lighter. The tone of black must depend (1) upon the colour
added to it; thus, for instance, an orange-coloured red, an
orange-coloured yellow, or a yellowish green, will brighten it,
whilst indigo, even if it does not heighten the tone, will at any
rate not reduce it as the first-named colours. (2) Upon the
force or brilliancy of the colour in juxtaposition with it; thus
bright colours, like orange and yellow, will tend by their
brilliancy to add force to black, whilst sombre colours, such as
blue and indigo, do not produce a similar effect.

Red and black.

48. Green (the complementary colour to red,) blends with
black, and makes it appear less reddish.

The red becomes more brilliant, and has less of an orange
or brown tone of colour.
Orange and black.

49. Blue (the complementary colour to orange,) blends
with black, and makes it appear less red or more blue.

The orange becomes brighter and yellower, or less brownish.
Yellow, inclining to green and black.

50. Violet (the complementary colour to a greenish yel-

PHYSICAL INVESTIGATIONS ON DYEING. 181

low,) blends with black, and the latter then acquires a violet-
coloured hue.

The yellow appears brighter, more greenish, or lighter.
Green and black.

51. Red (the complementary colour to green,) blends with
black, rendering it more violet or reddish.

The green inclines slightly to yellow.
Blue and black.

52. Orange (the complementary colour to blue,) blends
with black, and brightens it.

The blue inclines slightly to green.
Indigo and black.

53. Yellow, inclining to orange, (the complementary colour
to indigo,) blends with black, and brightens it considerably.

The indigo becomes brighter.
Violet and black.

54. Yellow, inclining to green, (the complementary colour
to violet,) blends with black, and brightens it.

The violet becomes more brilliant, lighter, and redder.

Article VI. — Upon the Influence of the Chemical nature of
coloured Bodies upon the Phenomena observed.

55. It is necessary to examine into the influence exercised
by the chemical nature of coloured bodies in juxtaposition on
their reciprocal modifications. The result at which I have
arrived is, that all the above-named modifications occur, let the
chemical nature of the coloured bodies in juxtaposition be what
it may, provided always, that the coloured substances that are
substituted for one another, are identical in colour.

Example. — The results were the same in using indigo for
Prussian blue or ultramarine ; likewise in substituting stripes
of orange coloured with minium for stripes coloured with anotto
or with woad-yellow reddened either by madder or cochi-
neal.

Article VII. — On the Reciprocal Influence of different kinds
of Colours appertaining to the same Group.

56. Whenever a great difference is produced by the juxta-
position of two colours, it is rendered appreciable by bringing

182 PHYSICAL INVESTIGATIONS ON DYEING.

the same colour successively in contact with the various colours
belonging to one group, for example :

1. Red and orange.

57. On placing a scarlet, or a crimson red in contact with
an orange, the red will invariably acquire a purple, and the
orange a yellow tone of colour.

2. Red and violet.

58. Analogous results are obtained on bringing a scarlet
and crimson red in contact with violet. The latter will appear
to be bluer, and the red more yellow or less purple.

59. These observations will clearly explain the cause of our
obtaining results conformable to the formula, notwithstanding
that colours or papers may have been used far from presenting
to the eye any very true colours.

60. The juxtaposition of coloured stripes affords a means
of demonstrating the difficulty of fixing the type of colours,
thus :

1. On taking red and bringing it in contact with an
orange-coloured red, the former will appear purple and the
latter more yellow, as I have already remarked; but on
placing the first-named red in contact with a purple red, the
latter will become more blue, and the former more yellow
or orange, so that the same red will be purple in one case
and orange in the other.

2. On taking yellow and putting it in juxtaposition
with an orange-coloured yellow, the former will appear
greenish and the latter more red; but again, on bringing
the first-named yellow in contact with a greenish yellow,
the latter will appear ^greener and the former more orange,
so that the same colour will, in one case, incline to green
and in another to orange.

3. On taking blue and placing it in contact with
greenish blue, the former will incline to violet, and the
latter will appear more yellow. If the same blue be
brought in contact with a violet blue, the former will incline
to green and the latter will appear more red; so that
the same blue will have a violet tinge in one case, and a
greenish hue in the other.

61. We thus see that those colours which are termed by

PHYSICAL INVESTIGATIONS ON DYEING. 183

painters simple or primitive, as red, yellow, and blue, insensibly
pass by the effect of juxtaposition into the condition of com-
pound colours, the same red becoming purple or orange, the
same yellow orange or green, and the same blue appearing
either to be green or violet.

AKTICLE VIII. — On the Interpretation of these Phenomena on
the Hypothesis that Red, Yellow, and Blue are Simple Colours,
and Orange, Green, Indigo, and Violet are Compound
Colours.

62. The instances in which I have applied the principle of
the modification that colours experience by juxtaposition, and
the explanation resulting from the manner in which we consider
the composition of white light physically, are further elucidated
by the terms adopted by painters and dyers, who only admit of
three primitive colours, red, yellow, and blue. As there may
be many who entertain the same opinion, but who might,
nevertheless, wish to understand the phenomena resulting from
the juxtaposition of colours, I will proceed to give an explana-
tion of them, conforming to the terms commonly used; and, for
the sake of perspicuity, I purpose dividing the whole into H\e
groups, beginning with those which comprehend the observa-
tions to which the above-named law most readily applies. I
will suppose, then, that orange is formed of red and yellow,
green of yellow and blue, and indigo and violet of blue and
red.

First Group. Two compound colours having a simple
colour as their common element.

It is very easy to verify the law by observing two colours
comprehended in this group : we perceive that, owing to their
reciprocal influence, they lose more or less of the colour
common to both ; and it is therefore evident that the degree in
which they differ will be proportionate to the amount of this
loss.

1. Orange and green.

These two colours, which have yellow as a common element,
lose it by juxtaposition, thus : the orange icill appear more red
and the green more blue.

184 PHYSICAL INVESTIGATIONS ON DYEING.

2. Orange and indigo.

These two colours, which have red as their common element,
lose it by juxtaposition : the orange appearing more yellow and
the indigo more blue,

3. Orange and violet.
As the preceding.

4. Green and indigo.

These two colours, which have blue as their common
element, lose it by juxtaposition: the green appearing yellower
and the indigo redder.

5. Green and violet.
Like the preceding.

Second Group. A compound colour, composed of one
simple and one compound colour.

1. Orange and red.

The orange loses its redness and appears yellower, and the red
becomes more blue, differing as much as possible from orange.

2. Violet and red.

Violet loses its redness and appears more blue, the red
becomes j^ellow, differing as much as possible from violet.

3. Indigo and red.
Like the preceding.

4. Orange and yellow.

Orange loses its yellow tinge and becomes redder, red
becomes more blue, differing as much as possible from orange.

5. Green and yellow.

Green loses its yellow tinge and appears more blue, the
yellow becomes red, differing as much as possible from green.

6. Green and blue.

Green loses its blue tinge and becomes yellower, the blue
becomes red, differing as much as possible from green.

7. Violet and blue.

Violet loses its blue tinge and appears redder, the blue
becomes yellow, differing as much as possible from violet.

8. Indigo and blue.
Like the preceding.

Third Group. Two simple colours.

1 . Red and yellow.
Red, on losing yellow, will appear more blue, and yellow, on

PHYSICAL INVESTIGATIONS ON DYEING. 185

losing red, will appear more blue ; or, in other words, the red in-
clines to purple and the yellow to green.

2. Red and blue.

Red, on losing blue, will appear yellower, and blue, on
losing red, will appear yellower ; or, in other words, the red
inclines to orange and the blue to green.

3. Yellow and blue.

Yellow, on losing its blue, will appear redder, and blue, on
losing yellow, will appear more violet ; or, in other words, the
yellow inclines to orange and the blue to violet.

Fourth Group. Two compound colours having the same
simple colours.

Indigo and violet.

As indigo only differs from violet in containing a larger
proportion of blue in comparison with the red, it follows that
the difference will be very considerably increased by the indigo
losing red and inclining to a greenish blue, whilst the violet,
acquiring more red, will incline to that colour. It is evident
that if the violet lost its red, or the indigo gained more red, the
two colours would approximate; but as they vary from one
another the first-named effect will be produced.

We may further explain this phenomenon by considering
indigo relatively to violet as blue; thus it will lose its blue,
that being common to both colours, and will incline to green.

Fifth Group. A compound colour and a simple colour
which does not occur in it.

If we adopt the hypothesis that orange, green, indigo, and
violet are compound, and red, blue, and yellow simple colours,
it necessarily follows, that, on bringing one of the four com-
pound colours (supposed to be perfectly free from any admixture
with a colour foreign to its two elementary colours) in contact
with one of the three simple colours not comprised in its
composition, we do not see any reason, by which to explain
the loss sustained by the compound colour of one, rather than
of another of its elementary colours, and why the simple colour
should be removed from one rather than another of the elemen-
tary colours. For instance, on placing green in contact with
red, there seems no reason that the green should tend towards
blue rather than yellow, or why the red should incline to blue
rather than to yellow.

186 PHYSICAL INVESTIGATIONS ON DYEING.

Article IX. — On the connection existing between my Obser-
vations and those previously made by Natural Philosophers.

63. Buffon* was the first who described, under the name
of accidental colours, several phenomena of vision, which,
according to his explanation, are all so far analogous, that
they result from too great vibration, or from fatigue of the eye,
differing in this respect from the colours under which those
bodies appear, which are coloured in a constant manner,
whether they decompose light by acting upon it by reflection,
refraction, or inflection.

64. Accidental colours may arise from different causes;
for example, they may be perceived under the following cir-
cumstances.

1. When the eye is compressed in the dark:

2. In consequence of a blow on the eye:

3. When the eyes are closed after having been for a
moment fixed upon the sun:

4. When the eyes are fixed upon a small square piece
of coloured paper laid on a white ground; if the square be
red it will appear bordered by a faint green ; if yellow it
will be bordered by blue; if green by a purplish white;
if blue by a reddish white; and if black by a vivid white:

5. If, after having observed the preceding phenomena
for a considerable time, we turn our eyes to the white
ground in such a manner as no longer to see the small
square of coloured paper, we shall perceive a square of an
extent equal to the former, and of the same colour as that
which bordered the small square in the preceding expe-
riment (4).

65. I could cite many other instances in which accidental
colours are produced, if I did not fear departing too far from the
principal object I had in view in my memoir, which was to
give an exposition of the law regulating the modifications mutually
experienced by differently coloured bodies in juxtaposition and when
seen simultaneously; before proceeding further, I must, however,
direct particular attention to the distinction of the two cir-

* See Memoir es de VAcadtmie des Sciences, 1743.

PHYSICAL INVESTIGATIONS ON DYEING. 187

cumstances, 4 and 5, in which Buffon observed accidental
colours.

66. I trust that I shall be able to prove, by means of the
details into which I am about to enter, that, owing to this
want of attention, one of the subjects of optics admitting of
the most comprehensive application has not generally been
treated with the precision and clearness necessary to exhibit
its importance to those who, without having made any obser-
vations upon the subject, have confined themselves to a mere
perusal of what has been written upon it. This distinction is
further necessary for the appreciation of the new facts that
my researches have added to the history of vision: I will
designate by the term simultaneous contrast the modification
of colour and height of tone experienced by two differently
coloured objects when seen simultaneously; by way of oppo-
sition, I will apply the term successive contrast to the phenomena
observed when the eyes, after having looked for a certain time
at one or more coloured objects, perceive images of a colour
complementary to that which appertains to each of these
objects. It is doubtless superfluous to remark that the fourth
instance above-mentioned appertains to a simultaneous contrast,
whilst the fifth refers to a successive contrast. I shall notice
the principal works and experiments instituted with reference
to accidental colours.

67. Father Scherffer, in 1754, laid much stress upon the
phenomena appertaining to a successive contrast, demonstrating
that a given colour produced an accidental colour, that, namely,
which we now call its complementary colour, and by this law
he rectified several observations of Buffon. Not content with
tliis, he sought to explain the cause of the phenomenon, in the
manner I shall explain in the following article. He only
slightly touched upon a simultaneous contrast. (See his memoir,
§ xv., Journal ale Physique, t. xxvi.)

68. ^Epinus* and Darwinf also turned their attention to
a successive contrast.

• Memoires de VAcademie de Pttersbourg et Journal de Physique, annie 1785,
t. xxvi., p. 291.

t Philosophical Transactions, vol. lxxvi. 1785.

188 PHYSICAL INVESTIGATIONS ON DYEING.

69. Count Rumford* made simultaneous contrast an object
of experiment and observation (3), and to these researches I
must refer, since they are more nearly analogous with my own
than any other that have been made upon the subject. It was
demonstrated by Count Rumford, after having observed that
a shadow in a ray of coloured light, (illumined by a ray of
white light of equal intensity with the former,) appeared tinged
with the complementary colour of the coloured ray, when it was
near a shadow of equal size produced in the white ray illumined
by the red ray:

(1.) That the result is the same when the ray of
coloured light is replaced by light transmitted through a
glass, or any other coloured medium, or by the coloured
light reflected by an opaque coloured body.

(2.) That if, in a circle of white paper placed upon a
large sheet of paper, lying on the floor of a room, two bands
of paper 6 lines in width and 2 inches in length be laid side
by side, one being covered with a powder of the colour a,
whilst the other is covered with a powder composed of white-
lead and lamp-black in such a proportion that the light re-
flected from this powder is equal in intensity to the coloured
light of A, a person looking with one eye through his hand
at these two stripes will see the one covered with gray
powder tinged by the complementary colour of a, which
will be as brilliant as A itself.

The author remarks that, in order to prosecute this expe-
riment successfully, it is necessary to take many precautions,
not only to avoid the light reflected from neighbouring objects,
but also to procure a gray capable of reflecting a light equal in
intensity to the coloured light. He observes that the difficulties
are very great if colours be taken that have been ground in oil,
owing to the latter giving a brown tinge to these colours, which
never retain the purity of the colours of the spectrum.

70. If it be true that the experiments of Rumford cor-
respond to those I have made upon colours in connection with
black and white, and that they afford a particular instance

* Experiment upon Coloured Shadows; Conjectures on the Principles of
the Harmony of Colours. Stamford's Philosoph. Papers, &c, vol. i. London,
1802. ^ ' ' '

PHYSICAL INVESTIGATIONS ON DYEING. 189

of the law of contrast as I have established it, it is not less
so that the law cannot be educed without making the series of
experiments which I have prosecuted. For as the experiments
of Rumford comprised the maximum of the phenomenon, it
could not be affirmed that, under ordinary circumstances, there
would be not only a modification of white and black by colours
in juxtaposition, but also a modification of the latter. We
have actually seen that colours brought in contact with white
become deeper, and that they grow fainter when in contact
with black, the contrast as I have demonstrated embracing
both colour and the height of tone in the colour.

71. Struck by observing in his experiments, that a coloured
ray developed its complementary colour, Rumford laid it down
as a principle, that two colours to be in harmony must both present
the respective proportions of the coloured light necessary to form
white. And he therefore recommended ribbons destined for
the toilet of ladies, and the colours in furniture, to be assorted
in accordance with this law. He likewise thinks that painters
may derive much benefit from an acquaintance with the prin-
ciple; it is evident, however, that this law of the harmony of
colours proposed by Rumford is nothing more than an ingenious
invention of fancy, and that as he laid it down, it could not
very easily be made to throw light upon the practice of paint-
ing. I shall, however, revert to this point in treating of the
applications of my own labours ; in the meanwhile I would draw
attention to the fact, that Rumford has not made any experi-
ment demonstrating the influence of two colours in juxtaposition,
or more generally, of two colours seen simultaneously.

72. M. Prieur, of the Cote-d'Or, is the author that has
treated most recently of accidental colours*. He has turned
his attention to these phenomena, under the name of contrasts,
which exclusively refer to simultaneous contrast; for instance,
a small stripe of orange paper appears red when laid upon
a yellow ground, whilst it will appear yellow upon a red
ground: according to the principle laid down by the author,
the accidental colour of the stripe must be that resulting
from its own colour, without abstracting from it that of the
ground. It seems, remarks he, that a certain fatigue of the eye,

* Annales de Chimic, t. liv. p. 5.

190 PHYSICAL INVESTIGATIONS ON DYEING.

whether produced instantaneously by the intensity of the light, or
more slowly by the prolonged vision, concurs in producing these
appearances. He admits, however, that excessive fatigue in the
organ would occasion a degeneration of the colours belonging to
another scale. And he finally adds, that the colours termed acci-
dental by Buffon, and on the subject of which Scherffer has given
an interesting memoir, belong to the class of contrasts, or at least
constantly follow the same law. It is evident that M. Prieur has
not made the distinction of the two kinds of contrasts that I
have established above.

73. Hauy presented, in his Traite de Physique, a resume
of the observations of Buffon, Scherffer, Rumford, and M.
Prieur ; but, notwithstanding the general clearness of style of
the illustrious founder of crystallography, there is an obscurity
in the absence of the preceding distinction ; and this is especi-
ally observable in his account of the explanations previously
given of these phenomena.

74. According to what has been said, we see,

1. That those authors that have treated of a contrast
of colours, have described two kinds of phenomena, with-
out distinguishing the one from the other:

2. That Scherffer has given the law of successive
contrast :

3. That Count Rumford has given the law of the mo-
dification experienced in a particular case by a gray band,
placed in juxtaposition with a coloured one:

4. That Scherffer first, and subsequently M. Prieur, of
the Cote-d' Or, with more exactitude, have given the law of
the modification experienced by a small extent of white or
coloured surface from the different colour of the ground on
which it is laid.

75. If, on the one hand, it be true, that in this case we
perceive, in the most striking manner, the modification which
the colour of the small extent of surface is susceptible of re-
ceiving from that of the ground, we cannot, on the other,
appreciate the modification experienced in the colour of the
ground by that of the small surface, owing to our seeing only
half of the phenomenon, and we should be greatly in error
were we led to think, that a coloured object cannot be modified

PHYSICAL INVESTIGATIONS ON DYEING. 191

by the colour of another, unless the latter be of infinitely
larger extent than the former. The manner in which I have
disposed coloured objects in my observations on simultaneous
contrast, has enabled me to demonstrate:

1. That it is not indispensably necessary to the modifi-
cation of the colour of one object by that of another, that
the former should be of greater extent than the latter, since
my observations have been made upon equal and merely
contiguous bands:

2. That one may perfectly judge of the modifications
experienced by contiguous bands, in comparing them to
those which are not in juxtaposition, a circumstance which
enables us to see the phenomenon of simultaneous contrast
in the most perfect manner, and to establish its general law :

3. That in increasing the number of bands not in con-
tact, or which are placed on either side of those which touch
each other, and on standing at a suitable distance for the
eye to embrace the two series of bands, we see that the
influence of one band is not limited to the next band with
which it is in contact, but extended to the second, third,
&c, although with continually increasing faintness. Now
this influence at a distance ought to be noted, in order to have
a just idea of the generality of the phenomenon.

Article X. — On the Physiological Cause by which the Contrast
of Colours may be explained.

76. ScheriFer has advanced a physiological explanation of
the successive contrast of colours, which seems satisfactory. It
is based upon this proposition, that if a double impression, of
which one is vivid and strong, and the other weak, be produced
upon one of the senses, we shall perceive the stronger of the two.
This occurs principally when both are of the same kind, or when
the powerful action of an object on one of the senses IS FOLLOWED
by another of the same nature, but infinitely weaker or less violent.
We will now proceed to apply this principle to the explanation
of the three following experiments on successive contrast.

First Experiment. — We must look for some time at a small
white square placed upon a black ground.

On ceasing to look at this, and turning the eye upon a

192 PHYSICAL INVESTIGATIONS ON DYEING.

black ground, we perceive the image of a square, equal in ex-
tent to the white square, but instead of being lighter than the
ground it will, on the contrary, be darker.

Explanation. — The part of the retina on which the white
light of the square acted at the first part of the experiment, is
more fatigued than the remainder of the retina, which has only
received a faint impression from the faint rays reflected by the
black ground ; the eye then being fixed upon the black ground
during the latter part of the experiment, the weak light of
this ground acts more strongly upon that part of the retina
which is still unexhausted, than upon that which has already
been fatigued, and hence arises the image of the black square
seen by that portion of the eye.

Second Experiment— W q must look for some time at a
small blue square on a white ground.

Turning the eye away from this, and fixing it on the white
ground, we perceive the image of an orange square.

Explanation. — The part of the retina on which the blue
light of the square has acted, in the first case, being more
fatigued by this colour than the rest of the retina, it happens,
in the latter part of the experiment, that the retina which is
fatigued by the blue, is consequently disposed to receive a
stronger impression from orange, the complementary colour
of blue.

Third Experiment — The eye must be fixed for some time
upon a red square on a yellow ground.

Turning it away, we must then fix it upon a white ground,
when we shall see the image of a green square upon a violet-
blue ground.

Explanation. — At first the part of the retina with which
we see the red is fatigued by that colour, whilst the part with
which we see the yellow is equally fatigued by the latter,
consequently, in the second part of the experiment, the por-
tion of the retina which received the impression of the red now
sees green, the colour complementary to it, whilst the portion
with which we have seen the yellow sees the violet-blue, its
complementary colour.

77. These three experiments, as well as the explanations
referring to them, taken at hazard from the memoir of ScherfFer

PHYSICAL INVESTIGATIONS ON DYEING. 193

from amongst many others analogous to them, will suffice, I
think, to demonstrate that it was actually the phenomenon of
successive contrast which specially occupied that ingenious ob-
server. Considering this, it is certainly strange that Haiiy, in
endeavouring to make known the explanation given by Scherffer,
should have spoken exclusively of the case of a simultaneous
contrast, a phenomenon which the latter naturalist has only
casually mentioned, as I have already remarked at 67 ; for the
rest, the following are the words in which Haiiy expresses him-
self on this subject, taking as an illustration the case in which
a small stripe of white paper is placed upon red paper ; " We
may," says he, " consider the whiteness of this stripe as being
composed of a bluish green and red. But the sensation of the
red colour, acting with much less force than that produced by
the surrounding colour of the same kind, is eclipsed by the
latter, so that the eye is only sensible of the impression of the
green, which being, as it were, foreign to the colour of the
ground, acts upon the organ with all its force."

78. Although this explanation appears to be a natural
consequence of the principle set forth by Scherffer, the latter
does not seem to me to have applied it to the explana-
tion of simultaneous contrast, and the passage quoted above
(67) of his memoir is very clear: this must principally take
place when they (the impressions) are both of the same kind,
or when the powerful action of an object on one sense is folloiced
by another of the same nature, but infinitely more weak, and less
violent.

79. Let us now see what difference exists between the
explanation of successive contrast, such as Scherffer has demon-
strated it, and that attributed to him by Fenay in the case of
simultaneous contrast. All the observations on successive con-
trast explained by Scherffer present this result, that the portion
of the retina, which in the first part of the experiment is
struck by a given colour, sees in the subsequent part of the
observation the complementary colour to the given colour, and
this new impression is independent of the extent of the coloured
object relatively to that of the ground on which it is placed,
or, to speak more generally, of the objects surrounding the
former.

194 PHYSICAL INVESTIGATIONS ON DYEING.

80. This is not expressed in the explanation attributed by
Haiiy to ScherfFer, thus:

1. The portion of the retina that sees the white stripe
placed on a red ground sees it as if it were of a bluish
green, that is to say, the complementary colour to the
ground. Now, according to the experiments of ScherfFer,
this portion, fatigued by white light, has a tendency to see
not a bluish green, but black, which is in some degree com-
plementary to white.

2. In order to admit the explanation attributed to
ScherfFer, it would be necessary that the object, whose
colour is modified by that of another, should in general be
of much smaller extent than the latter, since it is only by
this excess of extent in the modifying body that we can
conceive in general that excess of action which neutralized a
part of that of the first object; I say in general, because
there are cases, in which it might be said, that a much
brighter colour might modify one that was less so, although
it might actually occupy a very small space around it. In
reverting to what has been said, we shall perceive the differ-
ence between the explanation given by ScherfFer of suc-
cessive contrast and that attributed to him on simultaneous
contrast.

81. If we revert to this last explanation in order to test its
force, not under the circumstances related by these authors of a
small stripe appearing to be alone modified when seen on a
ground, but in the case of two stripes of equal extent being
mutually modified, not only when in contact, but at a distance
as shown by my observations; we shall readily appreciate the
difficulty that presents itself, thus :
Fig. 3. 1. Let us suppose that figure 3 represents the

_^^ image of a red stripe r contiguous to a blue stripe

b, depicted on the retina ; the former will acquire
yellow or lose blue, and the second will acquire
yellow or lose red. Now, it is the portion of the
retina on which the image of the stripe r is im-
pressed that will lose its sensibility for red, as it
is the part of the retina on which the image of
the stripe b is impressed that will lose its sensibility for

PHYSICAL INVESTIGATIONS ON DYEING. 195

blue; consequently, I do not perceive how it can be the part
r, which in reality loses its sensibility for blue, or how it
can be the part b, which loses its sensibility for red.

2. In my observations, where the stripes are of equal
extent, there seems no reason, as in the case of a small
stripe being laid upon a ground of great extent, that one of
the stripes should modify the other by the great fatigue each
one occasions the retina.

82. It was doubtlessly owing to the difficulties presented
by the explanation we have been considering, that the illustrious
author of the Mecanique Celeste was led to propose another,
which Hauy inserted in his Traite cle Physique at the end of
the one attributed by him to Scherffer: the case is still that of
a little stripe of white paper placed upon a red ground. The
illustrious geometrician supposes, as Hauy observes, "that there
exists in the eye a certain disposition, by means of which the
red rays comprised in the whiteness of the small stripe are
attracted, as it were, at the moment they reach this organ, by
the rays forming the predominating red colour of the ground,
so that the two impressions form only one, and that of the
green colour is enabled to act as if it were alone. According
to this method of understanding the subject, the sensation of
the red decomposes that of the whiteness, and whilst the actions
of the homogeneous rays unite together, the action of the hete-
rogeneous rays being disengaged from this combination produces
its separate effect."

83. I will not contest the truth of this explanation beyond
making the remark, that it admits by implication the necessity
of the modifying colour occupying a larger extent of surface
than the colour modified; it is probable, however, that such an
opinion would not have been advanced if the illustrious author
had been acquainted with the true explanation of Scherffer on
successive contrast, and if instead of citing a single experiment
of simultaneous contrast, which does not include more than half
the phenomenon, one had been laid before him in which differ-
ently coloured stripes of equal extent were seen to modify each
other, even when not in contact.

84. After having shown the insufficiency of the explana-
tions that have been given of simultaneous contrast, it only

196 PHYSICAL INVESTIGATIONS ON DYEING.

remains for me to speak of the connexion, which appears to me
to exist between the organ of vision and the phenomenon,
observed under the circumstances in which I have studied it.
All the authors who have treated of accidental colours agree in
considering them to be the result of fatigue of the eye; if this
be incontestably true of successive contrast, I do not believe it to
be so with reference to simultaneous contrast, for in arranging
the coloured stripes in the manner I have done, as soon as we
succeed in seeing all four together, the colours may be observed
to be modified before the least fatigue is experienced by the
eye, although I admit that it frequently requires several seconds
before these modifications can be perfectly well recognised.
But is not this time necessary as the same that we give to the
exercise of each of our senses, whenever we wish to give an
exact account of the perception of a sensation affecting them?
There is, besides, a circumstance that explains in many instances
the necessity of time. This is the influence of white light re-
flected by the surface modified, which is sometimes sufficiently
strong to weaken the result of the modification ; and the greater
part of the precautions proposed for seeing the accidental
colours of simultaneous contrast are therefore directed towards
the diminution of the influence of this white light. It is further
owing to this cause that gray and black surfaces contiguous to
the surfaces of very bright light colours, as blue, red, and
yellow are modified more than a white surface would be by
their vicinity. The following is an instance accidentally pre-
sented to my observation, and which will afford a good illustra-
tion of my idea. A coloured paper, the writing on which had
been traced with a pale gray substance, was presented to me
one evening at twilight ; on first looking at it, I could not trace
a single letter, but at the end of a few minutes I contrived to
read the writing which appeared to me to have been traced with
an ink of a colour complementary to that of the ground.
Now, I would ask, if at the moment when my vision was per-
fect my eyes were more fatigued than when I first looked at the
paper without being able to distinguish the letters upon it, and
see them of the colour complementary to that of the ground ?

85. I conclude definitely from my observations, that when-
ever the eye simultaneously sees two differently coloured objects,

PHYSICAL INVESTIGATIONS ON DYEING. 197

the analogous character of the sensation of the two colours
experiences such a diminution, that the difference existing
between them is rendered proportionately more sensible in the
simultaneous impression of the two colours upon the retina.

II. Applications.

86. The preceding observations, and the announcement of
the law, which by its generalizing character simplifies them so
much, will suffice to enable the reader to make all the applica-
tions of which they are susceptible, whether in assorting
coloured objects for the purpose of producing a definite effect
or in judging of the colours of dyed stuffs relatively to their
effect upon the eye without regard to their permanency.
Among the few examples to which I must limit myself, I will
cite the following :

Article I. — Application to the Art of Weaving coloured Stuffs.

First Application. — The assortment of coloured threads to
imitate the colours of a painting.

First Example.

87. A painter delineates two coloured stripes in a picture,
one being red and the other blue ; they are in contact, and,
consequently, the phenomenon of the contrast of two colours in
juxtaposition would occur, if the painter had not sustained the
red contiguous to the blue stripe by blue, and if he had not sus-
tained the blue stripe by placing red or violet near the red stripe.

88. Suppose a weaver desires to imitate the two stripes of
which we have spoken ; if he be ignorant of the law of the
contrast of colours, he will not fail, after choosing wools and
silks suitable to the model before him, to make two stripes pre-
senting the phenomenon of contrast, owing to his having
selected his wools and silks of only one blue and one red, with
the view of imitating the stripes, which the painter has only
succeeded in producing from two colours, (each of which the
eye regards as homogeneous through its whole extent,) by
means of an artifice, which no one would detect who was igno-
rant of the law of contrast.

198

PHYSICAL INVESTIGATIONS ON DYEING.

89. Suppose that the painter, instead of painting the two
stripes in colours that appeared to the eye to be homogeneous,
were to form them with a red and blue, that were not blended

on their contiguous bor-
Fig. 4. ders, the colours would

certainly in that case con-
trast. If now the weaver
be ignorant of the law
of this phenomenon, he
would not fail, in his at-
tempt to imitate his mo-
del, to blend yellow or
orange with the red, and
yellow or green with the
blue in the parts of the
stripes that come in con-
tact, and the result would
be, that the contrast
would be more or less
exaggerated, from the
effect of the picture hav-
ing been imitated by
working with homogene-
ous colours.

Second Example.

90. Let a paper, a b,
fig. 4, divided into 10
equal zones, be first
painted with an uniform
tone of any colour, for
instance, with an uniform
tone of Indian ink; let
the zones 2, 3, 4, 5, 6, 7,
8, 9, 10, receive a second
wash of the same uniform
tone; let the zones 3, 4,
5, 6, 7, 8, 9, and 10, re-
ceive a third, and so on

1—1

"?
e

"55

£~

■b

»o

■*

"«

~5S

"«

•<&

"C5

r— 1

PHYSICAL INVESTIGATIONS ON DYEING. 199

until 10 zones be procured, which gradually increase in depth
of tone, proceeding from the first onward. The remarkable
part of the phenomenon here is, that each zone will present at
least two shades, owing to the contrast produced by contiguity ;
for instance, in beginning from the first, the border b b of this
zone, contiguous to the border a' a' of zone 2, will appear lighter
than the border a a; and, consequently, two shades will be pre-
sented in zone 1, and the same in the others. But it is possible
that a larger number may be distinguished, especially in the
intermediate zones between 2 .... 9; provided they are of
sufficient breadth, and this is owing to the borders a ci a!' a" —
being lighter, and the borders b' bf b" b" — darker than the general
tone of the zone, when by reason of contrast, the middle of the
zones, being less affected than the borders, will present a third
tone of colour. It will be evident that the three tones, or the
two tones, as the case may be, presented by the zone, will not
terminate abruptly, but blend into one another.

91. Suppose a weaver were to copy the figure we have
composed, if he were unacquainted with the effect of the con-
trast of contiguous zones, he would exaggerate the effect in his
work, using probably at least twenty shades of the same colour
instead of the ten.

92. These examples convey a better idea than all the rea-
soning we could enter upon, of the fundamental difference
existing between the employment of colours in painting and in
weaving. A painter may, to a certain extent, succeed without
a knowledge of the law of simultaneous contrast; for if he
produce an effect with which he is not satisfied, depending
upon this phenomenon, his palette furnishes him with the
means of destroying or increasing it. The weaver has not the
same resources at his command ; he cannot replace his colours
without undoing his work^ and to avoid this necessity he ought
to be sufficiently well acquainted with the law of contrast, to
judge of the effect that will be produced by the vicinity of other
colours upon the coloured threads he employs in the imitation of
a certain part of his pattern, otherwise it will be impossible for
him to reproduce a faithful copy of the painting before him.

200 PHYSICAL INVESTIGATIONS ON DYEING.

Second Application. — To work on black with blue and violet.

93. The observations made at Article V. § 1, teach the
weaver that there are colours, such as yellow and orange,
which may be worked in with black without weakening its
intensity, whilst there are others, as blue, indigo, and violet,
which produce the contrary effect, by giving it somewhat of a
yellow, green, or olive tint. It is consequently necessary to
avoid making blue or violet contrast with black, when shadows
are to be produced in blue or violet stuffs, and the object re-
quired may be obtained by various means, as I am about to
indicate.

Article II. — Application to the Art of Printing Patterns upon
Coloured Stuffs or Painted Papers, and Coloured Inks upon
Coloured Paper.

94. In these applications, I exclude the consideration of
the chemical action that may exist between the substance of
the coloured stuffs or the painted paper, and that of the colour-
ing matter used, confining myself to the optical effect.

95. I have frequently been appealed to as an arbiter in
cases where persons having given to be printed various woollen
stuffs for furniture and ladies' cloaks, have had disputes with
the printer on the subject of the patterns, which were not of
the colour intended. I have often found that these complaints
depended upon the effect of the contrast of the colour of the
designs with that of the ground, and that if the printer were
reprehensible, it was not for having printed a different colour
from the one required, but for not having foreseen the effect
that would result from the contrast of colours, one of which was
to serve as a ground for another.

Examples.

1. When black patterns are printed upon red, crimson,
or amaranth grounds they appear green, owing to green, the
complementary to the ground, blending with the black (48).

2. For the same reason, black, when printed on violet
stuffs, or on dark green, loses all its force.

96. The method I have adopted in the cases in which I

PHYSICAL INVESTIGATIONS ON DYEING. 201

have been appealed to as an arbiter, was, by a simple process,
to prove that the designs which did not appear black were
actually so, and it consisted in cutting a piece of white paper
in such a manner as to cover the ground and only show the
pattern.

97. Similar difficulties have presented themselves in manu-
factories of paper-hangings, when it was required to produce
a design of a slightly yellowish-gray upon a green ground.
These designs, although actually gray, appeared, when once
printed, to be pink, owing to the complementary colour of the
ground. If they had been on a rose-coloured ground they
would have appeared green, for the same reason.

98. Finally, in wishing to impress letters upon coloured
paper, the rule to be followed is to choose a ground that is
of the colour complementary to that of the ink. Thus, we
must use violet-coloured ink for a yellow paper, and yellow
ink on a violet-coloured paper; red ink on green paper, and
green ink on red paper; orange-coloured ink on blue paper,
and blue ink on orange-coloured paper.

Article III. — Application to the Assortment of Stuffs with the
Woods employed in Upholstery.

99. Amongst the numerous applications of the law of
simultaneous contrast to the art of the upholsterer and house
decorator, there is one to which I shall call attention, since it is
too often disregarded, I allude to the assortment of stuffs with
the fancy woods used in upholstery, for making easy chairs,
sofas, &c. The principle to be followed in making the most
effective assortment is to choose violet or blue stuffs for yellow
icoods, as orange-wood, the root of the ash, Sfc. ; and green or yellow
stuffs for red woods, like mahogany. The colour of the stuff
must be as different as possible from that of the wood.

100. Many persons prefer, even for mahogany, to have a
crimson stuff, owing to its being one of the best colours for
standing light. Although this assortment is contrary to prin-
ciple, the bad effect admits of being diminished by placing a
broad green or black border, either a cord or a printed band,
between the crimson and the mahogany. The upholsterer

202 PHYSICAL INVESTIGATIONS ON DYEING.

often borders crimson either with a gold cord or band, fastened
on with gold-headed nails, or with yellow silk cord; and these
borders, although not complementary, have, at any rate, the
advantage of pleasing many by their brilliancy. There is one
assortment, however, which ought never to be made, namely,
that of yellowish red, as scarlet, flame-coloured, or light-red
stuffs with mahogany ; since their brightness is such that they
deprive the wood of that red colour which constitutes its
beauty, and makes it resemble oak or walnut.

Article IV. — Applications to the Painting of Pictures in gene-
ral, and to the Patterns of Woven Stuffs, and of Carpets in
particular.

101. There are, as is well known, two systems of painting,
that of plain tinting and of chiaro-obscuro. In the former, the
colours are not shaded, nor blended in one another, nor modified
by reflections, as in the second. Perspective is reduced in
these kind of paintings to an observation of mere linear per-
spective, and to the employment of vivid colours for the first
coat, and pale or gray colours for the last. If the choice of
the contiguous colours have been made in conformity with the
law of simultaneous contrast, the effect of the colouring will
be more striking than where the painting has been done
according to the system of chiaro-obscuro. In admiring the
beauty of the colours of those plain-tinted paintings, which we
receive from China, we ought, in comparing them with our
own, to take into account the system by which they have been
done, otherwise we might form an erroneous judgment in com-
paring pictures that have been painted according to such
different systems.

102. If the system of plain tinting is more favourable to
show the brightness of the colours used, it ought to be a
motive for the painter who follows the system of chiaro-obscuro
to make himself thoroughly conversant with the resources held
out to him by the application of the law of simultaneous
contrast. He ought to multiply his draperies of different
colours as much as his subject allows of it, and if he cannot,
he ought to bring out his colours by well-chosen and skilfully-

PHYSICAL INVESTIGATIONS ON DYEING. 203

handled reflexions : thus, for instance, green or yellow reflex-
ions bring out a blue drapery; yellow or orange reflexions
bring out a purple and violet drapery; purple reflexions bring
out a scarlet drapery. He will give effect to the freshness
of flesh colours by avoiding the contact of yellow or orange
draperies, as well as grounds of the same colours. He must
likewise avoid the fault common to decorative painters of using
pink or a light amaranth for the hangings of boxes in a theatre,
since these colours have the serious disadvantage of giving a
greenish tinge to the complexion.

103. Before entering upon the qualifications which ought
to be possessed by the patterns for hangings and carpets, in
order to insure the copies approaching as nearly as possible to
perfection, we must treat of the special points appertaining to
this species of labour. The weaver imitates objects with
coloured threads of a certain diameter. These threads passed
round the threads of the chain of the warp exhibit not a smooth
surface, but one that is furrowed with lines, some running
parallel with the warp and deeper than the others which are
perpendicular to them; the effect of these furrows is the same
that would be produced in a picture by a system of obscure
lines intersecting each other at right angles. There are, there-
fore, the following differences between a piece of weaving and a
picture: — 1st. That the former never presents the same fusion
of colours that the painter can so easily obtain by blending or
dividing his colours by means of a more or less viscid solvent.
2nd. That the symmetry and uniformity of the furrows in the
weaving are unfavourable to the production of these brilliant
lights and effective shadows exhibited by the painter, for while
the furrows darken the lights, the salient parts of the threads,
which are in the shadows, have the disadvantage of weakening
them by the light which they reflect. Add to this, that the
painter has other resources denied to the weaver for augmenting
the brilliancy of his lights and adding force to his shadows: as,
for instance, by the opposition of opaque and glazed colours.
He modifies an object of an uniform colour by varying the
thickness of the coat of paint laid on the canvas, as well as by
the direction in which he lays on the brush. If it were possible
to compose a piece of mosaic with such loose elements as tin

204 PHYSICAL INVESTIGATIONS ON DYEING.

threads used in the manufacturing of the Gobelin tapestry, a
work of the kind would rank between an oil painting and
tapestry, resembling the latter, owing to it having resulted
from the juxtaposition of coloured elements of an appreciable
extent of surface, and approximating to the picture by its
smooth surface, which is rendered brilliant by polish; besides
the opposition of opaque and transparent elements calls to
mind the contrast presented in a picture by the use of opaque
and glazed colours.

104. From this state of things, I conclude that, in order
to raise tapestry as nearly as possible to painting, it is neces-
sary, 1. That it should represent objects of such a size, that the
point where the spectator must stand in order to see them well,
should not admit of his distinguishing the coloured elements
from one another, nor the furrowed lines separating them;
2. That the colours should be as bright as possible, presenting
the strongest contrasts.

105. It is evident, therefore, that the patterns for tapestry
should not only be correct in design, and represent elegant
forms, but they should be composed of large objects, draped
rather than naked, with the dress covered with ornaments
rather than simple and uniform; and, finally, that the colours
should be varied and bright, presenting the strongest possible
contrasts.

106. Although the rules given are applicable in many
points to tapestry for furniture and to the patterns for carpets,
it may not be wholly useless to make a few remarks relatively
to the choice of colours, to the number of tones of the same
scale that may be used, and to the choice of the subjects to be
represented.

107. For the choice of patterns for damasks (tapestry)
for furniture, opposition of the grounds with the predominating
colour of the designs upon them is too often disregarded; for
instance, where a wreath of flowers is to be represented on a
crimson ground, the greater part ought to be composed of blue,
yellow, and white flowers; if red flowers are introduced, they
should border on orange rather than purple; whilst green
leaves laid directly on the ground conduce very considerably
to the beauty of the whole; where the ground is green or

PHYSICAL INVESTIGATIONS OX DYEING. 205

dead leaf, the predominating flowers ought, on the contrary,
to be pink and red.

108. In the pattern for a Savonnerie carpet, the size of the
figures ought to be in proportion with the exent of the whole;
attention must also be paid to the pattern on the parts con-
cealed by the furniture, so that a figure or any other object
composed of different parts, (distinct with respect to design
and colour,) may be presented to the eye of the spectator in
all its details when the carpet is laid down. In a large carpet,
or in one of only middling size, the figures and designs should
not be too large, nor should the divisions of one uniform colour
cover too large a surface. In a carpet of small extent the
designs of the Persian and Turkey carpets are preferable to
those which look more like a picture; and here we may as
well draw attention to the effect produced by palms, straight or
waving zones, Vandyck scallops, &c, in which yellow is con-
trasted with violet, orange with blue, and red with green.

109. Finally, in the case of patterns for tapestry and
carpets, where no attempt is made to compete wTith the effect
produced in tapestry or furniture-hangings of the Beauvais
style, or in carpets of the Savonnerie manufacture, a scale
composed of only four or five tones of colour will be sufficient
to give a very good effect, provided the choice of the conti-
guous colours has been judiciously made.

Article V. — Applications to Coloured Glass in Gothic

Churches,

110. We will now consider the preceding reflections with
reference to coloured glass, which combines so powerfully with
the architecture in imparting to large Gothic churches the
harmony that cannot fail to strike every one who enters these
vast edifices, after having admired the boldness and variety of
their external decorations; and which contributes to place
these erections foremost among works of art, alike by their
grandeur, the subordination of their different parts, and their
intimate relation to the purposes, for which they were intended.
The coloured glass of Gothic churches produces a strikingly
brilliant effect by intercepting the white light, and transmitting

206 PHYSICAL INVESTIGATIONS ON DYEING.

coloured rays, which are less glaring, and more in accordance
with the character of the place. If we were to seek for the
cause of this effect we should find that it depended not only
upon the happy contrast of colours displayed, but also upon
the contrast of their transparency with the opacity of the walls
around, and of the leaden frames enclosing them. The impres-
sions produced upon the eye, owing to this double cause increase
in intensity the oftener and the longer they are examined.

111. The windows in a Gothic church are generally either
circular, or pointed towards the top and with vertical sides.
The panes in the former usually represent large medallions
(roses), in which the yellow, violet, blue, orange, red, and
green, seem to sparkle as if from the finest precious stones.
The panes of the latter almost always represent, in the centre
of a border, or on a ground corresponding with the medallion
(rose) windows, a figure of a Saint perfectly in harmony with
the figures in alto-relief round the entrances of the church;
and these centre figures ought to be judged as parts of a whole,
and not like a Greek statue intended to be seen isolated from
every side.

112. The glass used in composing the different parts of a
human figure is of two kinds; either painted on the surface with
colours that have subsequently been glazed; or has the colour-
ing matter fused in with it; the former is generally used in the
composition of the naked parts of the figure, as the face, hands,
and feet, whilst the latter composes the drapery; the various
pieces of glass are joined together by a leaden frame. I have
been most particularly struck in the best specimen of these
panes representing the human figure, by the accurate obser-
vance of the proper relation of the size of the figures and the
intensity of the light to which they are exposed, to the distance
at which the spectator is placed, and which is so great that the
leaden plates enclosing each square of glass appear like a
narrow black circle, or a mere line.

113. It is not necessary to the effect of the whole, that
painted glass should present, when seen close, delicate touches,
careful lining, or much fusion in the colours, for it is intended
to be combined with the coloured glass of the draperies in
producing a whole, on the system of plain tinting, and it

PHYSICAL INVEST TGATIONS ON DYEING. 207

cannot be doubted that a painting on glass, executed in exact
accordance with the rules of chiaro-obscuro, would, to say-
nothing of the expense of the execution of the work, have this
disadvantage, that, the finish of the detail would be wholly lost
at the distance of the spectator, and that the whole would
appear less distinct : the first condition to be fulfilled by every
object of art intended to attract the eye is, that it should appear as
distinct and as clear as possible. Add to this, that paintings
on glass done according to the rules of chiaro-obscuro, are never
represented with those borders or grounds of rose or medallion-
formed panes (111)? which produce such a beautiful effect;
that they are less brilliant and less clear than the glass, in
which the colouring matter has been incorporated by means
of fusion by heat (112), and, finally, that they are less able to
resist the effect of time and weather.

1 14. Variety in the colours of the panes is so necessary to
the production of the best possible effect, that those which
represent entirely naked figures, edifices, or in fine, any objects
of extent of one uniform colour, are very far inferior, let the
execution be ever so perfect with respect to finish and truth of
invention, to panes composed of pieces of different colours
contrasting well with one another.

115. I infer, therefore, that the causes of the good effect
of coloured panes of glass in large Gothic churches, must be
referred :

1. To their presenting us with a very simple design,
whose different and well circumscribed parts may be seen
without confusion at a great distance.

2. To their composition of one totally formed of
coloured parts, distributed with a kind of symmetry, and
being at. the same time forcibly contrasted, not only amongst
one another, but also with the opaque parts circumscribing
them.

Article VI. — Application to the Distribution of Flowers in a

Garden.

116. Amongst the pleasures presented to us by the culture
of flowering plants, there are few that exceed what we expe-
rience from the sight of a multitude of flowers varying in

208 PHYSICAL INVESTIGATIONS ON DYEING.

their colour, form, and size, and in their arrangement upon the
stem that supports them, It is probably owing to the admira-
tion bestowed individually upon each, and to the attachment
bestowed upon them in consequence of the great care they have
required, that care has hitherto not been taken to arrange
them in such a manner as to produce the best possible effect
upon the eye, not only separately, but collectively. Nothing,
therefore, is more common than a defect of proportion observed
in the manner in which flowers of the same colour are made to
recur in a garden. At one time the eye sees nothing but blue
or white, at another it is dazzled by yellow scattered around in
profusion; the evil effect of a predominating colour may be
further augmented, when the flowers are of approximating, but
still different shades of colour. For instance, in the spring we
meet with the jonquil of a brilliant yellow, side by side with
the pale yellow of the narcissus: in the autumn, the Indian
pink may be seen next to the China rose and the aster, and
dahlias of different red grouped together, &c. Approximations
like these produce upon the eye of a person, accustomed to
judge of the effects of the contrast of colours, sensations, that
are quite as disagreeable as those experienced by the ear of the
musician, when struck by discordant sounds.

117. The principal rule to be observed in the arrangement
of flowers, is to place the blue next to the orange, and the
violet next to the yellow, whilst red and pink flowers are never
seen to greater advantage than when surrounded by verdure,
and by white flowers; the latter may also be advantageously
dispersed among groups formed of blue and orange and of
violet and yellow flowers. For, although a clump of white
flowers may produce but little effect when seen apart, it cannot
be denied that the same flowers must be considered as indispens-
able to the adornment of a garden when they are seen suitably
distributed amongst groups of flowers, whose colours have been
assorted according to the law of contrast; it will be observed
by those, who may be desirous of putting in practice the
precepts we have been inculcating, that there are periods of the
horticultural year, when white flowers are not sufficiently mul-
tiplied by cultivation to enable us to derive the greatest possible
advantage from the flora of our gardens. I will further add,

PHYSICAL INVESTIGATIONS ON DYEING. 209

that plants, whose flowers are to produce a contrast, should be
of the same size, and that in many cases the colour of the sand
or gravel composing the ground of the walks or beds of a
garden, may be made to conduce to the general effect.

118. In laying down the preceding rules, I do not pretend
to assert, that an arrangement of colours, different from those
mentioned, may not please the eye; but I mean to say, that in
adhering to them, we may always be certain of producing
assemblages of colour conformable to good taste, whilst we
should not be equally sure of success in making other arrange-
ments. I shall, however, revert to this point.

119. I will reserve for a special article, the consideration
of the number of plants in flower at the same time, which
admit of being grouped together, and of those details of execu-
tion which would here be out of place. I must, however, reply
to the objection that might be made, that the green of the leaves,
which serves, as it were, for a ground for the flowers, destroys the
effect of the contrast of the latter. Such, however, is not the
case, and to prove this, it is only necessary to fix on a screen
of green silk two kinds of flowers conformably to the arrange-
ment of the coloured stripes (Fig. 1 and 2), and to look at
them at the distance of some ten paces. This admits of a very
simple explanation, for as soon as the eye distinctly and simulta-
neously sees two colours, the attention is so rivetted, that con-
tiguous objects, especially, when on a receding plane, and where
they are of a sombre colour, and present themselves in a con-
fused manner to the sight, produce but a very feeble impression.

Article VII. — Applications to the Colours of Dress.

120. Many persons observe the fact, that a uniform com-
posed of cloths of different colours looks well much longer,
although worn, than one of only a single colour, even when the
cloth of the latter is identical with one of those composing the
former. The law of contrast perfectly explains the reason.

121. If we suppose a uniform to be made of cloth of two
colours, the one complementary of the other, as red and green,
orange and blue, yellow and violet, we shall find that the effect
will be most excellent from their mutually heightening one

210 PHYSICAL INVESTIGATIONS ON DYEING.

another; and supposing, further, that they are of equal stabi-
lity, they will present greater advantages, and appear good in
spite of atmospheric agents, longer than any other binary com-
bination of colours.

122. In a blue and yellow uniform, the blue gives to the
yellow an orange tint, which greatly heightens its effect not-
withstanding its tendency, as a dark colour, to make another
colour appear lighter; the yellow imparts in its turn a violet
tinge to the blue, which considerably improves its appearance,
and if the blue had an unpleasant: greenish tinge, it would be
neutralized by the yellow.

123. On the other hand, stains will always be less visible
on a dress of different colours than on one composed only of a
single colour, since there exists in general a greater contrast
among the various parts of the first-named dress than between
the stain and the adjacent parts, this difference renders the
effect of the stain less apparent to the eye.

124. For the same reason a coat, waistcoat, and trowsers of
the same colour can only be worn to advantage together when
all are new; for as soon as one of them loses its freshness, from
having been worn longer than the others, the difference will
increase by contrast. For instance, a pair of new black trow-
sers, worn with a waistcoat of the same colour, which is old
and a little rusty, will make the tinge of the latter appear more
conspicuous, at the same time that the black of the trowsers
will appear more brilliant. White, or even light gray trowsers,
would produce a contrary effect. We see from this, how advan-
tageous it is to let soldiers have winter trowsers of a different
colour from that of the clothes which they wear during the rest
of the year; and we can further understand the advantage
there is in wearing white trowsers with a blue, or indeed, gene-
rally speaking, with any dark coloured coat.

Article VIII. — Applications to the opinion we form of the
Colours of Dyed Stuffs, setting aside the question of their
stability.

125. In order to judge of the colour of a stuff, with
respect to its brilliancy, we must remove it from the vicinity of
the colours that might produce an effect of contrast upon it ;

PHYSICAL INVESTIGATIONS ON DYEING. 211

but, on the other hand, as we cannot judge of two patterns of
stuffs of the same colour except by comparing them together,
we must, in order to appreciate them correctly, take into
account the phenomenon of contrast that would be manifested,
if the two patterns were not absolutely identical with respect
to colour and height of tone.

First Application,

126. In the case of two patterns, appertaining to the
same colour, as two blues or two reds ; if the blue or the red
of the patterns compared together, be not identical in the two,
we must take into consideration the degree to which the effect
of contrast might exaggerate the difference ; thus, if one blue
be greenish, it will make the other appear less green or more
indigo, and even more violet than it really is, and reciprocally,
the first will appear more green than it would do when seen
separately. The same holds good for reds; if one is more
orange than the other, the latter will appear more purple, and
the former more orange than they really are.

Second Application.

127. Since the contrast of colours that are not analogous,
tends reciprocally to embellish and purify them, it is evident
that in wishing to form an accurate opinion of the beauty of
the colours in carpets, tapestry, and painting, we must take
into account the design and the manner in which the colours
are shaded, provided the objects compared together are not the
exact representation of the same subject; in fine, all other
things being the same, identical and unshaded colours, when
disposed in contiguous zones, will appear handsomer than if
each were seen separately upon a ground which would only
produce one single impression upon the eye. Colours arranged
in palms, as in the Cachmere shawls, and designs like those of
Turkey carpets, produce a more striking effect than if they
were shaded, and merged in one another, as is generally the
case in paintings. Consequently, if we wished, for instance,
to compare the amaranth-coloured zone of an oriental Cach-
mere with the differently coloured zones on an amaranth ground
of a French shawl, we must destroy the contrast of the colours

212 PHYSICAL INVESTIGATIONS ON DYEING.

contiguous to the amaranth zone, by concealing them by means
of a gray paper, cut in such a manner as only to show this
zone, provided, of course, that a similarly cut paper has been
laid over the ground, in order to expose the compared parts to
the same influence on the part of the contiguous objects.

128. The same method must be pursued in comparing the
colours of old tapestry with new colours, and for this reason :
time acts very unequally, not only upon the different sorts of
colours applied to stuffs, but likewise upon the tones or shades
of the same scale of colour. Thus the dark shades of certain
colours fade, as, for instance, the violets in general, while the
dark indigo blues, and the dark reds of madder, kermes, and
cochineal, stand fast. In the second place, the light tones of
the same scale of colour fade in a much shorter time than the
darker shades. From hence it follows, that the colours which
have resisted the destructive action of time the longest, being
more isolated from one another, darker, or less fused, appear on
that very account to be more brilliant than if they were other-
wise disposed.

Third Application.

129. On looking at the whole of the tones of most of the
scales of colour used in the manufacture of tapestries and
carpets, the phenomenon of contrast exaggerates the difference
observed between the extreme and middle tones of the same
scale. Thus, for instance, when all the shades of an indigo
blue are seen on silk, the light shades will appear greenish, the
browner violet, and the intermediate ones blue. Now the dif-
ference of the greenish and violet shades in the extremes, is
augmented by the effect of contrast. The same is the case with
the yellow scale, the light shades appearing greenish, and the
dark ones redder than they are in reality.

Fourth Application,

130. The contrast that increases the difference between
black and white, seen simultaneously beside each other, produces
an analogous effect upon the different tones of the same scale.
This may be deduced from the observation made above (90,) on
the subject of the series of 10 zones, which, seen separately,

PHYSICAL INVESTIGATIONS ON DYEING. 213

are of uniform tint, but which cease to appear so when conti-
guous to one another. Further, the same thing may be directly-
shown by the following proof. Suppose that in a series of blue
tones, designated by the numbers 1, 2, 3, 4, 5, 6, &c, be-
ginning with the light shades, we place the tone 2 between 3
and 4, the blue of 2 will be so much weakened as to appear
equal to the tone 1. This result affords a means of judging
more accurately than any other, provided the tones of the scale
are sufficiently numerous, and at equal distance from one
another. In short, if the tone 2, put between 3 and 4, appear
equal to 1, it will follow, that if the tones be equidistant,
3, when put between 4 and 5, will appear equal to 2, and 4
between 5 and 6 will appear equal to 3, and so with the rest.
If the tones too closely approximate to present this result, they
must be successively advanced, not by a single gradation, but
by two or three.

131. This method of judging of the equality of distance
among tones of the same scale of colour is founded upon the
fact, that it is easier to establish an equality of shades than to
estimate the distance separating the different tones of one scale,
when they are observed in the places which they must occupy
according to the respective intensity of their colour.

Resume, and some general considerations.

132. That which essentially distinguishes the observations
I have made in the first part of this memoir from those pre-
viously advanced, is the experimental demonstration that two
zones of different colours and of the same surface mutually
modify each other when seen simultaneously, not only when
in juxtaposition, but even when at a distance from one another;
and I will add, that these modifications likewise take place
on substituting leaves of coloured paper, 20 inches in length and
a foot in width, for the bands that have been used in the pre-
ceding experiments, the leaves that are used in the experiment
being placed at one yard from the contiguous leaves.

133. The law of these modifications being once established
we may foresee the changes experienced by any two given
colours from juxtaposition, when we know the complementary
to each one, and the height of their tone, since the changes

214 PHYSICAL INVESTIGATIONS ON DYEING.

which they experience will result from the complementary to
the one being added to the colour of the other, and since, if
the two colours be not of the same height of tone, the darker
will appear still deeper, as the other will appear lighter than
it really is, supposing always that the latter effect is not des-
troyed by the former.

134. I have drawn attention to the differences existing
between the preceding phenomenon, which I have termed
simultaneous contrast, and the phenomenon I have spoken of
as successive contrast; in the latter, the same part of the retina
which first sees an object under its true colour, sees its image
subsequently in the complementary to the true colour; whilst
in simultaneous contrast, two different parts of the retina
simultaneously see a different object, and see these objects
with modifications of colour and of height of tone which would
not be presented by them, if one were seen separately from the
other.

135. This distinction between the two kinds of contrast
throws much light upon the history of the labours undertaken
by different natural philosophers on accidental colours. Buffon
and Scherffer examined almost exclusively successive contrast;
whilst Rumford and Prieur of the Cote d'Or occupied them-
selves particularly with simultaneous contrast. Hauy, in
speaking of the theoretical Or explicative portion of accidental
colours, has attributed to Scherffer an explanation concerning
simultaneous contrast, which does not give an idea of that
which the latter Natural Philosopher advanced concerning suc-
cessive contrast. Finally, Laplace has proposed an explanation
of simultaneous contrast, which is not more satisfactory than
the preceding one, since it supposes in general that the modify-
ing colour must be of larger extent of surface than the colour
modified.

136. Simultaneous contrast, as I have considered it, is a
phenomenon of much more frequent occurrence than would be
supposed; in speaking of its relation to the organs of vision,
I have not pretended to advance a theory, but I have wished
to express a fact, which appears to me to be general, viz., that
when the eye is struck at once by two colours, which it views
with some degree of attention, the analogous character of these

PHYSICAL INVESTIGATIONS ON DYEING. 215

colours acts less powerfully upon the optic nerve than the
heterogeneous; or, in other words, the eye evinces less sensi-
bility in catching the analogies than the differences of the
colours, and this without our being able, generally speaking, to
say that the organ is fatigued.

137. In the second part of this memoir, I proved, by
examples, the number and variety of the applications of the
law of simultaneous contrast; if some of these might be deduced
from what was known of accidental colours independently of
my observations, it must be admitted that the majority could
not be made without the following propositions having been
previously demonstrated: viz., that two colours seen distinctly
and simultaneously are mutually modified, independently of their
respective extent, even when they are not in contact, and besides
when there is no ground for attributing their modifications to a

fatigue of the eye.

138. After establishing this proposition as a law, and
developing its consequences by application, it remains for me
to explain the opinion I have given on the system of the
harmony of colours advanced by Count Rumford, and which
I treated as an invention of fancy, and not as a demonstrated

fact. It is evident that this observer having remarked that
a white light contiguous to a coloured light is tinged by the
complementary of the latter, derived from that fact his idea
of the harmony of colours, and that he thus implicitly adopted
the opinion of those, who, like Darwin, suppose that the
beauty of contiguous colours depends upon the relative facility
with which the eye distinctly perceives them, and that, con-
sequently, the most beautiful assortments result from opposite
colours. I have devoted too much attention to the ideas
excited by vision, and to the effect of the arts comprised under
its domain, wholly to reject this opinion; but I cannot admit it
as a sole and sufficient principle, on which to explain the pleasant
or unpleasant sensations produced by such an assemblage of
colours, for, in many instances, we cannot deny the influence
exercised by the principle of the association of ideas. What-
ever it may be, let us then examine the evident facts connected
with an assortment of complementary colours and non-com-
plementaries, relatively to the influence which they mutually

216 PHYSICAL INVESTIGATIONS ON DYEING.

exercise in appearing more or less pure, more or less beautiful,
■without having regard to any systematic view.

Assortment of Complementary Colours.

139. That an assortment of complementary colours is never
disagreeable, is a truth that has long been known, but my
observations further demonstrate that, these colours increase in
beauty by their approximation to each other, heightening and
purifying each other, whatever the difference may be between the
different yellow bodies, or the different violet bodies, $*c, brought
into juxtaposition. It is, therefore, owing to the fact of our
certainty as to the agreeableness of the effect produced by the
mutual approximation of complementary colours, not only
when pure, but also when a mixture of black and white has
rendered them grey, that I have recommended their assortment
in preference to every other, for the distribution of flowers
in a garden, for stuffs for furniture, uniforms, and liveries;
besides this, I have further demonstrated and insisted upon
the economy presented by the adoption of these last-named
applications.

Assortment of Non-complementary Simple Colours.

140. I am now about to consider the assortment of non-
complementary colours, and by way of greater precision, I will
begin by speaking of simple colours, or more strictly defined, of
those which approach the nearest to the seven colours called
primitive. I will then treat of the assortment of two mixed
colours, and of that of one simple and one mixed colour. For
the sake of avoiding all unnecessary paraphrase, I will call a
colour mixed, when it is composed of two or more simple
colours.

141. On looking at the arrangements of the coloured
zones, which were made the subject of the 17 observations of
Article 1, §^ (10), it must have been made evident, that the
greater part of their tones which approached as nearly as possible
to the seven primitive colours, increased in beauty as they lost
their brown tinge, and acquired more intensity and brilliancy in the
same manner as has been observed in the assortment of comple-
mentary colours. It teas further remarked, that their optical

PHYSICAL INVESTIGATIONS ON DYEING. 217

composition appeared always, or almost always, to have been more
or less changed. So that we may draw from it the following
practical conclusion. If the non-complementary colours which
approach most nearly to the seven primitive colours, generally
speaking, increase in beauty by their mutual juxtaposition, they
nevertheless appear to the eye to have lost more or less of their
natural character.

142. The following are examples of three remarkable cases
presented to my observation by these colours.

1. Yellow and blue mutually increased in beauty by
juxtaposition, the former in acquiring a golden, and the
latter a violet tinge.

2. The preceding blue lost a part of its beauty in
acquiring a greenish tinge by its juxtaposition with violet.

3. The above-named violet and a dark indigo blue
reciprocally injured each other, the violet, by losing too
much blue, resembled violet faded by the light, whilst
the indigo blue, in losing its red, acquired an unpleasant
greenish tinge.

Assortment of Non-complementary Colours, either in the Com-
binations of one Simple and one Mixed Colour, or of two
Mixed Colours.

143. On placing in juxtaposition colours that have a cer-
tain analogy, as a simple red with a mixed red bordering on
orange or violet; a simple blue with a mixed blue bordering on
violet or green, &c, or two mixed colours, as an orange red
with an amaranth red, a greenish blue with a violet blue, &c,
&c. ; we may observe in the first place, that the two colours
will increase in beauty; in the second, that one only will
increase in beauty; in the third, that both colours will lose a
portion of their beauty. These three cases may be observed in
the arrangement of simple non-complementary colours, exhibit-
ing, however, this difference, that the two latter cases will
present themselves more frequently than in the arrangement of
simple non-complementary colours.

218 PHYSICAL INVESTIGATIONS ON DYEING.

Examples.
First Example,
144. On placing two roses in juxtaposition, one of which
is of a certain graduated tone of amaranth red, and the other of
a graduated tone of scarlet, and which, when seen separately,
are agreeable to the eye, the first will, according to the law,
appear more amaranth or rather more blue, and the latter more
yellow, which amounts to the same thing as if we were to say,
that both lost a portion of red. If these two roses are both at
the point where the addition of more blue or more yellow
would detract from their beauty, it is very evident that a
juxtaposition is injurious to both. If, however, they are
beyond the limits of which I speak, juxtaposition will not
injure their effect, although it augments the difference between
them, provided always that this is not carried beyond the limits
named. In fine, if a rose incline too much to yellow, or too
much to blue to be seen to advantage separately, it will increase
in beauty by being brought into juxtaposition with another rose
that is more yellow or more purple.

Second Example.

145. A yellow green in juxtaposition with a green border-
ing on blue, will lose blue, and will consequently appear more
yellow. There are certain yellow greens on which the con-
tiguity of another green would act favourably, and others again
on which it would have a bad effect. The first-named would
appear more beautiful on being brought near the yellow green
of young leaves, whilst the latter would acquire the yellowish
rusty green seen in certain leaves as they drop from the trees in
autumn.

Third Example.

146. A nasturtium, whose colour is so bright, produces a
disagreeable effect when combined with certain purple poppies,
the colour of which when seen apart is by no means deficient
in beauty. The nasturtium loses its brilliant red, and acquires
a dull orange tinge, whilst the poppy assumes an equally dull
tinge resembling wine lees.

PHYSICAL INVESTIGATIONS ON DYEING. 219

147. We see from these examples:

1. That in proportion to the analogy of the colours,
there is a greater probability that their mutual juxtaposition
will be injurious to the effect of one, if not of both.

2. That, if the law cannot, with a view of pleasing the
eye, prescribe arrangement of non-complementary colours
in as positive a manner as may be done with reference to
the assortment of complementary colours, it arises from the
impossibility in the present day, of designating with any
degree of precision, the innumerable colours of bodies sus-
ceptible of being classed under such arrangements, in con-
sequence of our inability to refer them to invariable types,
as for instance, Xewton's coloured rings. This is the reason
that in treating of the distribution of flowers in gardens
(118), I have only recommended an assortment of flowers
whose colours are complementary, at the same time, that I
admit the existence of many other assortments, productive
of a very agreeable effect. "We see further, from the same
examples, that in making ourselves masters of the law, and
considering well the effect that we may be desirous of pro-
ducing in placing two colours beside each other, we shall
not be guilty of any fault, I do not say with respect to the
harmony of these colours, but with reference to the best
possible disposition of which they are susceptible, for the
regulation and even augmentation of their splendour,
purity, and freshness, as well as for the correction of their
individual deficiencies. For the rest, in order to dissipate
all further obscurity, I will cite the following examples.

First Example.

148. In the case of a nasturtium, which has such a pro-
portion of red and yellow, that a little more red or more yel-
low would impair the beauty of the flower, the law indicates,
1. That the vicinity of blue, whose complementary is orange,
would produce a good effect owing to the blending of the
orange, which has the greatest analogy with the first-named
colour ; some persons object, however, to this assortment as too
harsh: — 2. That the vicinity of yellow or red would be inju-
rious to the effect, since their complementaries would cause the

220 PHYSICAL INVESTIGATIONS ON DYEING.

red or yellow of the flower to predominate too much : — 3. That
if one is obliged to surround nasturtiums with yellow, red, or
even orange, these colours must be so deadened as to heighten
the brilliancy of the nasturtium from the contrast of the dif-
ference in the height of the tone. This latter device of in-
creasing the brilliancy of colours has been known and practised
in all times, but I believe no attempt at an explanation of the
effect produced was made before the establishment of the law of
simultaneous contrast.

Second Example.

149. In the case of a blue agreeably tinged with violet,
we must have recourse to green or to grey, whether it border
on yellow or green, if we have no orange, yellow, or white to
put in juxtaposition with the blue ; the contiguity of violet
must be avoided, as destructive of that violet tinge of colour
which ought to be carefully preserved.

Third Example,

150. In the case of a green unpleasantly tinged with yel-
low, we must approximate yellow or a rusty yellowish grey.

Fourth Example.

151. In the case of a dull greenish blue, the unpleasant
effect of the colour will be neutralized by the juxtaposition
of green.

152. If these applications were considered in the art of
manufacturing the Gobelin tapestry, there could be no doubt
that a knowledge of the law of simultaneous contrast would
prevent much error and stumbling on the part of all, who made
it their guide in the assortment of coloured threads for copying
a given pattern in tapestry, and we ought, therefore, in future
to regard this law as one of the most valuable principles of
this art.

153. The knowledge of this law is not less indispensably
necessary when a comparison has to be drawn, with respect to
the beauty of colour, between the dyed woollen and silk threads
of carpets, tapestries, or shawls, whose designs are not iden-
tical, and whose colours are not uniformly shaded and fused

PHYSICAL INVESTIGATIONS ON DYEING. 221

into one another. In future, therefore, instead of asserting,
that the colours of oriental stuffs are superior to our own with
respect to brilliancy, the prescribed precautions ought to be
taken, in order to avoid all chance of illusion.

154. In speaking of these applications, with respect to
the painting of pictures, I think I have said enough to show
that painters, who are desirous of regulating their employment
of colours according to the law of simultaneous contrast, ought
to perfect themselves in colouring, just as in linear perspective
they study the principles of optics that refer to this department
of the art.

155. Patterns exercise so great an influence on the effect
produced in tapestry ai d carpets, that I have deemed it incum-
bent upon me, owing to the numerous observations I have had
occasion to make in the Royal manufactories, to offer some
reflections on the kind of painting most suitable for this pur-
pose. These remarks may interest artists who occupy them-
selves with works of this nature, and who endeavour to com-
prehend the principal object of this species of imitation. After
having determined what are the principal effects which they
ought to attempt to produce, they will see what points of ordi-
nary painting can be sacrificed to attain to the former ; and
they will thus be able to draw a conclusion as to what has to
be done for the perfection of the special part of their copy.
They will doubtlessly perceive, that patterns of flowers for
furniture do not require to be painted in the manner that a
pupil of V. Spaendonk would make a drawing for a work on
botany; that patterns of figures intended for tapestry must
not resemble miniatures in finish, and, finally, that in all these
productions, the objects ought to be as large as possible, and pre-
sent absolutely fine effects of colour.

156. I have established this proposition upon the nature of
the coloured elements employed by the manufacturer, (of tapes-
try, carpets, &c.,) and upon the distance at which these works
ought to be viewed. If I am not mistaken, reasons analogous
to those I have employed to prove this proposition, may pre-
sent the means of establishing, in an ostensible manner, the true
principles of several of the imitative arts, and these principles
being once deduced from the specialty of the art to which they

222 PHYSICAL INVESTIGATIONS ON DYEING.

refer, afford the means of distinguishing between the efforts by
which we may hope to arrive at true improvements, and such
as can only tend to a contrary result.

157. I have endeavoured to show, that some of the imita-
tive arts have a more intimate connexion with paintings done
in plain tinting than with those of the chiaro-obscuro style,
basing my propositions on the following facts : —

1. The eye experiences incontestible pleasure in seeing
different colours, and for this reason, the wainscoting of the
most elegant apartments is generally painted in various
colours.

2. This pleasure, which may be compared to the enjoy-
ment of taste, is augmented when bright colours are disposed
in such a manner, as to recall to mind an agreeable object,
even if it be not perfectly represented.

From this I have concluded, that when we wish to affect
the eye by colours under the following circumstances : —

1. When these colours are seen at such a distance, that
the finish of an elaborate painting would be lost, and,

2. When a painting is nothing more than an accessory
employed in the decoration of an object, the use of which
is at variance with the employment of any more elaborate
workmanship, which would besides be of too expensive a
nature ;

copies ought to be made in plain tinting rather than according
to the contrary method of painting ; and choice should be made
of such objects as are most remarkable from their beauty of
colour and simplicity of form, as flowers, insects, birds, &c.

In the course of lectures which I gave at Lyons in 1842
and 1843, I entered into an exposition of the contrast of colours
as applied to the manufacture of silk stuffs. It was with this
object in view that the Society of Agriculture and Useful
Arts, in the first place, and the Chamber of Commerce at
Lyons subsequently, petitioned the Minister of Commerce
that I might repeat the same course of lectures in that city,
that I have delivered every two years in Paris since 1828.
It was, therefore, in the wish of complying as far as I was

PHYSICAL INVESTIGATIONS ON DYEING. 223

able, with the intention of the industrial bodies at Lyons, that,
prior to the commencement of my lectures in that city, I
devoted myself to the researches which I deemed necessary to
prepare me for the instruction of the designer and the manu-
facturer, whose mutual co-operation is indispensable to the
perfection of stuffs presenting the best possible assortment of
colours, both as to their mixture and their mutual opposition.
These researches, which were made subsequently to the pub-
lication of my work on Simultaneous Contrast of Colours, in
1839, will compose the materials for the work I purpose to
bring out under the title of Theorie des Effets Optiques que
presentment les Etoffes de Soie; and of which I am going to
present a very concise extract to the Academy.

In my endeavours to reduce the optical effects of silk stuffs
to a theory, I soon perceived the necessity of placing them
relatively to the spectator under the most perfectly defined,
and the smallest number of conditions. There are four prin-
cipal conditions in which the same stuff may be seen by a
spectator standing with his face turned to the light, or, on
the contrary, with his back against the light. The importance
of distinguishing between these four conditions will be under-
stood when I have spoken of the effects of light reflected by
a system of contiguous parallel metallic cylinders.

First position of the Cylinders. — They are laid on a hori-
zontal plane, and their axis is in the plane of incident light.

First condition. — The spectator, standing with his face to
the light, sees the cylinders very strongly illumined, owing to
his receiving much light irregularly reflected.

Second condition. — The spectator, turning his back to the
light, sees the cylinders darkened, since he receives but little
light, and that irregularly reflected.

Second position of the Cylinders. — Their axis is perpendicular
to the plane of the incident light.

Third condition. — The spectator, having his face turned to
the light, sees the cylinders less strongly illumined than in the
first condition, since the only light that reaches him is reflected
by a narrow zone from the highest part of each cylinder.

Fourth condition. — The spectator, turning his back to the
light, sees the cylinders very strongly illumined, owing to

224 PHYSICAL INVESTIGATIONS ON DYEING.

each one being accompanied by a broad zone reflecting light
as by a mirror.

The cylinders when seen in the first and second position,
when the spectator faced the light, appeared to be unequally
illumined; but the difference in the quantity of the light
transmitted by them in the first and third conditions is much
less than will be the case on observing the cylinders when the
back of the spectator is turned towards the light, and for this
reason, that, in the second condition, they present the maximum
of shade, and in the fourth the maximum of light.

The effects of which I have been speaking may be observed
with cylinders of *6, *04, and *02 of an inch in diameter. By
the aid of two systems of metallic cylinders, which I place upon
the table, we are able to demonstrate that the optical effects
of the smaller cylinders are more strikingly marked than those
of larger diameter. I must add, that threads of smooth silk,
when arranged parallel to one another, have the same effect
as systems of metallic cylinders, and, on that account I purpose
speaking of the optical effects of metallic cylinders before I
treat of those exhibited by silk stuffs. It remains for me to
prove by experiment the truth of my assertion.

All woven stuffs are composed of two systems of parallel
threads, forming the warp and the woof; and are arranged
perpendicularly towards each other.

Silk stuffs are plain (a), or not figured, and (b) figured.

a. Plain or Unfigured Silks.

Plain silks are comprised under two heads: the first, show-
ing only on the right side one system of threads, constituting
either the warp or the woof; the stuffs of the second kind show
both the warp and the woof.

Plain Stuffs of the First Division.

First Section. — Stuffs whose effects correspond to those of
a system of parallel cylinders.

I am about to show the identity of the optical effects of
contiguous and parallel metallic cylinders, with those of satin
and cut or terry velvet.

PHYSICAL INVESTIGATIONS ON DYEING. 225

Satin.

Satin is a stuff, the warp of which is alone seen on the right
side, in the form of small parallel C3"linders, whose extremities
disappear in the interior of the stuff itself by the action of sizing
and dressing, which is indispensably necessary to secure the
threads in the places in which it is intended they are to remain.
The joinings of the sizing are irregularly distributed, in order
that, by concealing them as much as possible, the surface of the
satin may appear alike smooth and shining.

Satin is ordinarily made with the warp, but it may also be
made with the woof.

Cut Velvet. — Velvet termed Epingle, or Terry Velvet.

This is a fabric having transversely formed hollow flutings
or ribs which are produced by means of a cylindrical bar of
iron, which, after having been covered by the warp, is separated
from it in such a manner that the rib remains hollow throughout
its whole length, presenting on the exterior the appearance of
a cylindrical surface formed b}7 the warp.

In order that the optical effects presented to the spectator
should be what we have described, every rib formed by the
warp ought to exhibit the latter as nearly as possible in the
form of equal rings, parallel to one another, and perpendicular
to the axis of the cylinder which they represent.

Now, suppose we place two cuttings, a and b, of the same
satin upon any horizontal plane, in such a manner that the
threads of the warp or the woof of the satin, according as the
case may be, of the piece a shall be perpendicular to the threads
of the piece b, the effect will be identical with that produced
by metallic cylinders, when observed under the same circum-
stances.

The same results were obtained on making the experiment
with cut velvet, but less difference was observed between the
two pieces of velvet than had been noticed in the case of the
satin, owing to the surface of the cylinders for the velvet being
transversely ribbed instead of smooth, in consequence of these
cylinders being formed by threads wound perpendicularly round

226 PHYSICAL INVESTIGATIONS ON DYEING.

the cylindrical peg or shuttle, whose form they reproduce in
this manner.

In order that we may duly appreciate the influence that
flutings or transverse channelings may have on the effects of
light, we will now consider the manner in which light is
reflected in a system of transverse flutings.

Reflections of Light from transversely fluted Cylinders.

I place upon the table metallic cylinders having transverse
flutings of various depths, by means of which the following
facts may be demonstrated.

First position of the Cylinders. — They are laid on a hori-
zontal plane, and their axis is on the plane of incident light.

First condition. — The spectator, standing with his face to the
light, sees less light reflected than from the smooth cylinders,
owing to the flutings diminishing the extent of the surface,
which, in the smooth cylinders, reflected light as from a mirror.

Second condition. — The spectator, turning his back to the
light, receives a very strong reflection of light, as his eyes are
directed towards the face of each fluting, on which the light
falls.

This result is inversely the same as that observed in the
plain cylinders.

Second position of the Cylinders. — Their axis is perpendicular
to the plane of incident light.

Third condition. — The spectator, when facing the light, sees
the cylinders brighter than they were in the first condition:
the result is, therefore, inversely the same as that obtained
with the plain cylinders.

Fourth condition. — The spectator, when standing with his
back to the light, sees the cylinders less brilliant than in the
second condition, and much less so than in the plain cylinders.

We may, consequently, define the results of the reflection
of light by transversely fluted cylinders as the reverse of those
produced by smooth or plain cylinders.

Second section. — Stuffs, whose effects correspond to those
of a system of cylinders fluted perpendicularly to their axis, and
parallel with one another.

PHYSICAL INVESTIGATIONS ON DYEING. 227

While we are ignorant of the manner in which light is
reflected by cylinders, with smooth or transversely fluted sur-
faces, we must be at a loss to explain how cut velvet, with its
projecting ribs, will appear in the light, when compared with
satins whose surface is so uniformly smooth. Our astonish-
ment is still more increased when we observe reps (ribbed silks,)
which, like cut velvet, have marked ribs, acting upon light
differently from the latter tissues. After having compared
together the reflection of light on the surface of smooth cylin-
ders and on that of transversely fluted cylinders, we shall be
able by the aid of the lens to discover the analogy existing
between the surface of ribbed and fluted stuffs, bazinettes* , and
dimities], with that of transversely fluted cylinders, and our
surprise will then cease as we find an explanatory cause for
effects which had struck us as so strange.

Reps, (ribbed silk properly so called,) or reps par le trame,
(ribbed in the woof,) presents an appearance of ribs, the warp
of which forms the axis ; the intermediate space between the
threads of the warp forming the rib, gives rise to longitudinal
furrows. The woof entirely covers the warp on the right side,
in the form of cylindrical or flattened rings, each of which is
separated from the next by transverse furrows, generally more
strongly marked than the transverse furrows of the cylindrical
ribs of cut velvets.

I lay upon the table some specimens of reps par la trame,
(silks ribbed in the woof,) presenting precisely inverse pheno-
mena from those of satins and cut velvets.

It is certainly very remarkable, that these last-named stuffs
with their ribs should produce the same effect as satin, the sur-
face of which is the smoothest presented by any stuff, whilst
they exhibit the inverse effects to those of reps, (ribbed stuffs,)
to which they approximate so closely by their ribs.

Reps par la chaine, (stuffs ribbed in the warp,) bazinettes,
(which only differ from stuffs ribbed in the woof by the inequa-
lity in the width of the different ribs,) and dimities, (differing

* Bazinette, a muslin made of silk and wool, the ribs of unequal width. — Tr.

t Dimite, a muslin made of silk and cotton, (the ribs very wide but equal.)
Both are varieties of what in the trade in this country is known as Mousselinc
de soie. — Tr.

q 2

228 PHYSICAL INVESTIGATIONS ON DYEING.

in the size of the ribs,) produce the same effect upon light as
stuffs ribbed in the warp, and consequently as cylinders having
transverse flutinp;s.

Third section. — Mock velvet, or velvets with cotton backs.
There are certain stuffs termed mock velvet, which resemble
cut velvet, and like the latter have transverse ribs; instead,
however, of their being hollow, they are filled up by a cotton
or silk woof, in order to prevent the effect so easily produced on
the hollow ribs of cut velvet by external pressure or friction.
Mock velvets have a closer analogy with ribbed stuffs than
with cut velvet, when we consider the optical effects they pro-
duce, especially if we look at them with our face turned from
the light as in the second and third conditions; but if the
spectator face' the light, he will be able to observe specimens
of this stuff, which are more illumined in the first than in the
third condition, and they will then produce the same effect as
cut velvets.

Conclusion. — All plain stuffs which exhibit on the right side
only one system of the thread constituting them, act upon
light in a similar manner: as,

1. A system of smooth contiguous and parallel metallic
cylinders : for instance ; satins, whether showing the warp or
the woof; cut velvet, called terry velvet.

2. A system of fluted transverse and parallel metallic cylin-
ders. Reps, showing the warp or the woof; bazinettes, com-
posed of silk and wool; dimities, equally striped, silk and cotton.

3. The greater part of mock velvet stuffs act upon light in
the same manner as reps, but there are some which present to
a spectator, having his face turned to the light, effects analo-
gous to those of terry or cut velvet.

Plain stuffs of the second division. — The stuffs that simulta-
neously show both the warp and the woof are very numerous ;
such for instance are gauze, crape lisse, taffetas, including
Florence silk, sarsenet, and taffetta properly so called, louisines,
gros de Naples, pou de soies, satin turque ; various serges, com-
prising Levantine, Virginie, and Filoche*.

* Serge is a coarse black twilled silk, and Levantine, Virginie and Filoche'
mere varieties of the same, known to silk-merchants in this country by the
same names. — Tr.

PHYSICAL INVESTIGATIONS ON DYEING. 229

The surface of these stuffs may either be plain or both
striped and granular*. In all these cases, the optical effects
produced by the reflection of light are referrible to the pre-
ceding principles. These stuffs, therefore, when seen by a spec-
tator standing with his face to the light, exhibit both the warp
and the woof, and the effects produced vary with the position
of the warp relatively to the plane of the light, and according
to the predominance, subordination, or equality of the warp
with reference to the woof.

In order properly to appreciate the influence of each of the
elements of which I have spoken, with respect to the optical
effect of a pattern of a plain smooth stuff belonging to the
second division, we must first consider glace, or shot stuffs f,
that is to say, stuffs which have the warp of one colour, z,
and the woof of a colour, y, or the warp of one colour, z, and
the woof composed of two threads, one of which is of the
colour y, and the other of a colour x; and in order further to
take into account all the optical effects that may be observed,
we ought to have recourse to the principle of a mixture of colours,
and to the principle of regulating their contrast.

Conformably to the first of these principles, red mixed
with yellow gives orange; yellow mixed with blue, green;
red mixed with blue, violet; and, finally, red mixed with
green, yellow mixed with violet, blue mixed with orange, pro-
duce black or normal grey.

Further, conformably to the principle of simultaneous con-
trast of colours, when there are two parts of the surface of
the same stuff placed side by side, but so as to present two
unequally illumined, or differently coloured surfaces, these
surfaces will appear as dissimilar as possible, with respect to
brightness and colour, if one only or both are coloured ; and
in this case the modification is effected by the addition of the
complementary colour of one of the surfaces added to that of
the other surface. The following are the four principles to
which those effects are definitely referrible, which it has been
the object of the present work to explain.

* The name of the kind of silk answering to grenue is grenadine. — Tr.
t Silks shot with one colour are termed glace, or properly speaking, shot, and
chameleon when shot with several colours. — Tr.

230 PHYSICAL INVESTIGATIONS ON DYEING.

1. The principle of the reflection of light, by a system of
contiguous and parallel metallic cylinders.

2. The principle of the reflection of light, by a system of
metallic cylinders fluted perpendicularly to the axis.

3. The principle of the mixture of colours.

4. The principle of the simultaneous contrast of colours.

I will present some examples of the optical effects of glace
or shot stuffs.

First Example. — A piece of gros de Naples, the warp of
which is blue and the woof red, will appear violet to a person
turning his face towards the light ; if the warp be in the plane
of light, the violet will be redder than in the opposite case ;
this is in conformity with the principles of the reflection of
light by metallic cylinders, and with the principle of the mixture
of colours.

The same stuff seen by a person turned from the light,
appears red if the blue warp be on the plane of incident light,
and blue if the warp be perpendicular to this plane, and this
in conformity with the principles of reflection by a system of
metallic cylinders.

Second Example. — A stuff which has a blue warp, and a
woof formed of two threads, one of which is yellow and the
other red, will appear of a slightly coloured grey to a person
facing the light, in consequence of the three colours not being
exactly neutralized. These effects are produced conformably
to the principles of the reflection of light by cylinders, and to
the principle of the mixture of colours.

The same stuff seen by a person with his back turned from
the light appears, 1. blue, if the plane of the light be perpen-
dicular to a blue warp ; 2. yellow, if the warp be on the plane
of light, and if it be the yellow thread that is presented to the
eyes of the spectator ; 3. red, if the warp be in the plane of
light, and if the red thread of the woof present itself.

Such is the simple explanation of the effects exhibited by
the shot silk stuffs known as chameleon silks.

I should exceed the limits of this extract, were I to take
other examples of the glace or shot stuffs suited to the appli-
cation of the principle of the contrast of colours. I will,
therefore, present only three patterns of stuffs not shot, formed:

PHYSICAL INVESTIGATIONS ON DYEING. 231

(a) 1. One uniformly white stripe, and another like-
wise white, but open figured stripe ; the latter will appear
grey;

(b) 2. One uniformly thick yellow stripe and a white
open-worked stripe; the latter will appear lilac, from the
effect of the complementary to the yellow of the thick
stripe ;

(c) 3. One uniformly thick violet stripe and an open-
worked white stripe; the latter will appear of a straw
colour, owing to the effect of the complementary to the
violet in the thick stripe.

By an application of the researches I have made I have
been able to solve the following question: in manufacturing
a gros de Naples that is to be shot with two given colours,
which of the two should constitute the warp? my answer is, the
darker, or the least bright.

Examples. — Blue and orange shots, or blue and yellow,
violet and orange, and violet and yellow, are very beautiful
when the warp is blue or violet, but when the opposite is the
case, the effect is bad.

When a shot silk is to be made with one colour and white,
the colour must be used in the woof, and the white conse-
quently in the warp; the result is not contrary to the first
named, as I have demonstrated in my work.

Watered Stuffs.

The term watering, moire, is applied to designs produced
by means of pressure suitably applied to ribbed stuffs.

In order that the watering may produce its proper effect,
it is necessary that the ribs of the stuff should have a certain
saliency, and this can be effected by the pressure to which the
stuff is subjected being applied unequally to the different parts
of the same rib, and obliquely to its axis, as I am about to
show.

Watering presents different patterns, according as the stuff
is pressed, after having been folded in two longitudinally, or
after being placed in several transverse folds, or, finally, after
pressing two perfectly similar pieces with their right sides in
contact; modifications in the watering are likewise produced

232 PHYSICAL INVESTIGATIONS ON DYEING.

by drawing and stretching the stuff, or running the dye per-
pendicularly to the axes of the ribs towards symmetrically
arranged points, by which means undulating lines are formed
in the direction of this originally rectilinear axis.

Theory. — If the sides of the two right side surfaces in
contact were laid exactly the one over the other, and if one
piece of stuff were folded together, either in a transverse or
longitudinal direction, or if we had two like stuffs applied, the
one against the other, there would be no water if the perfectly
homogeneous sides did not mutually produce and receive im-
pressions perpendicular to the axes of the ribs, which I suppose
to be included in the same plane, and symmetrically arranged
relatively to the rings of the ribs produced by the warp; this
is the case with gros de Naples, which is a stuff peculiarly
fitted to receive the impression of watering ; if these conditions
be not fulfilled there will be a mere smoothing or crushing
of the salient parts, and the stuff will consequently resemble
smooth-faced tissues. As, however, this homogeneous con-
dition of the ribs, and pressure perpendicularly applied to their
axes, cannot be practically effected, one rib in coming in contact
with another or a different portion of its own extent exercises,
at different points of its length, an unequal pressure obliquely
to its axis, at the same time that it experiences a similar
pressure from the rib opposite to it; consequently, the initial
symmetry of the different parts of each rib is deranged.

Before examining the optical effects of the whole of the ribs
constituting a watered stuff, I will describe the modifications
undergone by a single rib in its whole extent, by the process of
watering.

The essential modification experienced by one of the ribs
in this process, is, that instead, as before it was watered, of
exhibiting on the right side a perfectly identical cylindrical
surface, having fine transverse furrows, it assumes a prismatic
form, appearing different at different points, and at different
points of view, the rib being undulated instead of rectilinear.

Thus, when on facing the light we look only at one side of
a stuff having its ribs perpendicular to the plane of light, and
being placed on a horizontal plane, there is one part that will
appear under the form of a dihedric angle, one of whose sides

PHYSICAL INVESTIGATIONS ON DYEING. 233

may be completely light and the other dark; another portion
of this rib will present one side on an inclined or horizontal
plane, which affords an excellent means of observing the effect
of the pressure on the whole of the threads, perpendicular to
the ribs, which, before the tissue was watered, constituted the
rings. In fact, the latter present a series of small shining and
satiny elliptical figures occasioned by the flattening to which
they have been subjected; these two portions come in contact
with a third, which appears, from the manner in which it
reflects light, to have been twisted, as it were, but which, in
consequence of the compression applied obliquely to its axis
by a rounded rib, resembles a furrow, one extremity of which
seems turned forward and the other backward. We may
observe by the lens that the small and silky ellipses of the
furrow are folded together in the direction of their smaller
diameter.

On ravelling out the threads of a coarse watered stuff, and
taking those on the interior of the rib, we may observe how
the whole of the threads are compressed, appearing prismatic
and twisted as it were, and being, besides, furrowed perpen-
dicularly to their length from the effect of the pressure of the
rings which partially cover them on the right as well as the
wrong side.

As the different ribs of a stuff that is not watered are all
parallel to, and dependent upon one another like parts of the
same system of a tissue, there will always be contiguous parts
belonging to different ribs which will necessarily experience
similar modifications, and in the same direction, from a like
action; and when we add to this, the effect of drawing or
running the dye towards points symmetrically placed along
the length of one rib, and which may be applied perpendi-
cularly to the axis of this rib, it will be easily understood that
contiguous and mutually dependent parts, experiencing the
same modification, will present an appearance of zones of a
certain width and a certain symmetry.

On examining with the lens a watered stuff, laid upon a
table, in such a manner that the ribs are perpendicular to the
plane of incident light, these effects will be made apparent;
when all the strongly shaded parts appear like the posterior

234 PHYSICAL INVESTIGATIONS ON DYEING.

faces of a certain number of dihedric angles of the contiguous
ribs, the parts in half shade will look like parts of the anterior
and posterior faces of dihedric angles made visible by the
inclination these portions of the ribs have received by the
pressure to which they have been submitted; and, finally, it
will be observed that the luminous parts belong to portions
of the ribs, which, having been strongly compressed, exhibit
the horizontal or slightly inclined face of a flattened prism.

On looking at the wrong side of a watered stuff the water-
ing is perfectly apparent, although there is not the same
inequality in the saliency of the different parts as on the right
side; we may likewise perfectly well distinguish the undulation
by the action of watering on the axis of the rib, which was
originally rectilinear. We will distinguish two kinds of
Watered stuffs: the first comprising monochromatic watered stuffs,
and the second, watered stuffs; as the process of watering is
alike applicable to monochromatic and to shot stuffs. The
question here arises: is the process alike advantageous to both
kinds? This subject has been treated in detail in my work,
and the researches to which its consideration have given rise
have led me to the conclusions I am about to give in a very
concise form.

There is this great difference between a monochromatic
watered stuff, and a shot stuff not watered, that the former
appears to the greatest advantage when presenting to the eye
broad plane surfaces covered with designs of great simplicity,
possessing a certain mobility and a variety of aspect which
does not injure the effect, whilst a shot material, not watered,
must be laid in folds, as when made into articles of wearing
apparel, before it will present the effects which make it so
much admired, and which exhibit a variability of colours
according to the position in which the spectator observes them,
and which thus acquire the apparent mobility of watered stuffs,
without, however, assuming the form of the undulating figures
that are characteristic of the last-named tissues. If, on the
contrary, folds are not injurious to the effect produced by a
watered stuff, it must be admitted, that it appears to the
greatest advantage when seen smoothly stretched, as in articles
of upholstery, or even as the cover of a book in the most
elegant forms of binding.

PHYSICAL INVESTIGATIONS ON DYEING. 235

We thus see that the most special use of watered and shot
stuffs is entirely in accordance with the preceding considera-
tions; add to this, that patterns of watering do not cut the
colour of the stuff more than by the opposition of shade to the
light, whilst the effect of a shot may present the most extreme
opposition of colours without losing their beauty.

The possibility of combining in one stuff the effects of
watering with those produced in a shot stuff, depends upon the
essential difference existing between these effects, and we can-
not therefore assert a priori, that the combination will, of neces-
sity, give rise to confusion.

I will now give an exposition of the facts demonstrated by
experiment regarding this question.

Monochromatic Watered Stuffs.

The beauty of a watered stuff is based upon a taste for
the design, and pleasure in the appearance of a simple form,
presenting an apparent mobility and variability, which in no
way detract from the good effect of the whole, and, conse-
quently, to attain to the maximum of the effect of which this
stuff is susceptible, it ought to present as simple an image as
possible, in order to exhibit a light, mobile, and, as it were, airy
appearance. This is the effect produced by watering in the
greater part, if not in all, monochromatic stuffs.

Shot-ivatered Stuffs.

The beauty of a shot stuff, consisting in the contrast of
its colours, its metallic brilliancy, and the lightness of its
shades, which call to mind the most varied tints of clouds
lighted by the sun, place it in the greatest possible opposition
to the beauty of the effects produced in watered stuffs. Besides
this, a shot-watered stuff, presenting as it does a great con-
trast of colours amongst the different parts of the design, loses
much of the beauty it would possess if it were monochromatic.

I conclude incontestibly from this fact, that every shot
stuff, in which the colour of the warp and that of the woof
are employed in the most suitable manner, loses by the process
of watering.

All shot stuffs do not, however, lose alike by the process of

236 PHYSICAL INVESTIGATIONS ON DYEING.

watering, and as in the opinion of some persons, watering may
add to the good effect of shot stuffs, I will speak of those cases
in which, if it be not decidedly advantageous, it is not at any

rate injurious.

The disadvantageous effect produced by watering will be in
proportion to the degree of opposition between the colours of
the warp and the woof; for instance, blue and violet, and blue
and green, produce shots, the watering of which is sufficiently
homogeneous to excite the admiration of many.

Finally, watering is decidedly advantageous to a shot stuff
rendered defective by any inequality in the threads, either by
lines or bars, since the process of watering destroys the effect,
by interrupting more or less the continuity of the lines.

B. — Figured Stuffs.

The last part of my work is devoted to the consideration of
figured stuffs. Before I speak of their effects, I will examine
the six general cases, which they present relatively to the dis-
position of the threads, and the influence thus exercised on
the optical effects produced, putting the subject of colour
entirely out of the question.

First case. — A figured stuff presenting but one sole effect,
either of the warp or the woof.

Second case. — A figured stuff, presenting simultaneously an
effect of the warp and the woof.

Third case. — A stuff, presenting an effect produced by the
warp upon a ground of taffetta silk.

Fourth case. — A stuff, presenting an effect of the warp upon
a ground of a kind of taffetta silk.

Fifth case. — A stuff, presenting the effects produced, both by
the warp and by the woof, upon a ground of a kind of taffetta.

Sixth case. — A stuff, presenting effects arising from a web of
taffetta on a ground of the same kind of web.

To this part of my work appertains especially the applica-
tion of the principle of the contrast of light, since two or more
colours may be employed in forming not only a shot, but defi-
nite and permanent figures.

The following table indicates the arrangement of the sub-
jects composing the work, of which I have given the shortest

PHYSICAL INVESTIGATIONS ON DYEING. 237

possible extract. In completing my work at a distance from
Lyons, I should have experienced great difficulties in speakino-
clearly of many circumstances connected with the weaving and
manufacture of silken stuffs, if I had not had recourse to the
aid of M. Piobert, our excellent brother Academician, whose
profound knowledge on these subjects and whose obligino-
assistance have contributed to render this work less imperfect
than it would otherwise have been, and I most gladly avail
myself of this opportunity to express my obligation to him.

TABLE.

First Point of View. — Smooth plain monochromatic stuffs considered relatively to

the effect which the warp and the woof may have upon the reflection

of light.
First Division. — Monochromatic stuffs, the essential optical effects of which may

be exclusively referred to the warp or the woof.
First Section. — Monochromatic stuffs with a smooth surface, and monochromatic

stuffs with a ribbed surface, the optical effects of which correspond

with those of a system of parallel cylinders.

1. Stuffs with a plain or smooth surface, as satins, showing the
warp or the woof.

2. Stuffs with a ribbed surface, as cut velvet, called terry velvet.
Second Section. — Monochromatic stuffs with parallel ribs, the optical effects of

which correspond with those of a system of parallel cylinders fluted
perpendicularly to their axes.

1 and 2. Reps ribbed in the woof, or in the warp and fluted.

3. Bazinettes (silk), muslin made of silk and wool.

4. Dimities (silk), muslin made of silk and cotton.

Third Section. — Monochromatic stuffs having parallel ribs, the optical effects of
which correspond both with those of a system of parallel cylinders, and
with those of a system of cylinders fluted perpendicularly to their
axes.

Certain Mock or Imitation Velvets.

Fourth Section. — Monochromatic stuffs having parallel ribs, the optical effects of
which correspond with those of a system of parallel cylinders, fluted
in such a manner as to appear equally light in the first and third
conditions.

Certain Mock or Imitation Velvets.

Second Division. — Monochromatic stuffs, whose optical effects are referrible both
to the warp and the woof.

238 PHYSICAL INVESTIGATIONS ON DYEING.

/ Gauze.

Crape lisse.

Florence silk.

Sarsenet.
First Section.— i Taffetta.

Louisine.

Gros de Naples.

Pou de soie.

v Satin turque. (Also Turquoise, so called from the colour.)

, „ . (Serge — Coarse black twilled silk.

Second Section. — < . .

Third Section.- FikchT } varieties of the same.

Second Point of View. — Smooth plain stuffs, the warp and woof of which are
apparent, and exhibit two different colours.

Shot Stuffs.
First Section. — Shot stuffs having a monochromatic woof.
Second Section. — Shot stuffs having a bichromatic woof.

Third Point of View. — Plain monochromatic stuffs, or shot stuffs considered
relatively to the process of watering.

Watered Stuffs.

First Division. — "Watered stuffs.

Second Division. — Shot watered stuffs.

First Section. — Shot watered stuffs having a monochromatic woof.

Second Section. — Shot watered stuffs having a bichromatic woof.

Fourth Point of View. — Stuffs considered with reference to fixed patterns, that
is to say, such as preserve their tints whatever be the position in
which they are considered.

Figured Stuffs.

First Division. — Monochromatic figured stuffs, the optical effects of which are
exclusively referrible to the warp or to the woof.

Second Division. — Monochromatic figured stuffs, whose optical effects are simul-
taneously referrible to the warp and the woof.

Third Division.— Figured stuffs, whose optical effects are referrible to the
different tones of the same colour exhibited in the threads.

Fourth Division. — Figured stuffs, whose optical effects are referrible either to
the threads of one colour, or of several colours being combined with
white, black, or grey threads, or to threads of many colours com-
bined together, including white, grey, and black.

The work, of which the above is a short extract, was pub-
lished in the last session of 1846, at the expense of the Chamber
of Commerce at Lyons.

239

VI.

MEMOIR.

ON THE LATENT HEAT OF STEAM AT DIFFERENT

PRESSURES.

By V. Regnault.

My object in the present memoir is to determine the amount
of heat necessary to convert 1 kilogramme of water at 0° into
steam at different pressures. I shall express these quantities of
heat by designating the number of kilogrammes of liquid water
they are able to heat from 0° to 1° centigrade.

In the middle of the last century, Black was the first who
made the important observation that, a large quantity of heat
was absorbed in the formation of steam, or instituted experi-
ments for the purpose of measuring this quantity. His manner
of proceeding was as follows. He placed a small vessel full of
water on the top of a cast-iron stove, in which afire was kept up
with sufficient regularity to admit the assumption, that the
vessel received in equal periods of time pretty nearly equal
quantities of heat; and he then made comparisons between the
time it required for the water to come to the boiling-point, and
the time required for the same water completely to boil off.
This experiment, although sufficient to demonstrate the absorp-
tion of heat which takes place during evaporation, is too rough
to give the result in strictly accurate measurements. Black

240 LATENT HEAT OF STEAM

was himself aware of this fact, and he subsequently in concert
with Dr. Irvine, instituted new experiments, in which he made
use of the method of mixtures, which he had devised for the
purpose of determining the specific heat of bodies. He made
use to this effect of an ordinary still, and measured the elevation
of temperature experienced by the cold water surrounding the
worm, by the condensation of a certain quantity of steam.
Black thus obtained 530 degrees Cent., which is much too low,
owing to his not having taken into account the necessary cor-
rections*.

The celebrated Watt made experiments on this subject at
various intervals, and was first led to direct his attention to the
subject at the suggestion of Black, whose pupil he had been.
Watt's first observations date from 1765, and yielded him the
number 766, which, however, he himself regarded as inaccurate.
He resumed the subject in 1781, and then made the experi-
ments, described by himself in the article Stea?n in Robison's
Mechanical Philosophy. (Brewster's Edition, vol. ii., p. 5.)
The mean of 11 determinations gave the number 625*2, but this
was regarded by Watt as too low, and he adopted as a more
probable number, 633*3.

Rumford attempted to determine the same element by means
of his calorimeter. Water was put in a glass retort, the neck
of which rising up was inserted by a cork in the mouth of the
worm of his calorimeter.

Rumfordf thus found the numbers 669*0, 670*8, and 671*9
in three different experiments. These numbers are too high.
In the method pursued by Rumford, a certain quantity of steam
was condensed in that part of the neck of the retort within the
calorimeter, which imparted heat to it, although it fell back into
the retort in a liquid state.

Dr. Ure published a memoir in the Phil Trans. 1818,
p. 385, in which he proposed to determine the latent heat of
evaporation of a certain number of volatile substances. Ure's
experiments were made with a simple apparatus, which, how-
ever, was ill adapted for yielding precise results. It consisted
of a small glass retort, having a short neck which entered into a

* Lectures on the Elements of Chemistry, vol. i.
t Biot, Traite de Physique, t. iv., p. 710.

AT DIFFERENT PRESSURES. 241

receiver of thin glass, of a spherical form, and 3 inches in
diameter. This receiver was surrounded by a certain quantity
of water enclosed in a glass cylindrical vessel. 200 grains of
the liquid to be tested was put in the small retort, and rapidly
distilled by means of an Argand lamp. The water of the small
calorimeter was continually agitated by means of a thermo-
meter, which indicated its temperature. Dr. Ure by this
method found the number 637*5 as the representative of the
quantity of heat required to reduce a kilogramme of liquid
water at 0° to steam at a pressure of 760 millimeters. It must,
however, be observed that the results arrived at by Dr. Ure
were calculated with an incorrect formula, as has already been
mentioned by M. Brix {Poggendorffis Annalen, LV., p. 351),
and that the preceding number becomes reduced to 593*4
on applying the true formula to the numerical data of his
experiment.

Dr. Ure states in the memoir already cited, that Lavoisier
and Laplace had found the number 655, but I have sought in
vain for this determination in the memoirs published by these
illustrious authors.

T\re find in the Traite de Physique, by Biot, t. iv., p. 713,
that Gay Lussac on the one hand, and Clement and Desormes
on the other, have obtained the same number, viz. 650; but I
have nowhere been able to find the details of their experi-
ments.

The most important and recent experiments made on the
subject under consideration are due to Messrs. Despretz and
Brix.

In a first series of experiments {Annates de Chimie et de
Physique, t. xxiv., p. 323) Despretz made use of an apparatus
similar to the one employed by Rumford. The mouth of the
retort enclosing the liquid entered into a copper worm which
passed through an oblong and rectangular box of the same
metal, and came out through one of its walls; the condensed
liquid passed through the worm and fell into an external vessel
so arranged as to receive it. The box held 2703 grammes of
water. Despretz found the number 631 for the total heat.

In a second series of experiments, Despretz made use of a
calorimeter of much larger dimensions, which was capable of

242 LATENT HEAT OF STEAM

containing 30 liters of water. The condensed liquid remained
at the bottom of the worm and could be weighed with exact-
ness. This series gave a higher number than the former,
namely 640.

M. Brix in his memoir {Poggendorff's Annalen, LV., p. 341)
entered into an elaborate discussion of the different causes of
error that may affect the results obtained by the method of
mixtures, when used to determine the latent heat of steam, and
he endeavoured by a long course of careful investigation to
render the sum of the errors as small as possible. M. Brix's
principal object was, however, to determine the latent heat of
other liquids than water; and it was probably on that account
that he made use of an apparatus of such very small dimen-
sions. (Fig. 1, Plate I.)

In this case, the sum of the corrections necessarily acquires
a very large relative value, and consequently throws more
uncertainty upon the absolute value sought.

However this may be, M. Brix deduced from his experi-
ments on steam, the same number 640, which was obtained by
M. Despretz in his second series of experiments.

It may be seen from the preceding remarks that the numbers
obtained by different experimentalists differ considerably from
one another; but it must be observed that the experiments of
Messrs. Despretz and Brix, which merit most attention, lead to
the same result. The number 650 has been generally adopted
in practical mechanics.

The labours that I have enumerated were solely directed to
the determination of latent heat of steam under ordinary atmo-
spheric pressure; while the experiments, whose object it has
been to determine the same element under different pressures,
are very imperfect, and much less numerous.

Watt admitted that the quantity of heat necessary to convert
1 kilogramme of water at0° into steam at any pressure, is constant;
consequently, this quantity is always the same whatever be the
temperature of the steam, provided the latter is in a state of
saturation.

This proposition which is known as Watt's law ought rather
to be regarded as an hypothesis of this celebrated mechanician,
since he did not establish it by direct experiment, but simply

AT DIFFERENT PRESSURES. 243

made one experiment under a lower pressure than that of the
atmosphere, and this was admitted by Watt himself to have
been very imperfect*.

Southern and Creighton made more carefully-conducted
experiments in 1803, to determine both the density of steam at
different pressures, and its latent heat under the same con-
ditions. Their observations are incorporated in a letter ad-
dressed to Watt, and which has been published by Kobison, in
his Mechanical Philosophy, vol. ii., p. 160.

In a first series of experiments, Southern found that the
latent heat of steam may be represented by the following
numbers :

515*5 at a pressure of 1016 Millimeters.

541-1 „ 2032 „

534-0 „ 3048 „

In a second series, which he regards as more exact than the
former, the same physicist found :

523*3 at a pressure of 1016 Millimeters.
523-3 „ 2032 „

52?7 „ 3048 „

Southern concludes from these experiments that the latent
heat of evaporation, that is to say, the heat absorbed in the transi-
tion from the liquid to the gaseous state, is constant at all pressures,
and that the total quantity of heat may be obtained by adding to
the constant latent heat the number representing the temperature of
the steam.

This law is known to physicists as Southern's law. Clement
and Desormes made new experiments on this subject in 1819.
They made use of a large boiler, in which the water might be
brought into ebullition at different pressures, and caused equal
quantities of steam to be condensed in a refrigerator containing
the same quantity of cold water, taken at the same temperature
in each experiment. They thus discovered that the refrigerator
acquired the same temperature at the close of every experiment,
whatever the pressure might have been at which the steam had
been formed. The experiments of Clement and Desormes thus
confirmed Watt's law.

* See " Mechanical Philosophy " by Robison, vol. ii., p. 8.

R 2

244 LATENT HEAT OF STEAM

The labours of these physicists are only known through an
extract published by M. Thenard in his Traite de Chimie, vol. i.,
p. 78, and communicated directly to him by one of these obser-
vers. This extract will be found in the accompanying note*.

In order to complete the historical sketch of the attempts
that have been made to determine directly the question before
us, I must not omit to mention that M. Despretz states in his
Traite de Physique, 4th Edition, p. 212, that he had made some
experiments to determine the latent heat of steam at great pres-
sures, but that he had been arrested in his course by the dif-

* The experiments of Cle'ment and Desormes were made in the following
manner. They made use of a large boiler, in which the steam might be made
to acquire an expansive force of several atmospheres. A copper pipe of a
small bore was made to conduct the steam into a trough, containing a mass of
cold water, weighing 290 kilogrammes ; the escape of the steam being regulated
at will by means of a cock. A manometer, placed upon the conducting steam
pipe, indicated the degree of tension, whilst a thermometer marked the amount
of temperature.

In the first experiment, the steam had a tension of 4 atmospheres, and a
temperature of 1 52°^-. 1 4-g- kilogrammes of steam were introduced into the
cold water; the time occupied in the process did not exceed two minutes.
The temperature of the water was previously 20°, and subsequently 49°^.

The quantity of heat before the experi-
ment is represented by 290 kil.at + 20° = 5800 units.

After the introduction of the steam, by 304. 50 „ at + 49°i-= 14972 j „

The 14| kilog. have thus yielded 9772.

and consequently 1 kilog. yielded 632.

It was thought that something ought to be added to the immediate result
of the experiment for the loss of heat sustained during its continuance, but the
period was so short, that Messrs. Clement and Desormes believed that they
indicated its maximum by raising the quantity of the constituent heat of
1 kilogramme of steam to 650 units, their unit of heat being equal to that
whicli is sufficient to raise I kilogramme of water one degree of the centigrade
thermometer.

The same experiment, repeated on the same quantity of steam, with the
same apparatus, the same day, and with an equal quantity of cold water, in
fact, under perfectly analogous circumstances, with the sole exception that the
steam had different tensions and different temperatures, had precisely the same
results as those yielded by the previous experiment. It was far more easy to
perceive the similarity than to appreciate any differences that might have pre-
sented themselves, and this very circumstance, more than the size of the appa-
ratus and the care directed to the execution of this curious experiment, seems
to place the result beyond all doubt.

AT DIFFERENT PRESSURES. 245

ficulty of constructing an apparatus capable of resisting such
high pressures. It would appear that M. Dulong experienced
similar obstacles (see Cours de Physique, by M. Lame, t. i.,
p. 487, le edition).

Mechanicians have in general admitted Watt's law, which is
very available in calculations, and which moreover appears to
be confirmed by the practical observation, that it requires
nearly the same quantity of fuel to form one kilogramme of
steam, whether at low or high pressure.

M. de Pambour {Traite de Locomotives, fyc., 2e edition,
chap, ii, and vii) found Watt's law confirmed by the experiments
which he made on a locomotive. This skilful mechanician
observed that steam, when forming in a boiler under an absolute
pressure varying from 2*7 to 4*4 atmospheres, and escaping into
the atmosphere with an absolute pressure of 1*40 to 1*03 atmo-
spheres, presents precisely the same temperature at the time of
its escape, as if it were still in a state of saturation. This cir-
cumstance coincides with Watt's law, whilst, according to
Southern's law, the steam should escape at a higher temperature,
which would be the same in theory as that possessed by the
steam when in the boiler. In order, however, to verify
M. Pambour's conclusions in all respects, the steam must be
assumed to experience no refrigeration during its passage
through the apparatus, but this is a condition which it is ex-
tremely difficult to realize even with locomotives; moreover, it
must likewise be assumed that the steam enters the engine
perfectly dry, without the admixture of any liquid particles.
This is also a very difficult condition to fulfil, especially in
locomotives, owing to the rapid play of the slide valves; and
the presence of a very small quantity of liquid water, is suf-
ficient to explain the refrigeration of the steam and its con-
dition of permanent saturation, even assuming Southern's law
to be correct.

Finally, several authors have tried to demonstrate, a priori,
the correctness of Watt's or Southern's law, by basing their
arguments on more or less ingeniously conceived physical or
mechanical grounds. I will not pause to discuss these specula-
tions, which always depend upon principles open to contest;
it is evident that the question, engaging our attention, can only

246 LATENT HEAT OF STEAM

be decided by direct experiments, and the historical detail that
I have given of the results yielded up to the present time,
shows that they are inadequate to the solution of the difficulties
presented by the subject.

Before I proceed to describe the apparatus which I have
employed in the determination of the latent heat of steam at
different pressures, it would seem expedient that I should pause
for a moment to discuss the method of mixtures, in order to
establish the causes of error and uncertainty to be avoided ; by
which means the object of my mode of arrangement will be
better understood.

The apparatus employed for determining latent heat of
vapours by the method of mixtures consists, essentially, of a
retort A (fig. 2, Plate I.,) containing the liquid whose latent
heat of evaporation is to be determined, and a calorimeter C
enclosing a worm in which the evaporated liquid may be con-
densed. The retort communicates with the worm by means
of a neck a b c, bent at b. This neck is made of one rising
portion a b, disposed in such a manner that vapour condensed
within it may fall back into the boiler, and of a descending
portion b c, which is very short and goes into the worm. The
direction changes abruptly at b; all the steam which is con-
densed beyond b, falls into the worm, and is considered as having
parted with its latent heat to the calorimeter. The condensed
liquid remains in a box B placed at the end of the worm, and is
made to flow out after the conclusion of the experiment in such
a manner that its weight may be determined.
Let:

p be the weight of the liquid condensed in the worm ;
P 9i „ water of the calorimeter, augmented

by the weight of the water, which owing to its capacity for
absorbing caloric, is equivalent to the calorimeter and to the
worm ;

t0 the initial temperature of the water of the calorimeter;
tY the final temperature ;
T the temperature of the steam ;

c the specific heat of the fluid, that of water being taken as
the unit ;

A the latent heat of the steam,

AT DIFFERENT PRESSURES. 247

We thus have :

p\ +Pc(T—tl)^F(t—t0)9

whence is follows that :

This expression requires, however, several corrections before
it can give the actual latent heat sought.

1. The fluid always requires a somewhat long period of
time for its distillation, during which the calorimeter loses a
certain quantity of heat by radiation, and by the contact of
the surrounding air; the final temperature t, will, therefore, be
lower than it would have been, if this loss of heat had not taken
place.

Rumford proposed to eliminate this cause of uncertainty,
by taking the water of the calorimeter at an initial temperature,
lower by several degrees than the temperature of the surround-
ing air, and continuing the passage of the steam until the
temperature of the water exceeded that of the air, by the same
number of degrees by which it had previously been below it.

The absolute value of the correction would certainly be
considerably diminished by this method, but it would not be
wholly removed. The quantities of heat gained, or lost in the
two parts of the experiment would perhaps be equal, if the
refrigeration and the heating wrere both effected solely by
radiation; but they differ in reality when the experiment is
made in the air, and especially, when the latter is agitated.
Besides, it is only in rare cases, that water can be introduced
into the calorimeter at 5° or 6° below the temperature of the
free air, without causing a deposit of dew to be formed on the
surface of the calorimeter. This dew, by its subsequent evapo-
ration in the course of the experiment, becomes a new source of
disturbance.

This correction is generally determined by calculation,
applying the law of Newton to the refrigeration of the vessel;
that is to say, the rapidity of the refrigeration of the vase is
supposed to be proportional to the excess of its temperature
above that of the surrounding medium. If we designate the

248 LATENT HEAT OF STEAM

temperature of the surrounding medium as t, the variable tem-
perature of the calorimeter as t, and the time as x, we shall have

dsk- A <<-*>

The constant A is determined by direct experiments on the
refrigeration of the apparatus.

On dividing the length of the experiment into elements,
A x= 1', we may suppose t to be constant and equal to its mean
value, during n minutes. The value of the correction is there-
fore:

X A t=A (t—r) n A x.

The law of Newton would perhaps give this correction with
sufficient exactness, if the refrigeration took place in vacuo, or
at any rate, in an atmosphere slightly agitated; but it becomes
inexact when the air is in a state of agitation, and more espe-
cially, when its motion varies at different moments during the
experiment. It often even becomes extremely difficult to know
what value of t must be adopted for the temperature of the
external air.

I know of no means of determining this correction with
precision, and the experiment must therefore be so conducted
as to render the correction as small as possible. The most
certain means of attaining this result, is by using an apparatus
of considerable dimensions. The rapidity with which the
calorimeter cools, may also be slightly retarded by placing the
vessel within a second one formed of very thin metal, and lined
with cotton wool, or swan's down.

2. A second correction, more uncertain even than the first,
is produced by the pipe, which carries the steam into the worm,
bringing with it at the same time, a certain quantity of heat by
conduction. It is impossible, in the ordinary manner of con-
ducting this experiment, to determine this correction with any
certainty.

The quantity of disturbing heat thus introduced, varies with
the position of the bend b on the pipe. The portion a b of the
tube may be supposed to be heated by the steam condensed
along the walls, and which falls back into the retort. Even
supposing that no particle of steam passes beyond the plane

AT DIFFERENT PRESSURES. 249

a /5, which intersects b, the calorimeter will not the less receive
a certain quantity of heat, brought to it by internal communi-
cation along the tube b c.

It is evident that this portion of heat will attain its maxi-
mum when the bend b is immediately at the entrance of the
calorimeter, and that it will diminish, on the contrary, in pro-
portion as b c is lengthened: it would likewise diminish if the
tube b c were made of a substance, which was a bad conductor
of heat.

But on the other hand, if b c be of any considerable length,
a cause of error arises, which acts in a contrary manner to the
former. All the steam that passes beyond the limit a /3 is
considered as acting upon the calorimeter, but a portion of the
heat given off by this steam during its passage into the tube
b c, is lost by radiation, and by contact with the cold external
air. This loss of heat is great in proportion to the length of
the part b c of the tube. It will easily be conceived, that in
every experiment made on a definite fluid, there is a position
of the bend b, in which the quantity of heat gained by the
first cause, and the quantity of heat lost by the second,
counteract each other, but it is difficult to determine this posi-
tion a priori.

3. The steam which penetrates into the calorimeter neces-
sarily carries with it particles of fluid. Even if we suppose
the steam to be perfectly dry in that part of the boiler where
it passes into the neck a b, it must necessarily experience a
refrigeration in traversing this tube, which will determine the
precipitation of a part of the steam in a liquid condition. The
greater portion of this fluid returns to the boiler along the sides,
but an appreciable part remains in a gaseous condition, and is
carried into the calorimeter by the current.

The inconvenience of which I have spoken is manifested,
especially, at the beginning and the close of the experiment.
At the outset the retort contains air : this air is expelled by the
steam which is developed, but on its disengagement in passing
through the worm, it necessarily brings with it a certain
quantity of steam, which is then condensed. The heat yielded
by this condensation is not taken into account, since the initial
temperature t, is not observed until the distillation has become

250 LATENT HEAT OF STEAM

very active. Besides, during the first few moments of the
distillation, the tube a b is not heated, the condensation of the
steam is abundant, and much liquid is consequently yielded.
The experiment is usually terminated by removing the fire
from under the boiler, but the distillation instead of being
immediately stopped, is gradually diminished, whilst its last
stage is attended by perturbations analogous to those which
took place at its commencement.

4. In the experiments that have been made to determine
the latent heat of steam under pressures more considerable than
those of the atmosphere, steam is developed in a boiler at high
pressure; but immediately on its escape from the boiler, this
steam falls to the pressure of the atmosphere in the pipe which
carries it to the calorimeter. It is difficult to explain what
occurs owing to the enormous expansion to which the steam is
thus subjected; the result may probably be a considerable loss
of heat, especially, when the conducting pipe is very long, and
such must inevitably be the case, where a steam-engine is used
in the experiment.

It appears absolutely necessary that the steam should be
conducted to the worm with the same elastic force which it
possesses in the boiler.

5. The water which is collected in the box B of the worm
has not precisely the same temperature as the surrounding
water of the calorimeter, at the moment when the maximum
temperature of the latter is observed. When the experiments
are made solely at the pressure of the atmosphere, we may
compute the difference of the temperature with sufficient pre-
cision, by means of a small thermometer, the bulb of which is
placed in the centre of the box B, but this method is not
applicable when steam condenses at a high pressure. Having
made these preliminary remarks, I shall now proceed to the
description of the apparatus.

The most essential parts are, —

1. A boiler, A.

2. A condenser, D.

3. An air receiver, E F, which acts in the place of an arti-
ficial atmosphere.

4. A system of two precisely similar calorimeters, C and C.

AT DIFFERENT PRESSURES. 251

5. A cock, R, by which the steam is distributed.

6. A mercury manometer.

7. A forcing air-pump.

1. The boiler consists of a vat made of iron plate, 12 mm.
in thickness; the diameter of this vat is about 0m,64, and its
height 0m,80. It is surrounded by a ring of cast iron to which
the cover of the boiler is attached by means of strong pegs
screwed on, and by a joint. The cast iron cover, about 30 mm.
in thickness, has two small tubes. The upper tube is closed
by a cast iron plate, to which three iron tubes are hermetically
attached at their base. These tubes are intended to receive the
thermometers employed to indicate the temperature of the
steam, and of the water in the boiler. Two of the tubes have
an internal diameter of about 10 mm. ; the one is carried to
the bottom of the boiler, whilst the other does not descend
more than half way, and does not reach the level of the water;
both are intended to have mercurial thermometers. The third
tube, which has an internal diameter of about 30 mm., is made
to enclose the bulb of an air thermometer.

The boiler is capable of containing about 300 liters; 150
liters are poured in through an opening in the cover. This
aperture is kept closed during the experiment by means of a
screw-peg, and an annular lead disk. The boiler must be placed
on a brick furnace, which has a high iron-plate chimney, and a
register, intended to regulate the draft. The bottom of the
boiler and about one decimeter of its lateral walls are alone
exposed to the direct action of the fire, in order to prevent
the steam being over heated. The fuel generally employed is
coke, although a mixture of this substance and pit coal is some-
times used.

A copper tube T T' enters a tubulure T, and is intended to
carry the steam to a regulator R.

It is requisite to take precautions to guard against the con-
densation of the steam in the conducting pipe, as well as against
the escape of the liquid water. The first condition necessary
to effect this is to prevent the refrigeration of the conducting
pipe, by the surrounding air. For this purpose the apparatus
is so arranged, that the steam which is to be carried to the
calorimeters is enveloped, during the whole of its passage, by a

252 LATENT HEAT OF STEAM

dense stratum of steam, which is not otherwise employed in
the exj)eriment, but possesses the same temperature as the rest,
being drawn from the same boiler.

The pipe which carries the steam to the distributing cock
R, and consequently also to the calorimeters, has an internal
diameter of 35 mm., and is placed in the axis of the pipe TT',
which is nine centimeters in diameter. The pipe TT' does not
communicate immediately with the regulator R, as may be seen
in fig. 5, which gives a sectional view of the regulating cock;
but a small tube carries off the steam and conveys it directly
to the condenser. The interior tube O O' (fig. 6, Plate I.,) does
not terminate at the tubulure T, but penetrates into the interior
of the boiler, passing twice round it in the form of a worm, its
opening being in the centre of the steam-receiver of the boiler.
Thus the steam which is drawn from the* centre of the boiler,
passes through a worm four meters in length, round the interior
of the boiler ; reaching the regulator R without experiencing
any cause of refrigeration, since the tube 00" which it tra-
verses is in the axis of a box filled with steam, having the same
temperature.

2. The condenser D is a cylinder made of iron plate, 12
mm. in thickness. It is placed in a large reservoir also made
of iron plate, and filled with cold water, which keeps the
cylinder constantly at a low temperature. An index for the
level of the water is fitted to one of the walls of the condenser,
and shows at every moment, the quantity of water which has
passed into the condenser, and the manner in which the distilla-
tion is proceeding.

The condenser D is capable of containing 60 liters; the
surrounding water must be kept at a low temperature, and
must consequently be continually renewed. A reservoir X
furnishes the water necessary for this purpose.

3. The air receiver E F, is composed of a cylinder made of
iron plate, 12 mm. in thickness; the diameter of this cylinder is
about Om,72, its length lm,40, and its total capacity is conse-
quently about 600 liters. The cylinder is placed in a basin Y,
being supported by two iron bars. The air receiver communi-
cates with the forcing air pump, and is connected with the
boxM.

AT DIFFERENT PRESSURES. 253

4. The arrangement of the calorimeters will be more easily
understood by a reference to fig. 4, Plate I., which represents a
vertical section of one of these instruments, and a portion of the
main cock.

The calorimeters consist of two red copper cylinders, having
very thin metal covers. A cock s, placed at the lowest point of
the cylinders, allows of the escape of the water contained
within them.

The worm consists of a first bulb A made of red copper,
2 mm. in thickness, into which the steam to be condensed passes
directly. The water, as well as the steam that has not been
condensed, passes through the tube gh into a second bulb B,
similar to the first, which has at its lower part a cock rx placed
on the outside of the calorimeter. The same bulb B has an
upper tubulure a, by which it is connected with a copper worm
passing from the calorimeters along the tube de yr This tube
de Yj has a band by which the worm may be adjusted to copper
tubes, and made to communicate with the box M. The tube
de 7 of the worm is maintained in the axis of the calorimeter
by a narrow band of copper b d c.

An agitator or fan, composed of two disks of fluted copper
as seen in fig. 5, Plate I., serves to blend together the strata of
the water in the calorimeter during the experiment. The two
disks are raised on vertical metallic rods, so arranged that when
the agitator is at the lowest part of its course, the lower disk is
two centimeters from the bottom of the calorimeter, whilst the
upper disk is about half-way. By this means a perfectly equal
agitation is procured in both calorimeters.

The whole course of the agitators is equal to the half of the
height of the calorimeters.

The same volume of water is introduced into the calorimeters
at every experiment, being measured by means of a gauging
vessel H.

At the lowest point of the gauging vessel, a cock with three
branches is soldered on, so placed, that the water may be con-
veyed at will to the calorimeter C, or to the calorimeter C.

5. The cock R, by which the steam is distributed, will best
be understood bv figs. C> and 7, Plate L, which represent two of

254 LATENT HEAT OF STEAM

its vertical rectangular sections, and by fig. 4, Plate I., which
shows the mode of its disposition between the two calorimeters
C and C.

This cock is composed of a bronze box, cast in a single piece,
and composed of a first annular space J Inm, to which the steam
passes directly by the pipe O' O, and a central space h g i k,
which is slightly conical and serves to enclose a hollow cock
a b d e, a section of which is given in fig. 8, Plate. Two tubes
oC, and o'C', passing to the same basin, establish the communi-
cation between the central space h g i k and each of the calori-
meters ; whilst the hollow cock ab d e establishes a communica-
tion by the lateral opening O, either with the calorimeter C, or
with the calorimeter C, or allows of the complete interception
of the passage of steam into the calorimeters.

The branch of the cock passes into the tubulure B, (fig. 6,
Plate I.,) which is hermetically closed by means of a ring of
hemp, covered with melted caoutchouc, which is rammed into
the space B by means of the screw-peg A. This branch has a
handle o2 m (fig. 4), by which the cock is worked.

Considerable difficulty was experienced in connecting the
tubulures oC and o'C with the worms, since I required that all
the joined parts should be placed in the interior of the calori-
meters in order to keep the external parts within small dimen-
sions. The mode of adjustment, which I adopted, will be easily
comprehended by figs. 4 and 7.

The tubulures oC and o'C terminate in two joined pieces
Css', passing to the same basin, and supporting one turn of the
interior screw and one turn of the exterior one.

The communication with the worm is established by means
of this turn of the interior screw. For this purpose, the tube
i i' of fig. 4, one portion of which is represented by r q
in fig. 7, terminates in a plane disk, which fits into a rabbet
in the interior of the joint Css' at the end of the turn of the
screw.

A small twist of hemp steeped in a solution of mastic with
red lead and surrounding the tube q r, is compressed by means
of a nut D against the plane disk, terminating the tube q r,
and thus effects a hermetical closure.

AT DIFFERENT PRESSURES. 255

As the tube q r is inclined towards the bulb A, in order to
facilitate the flow of the water yielded by the condensation of
the steam, it was found necessary to give a similar inclination to
the pieces Css'. This circumstances increased the difficulties
attending the construction of the part of the apparatus under
consideration.

It now remains to join the whole of these pieces to the wall
of the calorimeter. This wall has a circular orifice a little larger
round than the turn of the external screw of the joint Css'. A
ring of greased copper is attached to the exterior of the calori-
meter between its side and the disk s s'. A second ring, similar
to the former, is placed upon the inner wall of the calorimeter,
and the whole system is tightened by means of an annular nut
E E' which works upon the turn of the external screw of the
part Css'. As, however, the partition of the calorimeter is ver-
tical, and the direction of q r is inclined, it was found necessary
to insert a metallic wedge-shaped ring F F' between the nut
E E' and the wall of the calorimeter.

This mode of arrangement presented some difficulty, since it
was necessary for the parts to be kept hermetically closed under
pressures of from 15 to 20 atmospheres, whilst they were tra-
versed by steam having a temperature of 200°. The object of
placing all the metallic joined parts in the water of the calori-
meters was, however, effected, by which means it was found easy
to determine at every moment, whether there was any escape
of heat, thus removing a cause of anxiety, that would have
existed if these parts had been placed on the exterior, owing to
the small quantity of heat which they would have been able to
convey to the steam.

6 and 7. The mercurial manometer and the forcing air-
pump have been already described in my Memoirs on the com-
pressibility of gases, and on the elastic force of steam.

The manometer communicates with the whole of the ap-
paratus by means of a tube which passes into the tubular
box M.

All parts of the apparatus have been constructed in such a
manner as to resist a pressure of 20 atmospheres ; but the prin-
cipal difficulty consisted in making the whole completely imper-

256 LATENT HEAT OF STEAM

vious to the air, and able to resist the access of currents of air,
which might pass through the small fissures of the metal, or
between the numerous joints which it was impossible to dis-
pense with.

The cylinders of fluted iron plate were frequently moistened
externally and internally with a solution of sal ammoniac, and
then exposed for more than a month to the oxidising action of
the atmosphere. By this means the greater number of the
crevices in the fluted joints were closed up by rust.

The apparatus was then finally put together, and air at a
pressure of 5 atmospheres was compressed into the interior by
means of a forcing air-pump. The basins, calorimeters, &c, &c,
were filled with water in order to subject all the parts capable
of being put under water to its influence ; whilst those portions,
which did not admit of being thus submerged, were carefully
examined, after being moistened with soap and water. A great
number of crevices were thus detected, principally in the iron
plate; but it was found easy to fill them up by ramming the
metal with a graver. A few chinks were also discovered in the
bronze parts of the apparatus, and although they might have
been stopped by tin solder, I thought it was most prudent to
have new parts cast since they would necessarily be very
strongly heated when the apparatus was in use.

The copper tubes had been previously tested in a hydraulic
press to 20 atmospheres.

After these preliminary trials the apparatus was again put
together, and air introduced into the interior at a pressure of
10 atmospheres, and then left during 24 hours. At the end of
that time the column of the manometer had only descended
1 decim.; and this slight diminution of the pressure had in a
great measure been occasioned by a depression of the tempera-
ture, and probably also by an absorption of oxygen owing to the
humidity of the metallic walls. It was otherwise impossible to
discover any crevice by means of soap and water.

After this trial, the apparatus was considered fit for use.

AT DIFFERENT PRESSURES. 257

I. Total heat of Steam under the ordinary Pressure of the

Atmosphere.

I will begin by describing the experiments that have been
made under the ordinary pressure of the atmosphere, in which
case a great part of the apparatus becomes useless.

The apparatus is put in communication with the external
air by removing the cock R4 from the air receiver; it is also left
in communication with the mercurial manometer, in order to
show that there is no sensible excess of pressure in the appa-
ratus during the distillation.

About 150 liters of water are introduced into the boiler, and
two mercurial thermometers are so arranged in the iron tubes,
that the column of mercury hardly projects beyond the boiler.
These thermometers are observed by means of a horizontal lens.

All the parts of the apparatus which are traversed by the
current of steam and exposed to the external air, are enclosed
in several folds of flannel and list. This covering has been put
round the tube TT', the distributing cock R with its appendices
i i' leading to the calorimeters, and finally round the return pipe,
which carries the steam to the condenser D.

The distributing cock is so placed that the steam cannot
reach any of the calorimeters, but must pass directly to the
condenser after having traversed the cock R. The distillation

is carried on in this manner for 4 of an hour, or a whole hour,
until 20 or 30 liters have been conveyed to the condenser; the
air is completely expelled from the boiler, and the different
parts of the apparatus are placed in a stable condition of
temperature.

The amount of cold water is introduced into the calorimeter
and the agitators are made to work. The thermometers of the
calorimeters are so arranged that the upper extremities of the
columns scarcely project beyond the stoppers, by which they
are attached to the cover of the calorimeter, and the degrees
are observed by means of horizontal lenses which move along
vertical columns.

A preliminary experiment is made, consisting in an observa-
tion of the heating which is induced in the course of 5 minutes
in each of the calorimeters, whilst the water is being continually

258 LATENT HEAT OF STEAM

agitated. The temperature of the water of the calorimeters is
below that of the air, and consequently tends to rise by contact
with the surrounding air ; moreover, a certain quantity of heat
is constantly yielded to the calorimeters by internal conduction
through the appendices iif of the distributing cock. The
amount of these two quantities of heat is ascertained by an
observation.

This observation, combined with another previously made,
with respect to the heating of the calorimeters produced under
analogous circumstances — but merely by contact with the air,
the boiler not being heated — yields the necessary elements for
calculating the quantity of heat which passes by conduction
along the appendices i i', when the distributing cock is traversed
by a current of steam at 100°; so that this element of correction
is determined by the other experiments.

The distributing cock is then turned so as to make the
steam pass into the calorimeter C; the quantity of the water
condensed, being ascertained by the reading of the thermometer
of this calorimeter. When the requisite elevation of tempera-
ture has been obtained, the cock is closed by being brought into
its primitive position; the steam continues to pass through the
cock R, but the whole amount enters the condenser D, whilst
in the other position of the cock, a portion of this steam passed
into the calorimeter C. The proportion of steam which passes
into the calorimeter, may however be increased or diminuished
at will, and, consequently, also the time required for the water
of the calorimeter to rise the same number of degrees. If it be
desired to prolong this period, the cock Rx of the condenser is
left entirely open, whilst the distributing cock R is only par-
tially opened. If, on the contrary, it is required that the time
should be very short, the cock Rx is partially closed, and the
distributing cock R entirely opened. The total quantity of
steam, which passes into the apparatus in a given time, may
further be increased or diminished by increasing or diminish-
ing the fire under the boiler.

At the moment the cock R is closed, the time must be
noted, and the thermometer of the calorimeter C then immerged
in such a manner, that the column may only slightly project
beyond the stopper; the temperatures indicated by each of the

AT DIFFERENT PRESSURES. 259

thermometers of the two calorimeters being observed from
minute to minute. The times at which the observations are
made, are indicated by the chronometer which strikes minutes.

The mercury in the thermometer of C becomes stationary
after one or two minutes, and it then begins to fall. The
readings of this thermometer, as well as of the thermometer of
C, are continued for five minutes.

The water condensed in the calorimeter C is suffered to
flow, and is collected in a bulb placed below the cock rv and
having a thermometer with a very small cylindrical reservoir
suspended within it. As soon as the full current ceases to flow,
the cock rx is closed, the water in the bulb is agitated, and the
temperature instantaneously observed. The bulb is placed
below the cock, and the reading of the thermometers continued
for five minutes. At the end of that time, the cock rx is reopened,
and the few drops of water that have escaped along the walls
are collected again. The small quantity of liquid which moistens
the interior walls, is considered as being the same in all the
experiments, and, consequently, as forming a part of the ap-
paratus.

The quantity of water collected in the bulb must be weighed
in a scale, by an assistant.

The agitator must be continually worked by a uniform
motion during this interval, and it must still be moved in the
same manner during the second part of the experiment, when
the calorimeter C is in its turn in operation.

The fuel is replaced in the furnace, and when the distillation
has again become regular, which may be observed by the
manner in which the level of the water rises in the indicator
of the condenser, the thermometers of the calorimeters C and
C are simultaneously noticed, and the distributing cock R is
then replaced in the position in which it conveys the steam
into the calorimeter C. Tins is managed precisely in the same
manner as has already been described in the case of the calori-
meter C.

Finally, when the condensed water has been removed from
the calorimeter C, the simultaneous fall of the thermometers of
C and C is observed during five minutes, this cooling being
due to the excess of the temperature of the calorimeters over

S 2

260 LATENT HEAT OF STEAM

the surrounding air, notwithstanding the equal quantities
conveyed to them by conduction from the distributing cock R.

We will now consider the means, by which the principal
causes of error to which I have already alluded have been
obviated in this mode of operation; and how the experiment
itself furnishes the elements of the different corrections which I
have enumerated.

The object aimed at in the disposition of the interior tube,
which conveys the steam to the very centre of the boiler, and
then winds in the form of a worm more than four meters along
the interior of the boiler, is to avoid the escane of the water

3 J.

projected by the agitation of the liquid in the boiler. This
bubbling agitation must, however, be inconsiderable, as the
ebullition is effected under a regular pressure, unexposed to
those sudden alterations, produced at every moment in the
boiler of a steam engine by the play of the slide valves.

A thick investment of vapour passing directly to the con-
denser, and possessing the same temperature, moreover surrounds
and accompanies the efficient steam throughout its whole course,
even to the small appendices i i', which communicate with the
calorimeters, and are only four centimeters in length. By way
of greater security, all the metallic parts are enclosed in a
woollen covering of considerable thickness.

Finally, the perturbations of which I have spoken, as neces-
sarily present in the ordinary mode of operation, at the begin-
ning and end of the experiment, do not present themselves here,
since this experiment has neither beginning nor end; the steam
not being brought to the calorimeters until the distillation has
become regular, and all the parts have resumed their normal
temperature.

When the temperature of the calorimeter is higher than that
of the surrounding air, a certain quantity of heat is lost by
radiation, and also by contact with the surrounding air, but a
small quantity of heat is also gained by internal conduction
along the appendices itv, which join the calorimeters to the box
enclosing the distributing cock R.

If the law of Newton may be applied with sufficient
approximative correctness to the cooling of a body in more or
less agitated air, for the small excess of temperature which we

AT DIFFERENT PRESSURES. 261

obtain in our experiments, we may represent the sum of the
two first quantities of heat by an expression of the form

A6=A.6.Axi (1)

in which

6 represents the excess of the temperature of the calori-
meters over that of the surrounding air:

Ax the element of the time :

A a constant which must be determined by direct expe-
riments.

If, on the contrary, it be necessary to separate these two
quantities of heat and to apply to each its own law, we may
always set down the quantity of the heat lost by radiation as
equal to

A.O.Ax.

The cooling occasioned by the surrounding air will be ex-
pressed, according to Dulong and Petit, by the form

npcQ^.Ax,

in which/? represents the elasticity of the fluid; the exponent c
has the same value for all bodies, but changes from one gas to
another; n is constant for the same body, but changes with the
nature of the elastic fluid.

In our experiments we may suppose

wpc=B,

B being a constant quantity for the same state of agitation

of the air, so that the cooling produced by the surrounding air

will simply be

B 6}'233Ax.

Finally, the quantity of heat yielded by conduction for
a constant difference between the temperature of the dis-
tributing cock II and that of the calorimeter, is merely pro-
portioned to the time: it is represented by

KAx.

The first question to be solved is, therefore, to ascertain
whether cooling by radiation, and by contact with the sur-
rounding air, may be represented with sufficient exactness by
an expression of the form

Ad=A.0.Ax,

262

LATENT HEAT OF STEAM

or, if we must have recourse to a more complicated formula, as
A0= A0Ax+BF>™. Ax= A0(1 + ^-&>™)Ax, (2)

A J B

comprising two indeterminate constants, A and -r.

For this end I poured water, at higher and higher tempe-
ratures, into the calorimeters, and observed the cooling expe-
rienced simultaneously by the thermometers of the two calori-
meters at 30'; the water being continually agitated. The
rapidity of the refrigeration was assumed to be uniform during
this interval of time, whilst the temperatures of the calorimeters
and the surrounding air were supposed to be equal to the
means of those which had been observed at the beginning and
end of the 30', so that the refrigeration for 1' was obtained
by taking the A of the total refrigeration obtained.

Mean Temperatures.

Excess of the
Temperatures
of the Calori-
meters over the
surrounding
air.

Decrease of the

Value of A in the

■~ ^
Z S

£ <o

Of the
Air.

Of the
Calorimeters.

Temperature
during 1'.

formula
A0 = A. 6. Ax.

2
3
4
5

o
4-83
4-55
4-63
473
4-84

9°30

8 30

11-26

1495

19-57

o
9-95

8-87
14-74
1790
16-89

4-47

3-75

663

10-22

14 73

o
512

4-32
10-11
1317
12-05

0'00°586
000453
0-00817
0-01370
0-01983

0-00646
0-00498
0*01298
0-01897
0*01779

0-001311
0001209
0-001232
0-001340
0-001346

0-001262
0-001153
0*001284
0*001440
0-001476

Means

0-001287

0001338

The two last columns of this table give the values cal-
culated for the co-efficient A in the formula

AO-K.O.Ax.

It will be observed, that these values are not identical, they
do not, however, differ very much, if their extreme absolute
smallness be taken into account.

We may apply to these experiments the more complicated
formula

AT DIFFERENT PRESSURES.

263

calculating the two constants, A and -r-, according to two of

the observations, and see whether the three others give the
same value for A. I have thus found that the second formula
does not represent the observations better than the former,
and that the differences are owing much less to the inexactitude
of the formula (1) than to the changes which occur in the state
of the agitation of the air, and which cause the co-efficients to
vary in a very marked manner.

Several other series of observations, made in the same
manner, have led to the same result; the variations of A were
not sensibly greater when the formula (1) was applied to the
observations than when the formula (2) was used.

When the calorimeters exhibit a lower temperature than
the surrounding air, and, consequently, when they become
heated, the co-efficient A of the formula (1) is very different
from that which is applied in case the temperature of the
calorimeters is above that of the surrounding air. This will
be seen by the following table.

"S
B

Mean Temperature

Excess of
the Temperature
of the Calori-
meters over the
surrounding
air.

Variation of the

Value of A in the

W
o

(i CO

•° B.
II

Of the
Air.

Of the

Calorimeters.

Temperature
inl'.

formula
AS = A. 6. Ax.

1
2

12°20
12-20

20°17
1978

o
798

829

+7*97
+ 7-58

o
-4-22

-3-91

001114
001087

0-01005
000913

0°001398
0-001434

0-002570
0-002359

Means

0001416

0*002465

Thus, the co-efficient A has a much higher value in those
experiments in which the temperature of the calorimeter is
below that of the surrounding air, than in those in which its
temperature is higher.

I think I may conclude from all these observations, that the
formula (1) may be adopted in the calculation of the cor-
rections, with the condition that the value of the co-efficient A
shall not be fixed, but be determined, as far as possible, in each

264 LATENT HEAT OF STEAM

experiment by simultaneous observations, conducted under
the same conditions of agitation in the air. The general for-
mula which represents the refrigeration during the element of
the time, Ax, is, therefore:

-Ad=:A0.Ax-KAx,

A and K being co-efficients having special values for each
experiment. We shall see how they may be calculated by the
experiments themselves.

Let

t0 be the initial temperature of the water of the calorimeter C,
tY the maximum „ „ „

after the passage of the steam.

r the temperature of the surrounding air,
T the temperature of the steam in the distributing cock R,
t0' and tj* the initial and final temperatures of the ther-
mometer of the second calorimeter.

We will consider, at several separate periods, the double
experiment which I have already described.

First period. — The two calorimeters are at the temperatures
t0 and t0' below the temperature t of the air; steam passes through
the distributing cock directly to the condenser. The heating
experienced during these circumstances by the two calorimeters
during five minutes is then observed. If we assume the element
of the time, Ax, to be equal to 1 minute, and designate the -^th
of the refrigeration which takes place during 5 minutes by A6
and Ad', we have the two equations:

J0=A(t-*o)+K,
J^=A(t-0 +k. w

As the two calorimeters have sensibly the same tempera-
ture, and are arranged in a symmetrical manner with reference
to the distributing cock R, we may suppose that K has the
same value in the two equations; which will, therefore, suffice
for determining the two constants A and K.

Second period. — The steam passes into the calorimeter C,
whose temperature rises to the maximum tv The temperature
rises nearly regularly whilst the cock R is open, but as soon

AT DIFFERENT PRESSURES. 265

as this is closed it only rises very slowly, attaining its maximum
at the end of 2 or 3 minutes.

We will divide into three parts the interval of time which
passes between the observations of the initial and the final
temperatures.

1. The time m during which the water of the calorimeter
C rises from the initial temperature t0 to the temperature t
of the air; the amount of heat gained is represented by

(aI/+k), (4)

A and K having the values which have been determined by
the experiments of the first period. It would, however, be
better, to suppose that K alone was known by these experi-
ments, and to determine A by the simultaneous observation
of the heating of the calorimeter C, to which we may apply
the formula

J^m^^+KV (5)

2. The time mv which has intervened between the moment

when the calorimeter acquired a temperature equal to that of

the surrounding air, and the moment when the cock was closed,

and when the temperature of the calorimeter is nearly at its

maximum tv We will here suppose that the cooling is the same as

t — T
if the excess of temperature had been constantly -^ — - when we

should have for the cooling,

mI(AI-^I-K1> (6)

Ax has no longer the same value as in the first part; its value is
deduced from observations made during the third period.

With respect to the quantity K15 we will suppose it to be
proportional to the difference of the temperatures of the cock
and of the calorimeter, and consequently we shall admit it to be
as follows:

T tY+r
2

K having the same value as before.

3. The time m2, which intervenes between the closing of

266 LATENT HEAT OF STEAM

the cock and the observation of the maximum ; the refrigera-
tion is represented in this case by

A1 having the same value as in the preceding experiment, and
being calculated from the observations of the third period.

Third period. — The condensed water has been removed from
the calorimeter C ; the simultaneous refrigeration of the two
calorimeters is then observed during five minutes. We thus
have for the calorimeter C,

J^A^-^-K^1, (8)

from which we deduce the value of Ax required in the formulas
of the second period.

The calorimeter (7 is in conditions very nearly similar to
those which occurred in the first period ; thus we have

Jd>^A'(t0'-T) + K. (9)

K may be considered as the remaining constant ; A' will differ
but little from A, but it is requisite to deduce its value from
the preceding equation, and use it in the following period in
which the calorimeter C is brought into action.

Fourth period. — The steam passes into the calorimeter C,
the calorimeter C being in a passive state.

The interval of time comprised between the opening of the
cock, and the observation of the maximum temperature may be
further subdivided into three parts.

1. The time m', during which the water of the calorimeter
C rises from its initial temperature t'0 to the temperature t of
the surrounding air; the amount of heat gained during this
period is

™/(A'-^-0 + K). (10)

2. The time m\, intervening between the moment at which
the water of the calorimeter rises to that of the surrounding
medium, and that in which the regulator is closed, and the
temperature is near its maximum; the loss of heat is re-
psesented by

*, + T , (11)

))l

T- l \
(A' *'-t K 2 )'

AT DIFFERENT PRESSURES. 267

The value of the co-efficient A\ is deduced from the obser-
vation of the simultaneous refrigeration of the calorimeter C,
which yields the equation,

^Bl'.ft-Tj-K^. (12)

3. The time m'2 which passes between the closing of the
regulator and the moment in which the maximum of the tern-
perature of the calorimeter C is observed; the loss of heat
during this time is

^(A'^-T)-]!^), <13>

A\ having the same value as above.

Fifth period. — The water has been removed from the calori-
meter C ; the sinking of the thermometers of the two calorime-
ters is observed for 5', both calorimeters containing water at a
higher temperature than that of the air. The formulae which
apply in this case to the cooling in 1/ are,

T-t
for the calorimeter C, AO^. A, (tx — t) — K ^ — -S

(14)

These two expressions may serve to calculate new values for
A\ and K, which may be compared to those previously obtained.
It is evident that the values of t0, t0'9 tv t{> r change in a slight
degree during the course of these experiments; in each case
they are taken as yielded by direct observation.

Finally, we have already said, that from the moment at
which the thermometer of the calorimeter attained its maximum,
the same thermometer was observed from minute to minute,
until the removal of the water which was yielded by the con-
densation of the steam. Suppose these observations to be
made during m3 minutes, it is evident that if the water con-
densed were at the same temperature as the water of the calori-
meter, the refrigeration would be the same as in the observa-
tions made after the removal of the water : according to which
we should have for the refrigeration

j*(a1(*1-t)+kJ^).

268 LATENT HEAT OF STEAM

By deducting from the value of this expression the actual
refrigeration observed, we obtain the small quantity of heat
yielded by the condensed water to the calorimeter during this
interval of time : and this we will represent by q.

We have thus definitively for the small quantities of heat
which must be added to the increase of temperature, {t1 — t0),
(^1 — *o ')> by direct observation.

for the calorimeter C,

SJ0= _„,(aI^ +K)+mi(vA1V-K-rrJ

+m2(A1(/1-T)-K^i)+?.
for the calorimeter C,

XA*= -^(Al^Lo +k) +^A^-K-t^J ,

(15)

(16)

In a word, the principle of this method consists in deter-
mining the corrections which must be applied to the calorimeter
in actual operation, based upon observations simultaneously
made on the second calorimeter, which is in a passive con-
dition, that is to say, subject to the same circumstances of
external perturbation, with this difference, that it does not
receive any steam; so that the variations are only produced by
disturbing causes, which act simultaneously on the first calo-
rimeter.

The temperature O of the condensed water is observed in
the bulb o immediately after its escape from the calorimeters.
This temperature always differs, in a slight degree, from the
maximum temperature of the calorimeter, but the difference
rarely amounts to 1°.

The observed temperature O is not absolutely exact ; it is
evident that it sinks a little during the flow of the water.
Some direct experiments, made under circumstances as nearly
as possible similar to those existing in actual experiments, have
shown that this decrease of temperature varied from 0o,10 to

AT DIFFERENT PRESSURES. 269

0o,20, according to the excess of the temperature over the sur-
rounding air. I made this small correction in each case for the
observed temperature ©; the uncertainty resulting from this cir-
cumstance is, however, wholly insignificant; and it easy to show,
that it cannot induce an error of 20,000 on the latent heat
sought *.

It is often necessary in practice, owing to the small value
of these corrections, to deviate, in a slight degree from the
method that I have described in the calculation of the correc-
tions. Thus the equation 3 and the equation 5 rarely differ
sufficiently, in a numerical point of view, to be considered as
distinct equations, and to serve for the determination of the
two constants, A and K. The same may be said of equations
12 and 14, when compared together. But as we may suppose,
without any sensible error, that K is constant for excesses of
temperature T — t that vary but inconsiderably, it is best to
choose amongst all the equations, deduced from experiments
made under the ordinary pressure of the atmosphere, those
which appear most suitable for determining this quantity, and
adopting the mean of the values deduced from them for all
the experiments. In this manner it has been found, that
K=0°-0040 for a difference of temperature T-t=8S°.

The value of K being once determined, the equations 3, 5,
and 8 yield, in each experiment, the special values of A and A7.

The values of A and A' oscillate about 0°-0022.
Ax and A\ „ 0°-0014.

* I had proposed to determine directly the difference of temperature Q — t1
existing at the moment in which the final temperature of the calorimeter is
observed, between the condensed water and the surrounding water. With
this view, I placed a thermo-electric element of iron and copper in the centre
of the bulb B ; the second segment of the element being placed in the water of
the calorimeter and the copper thread which connected the two segments passed
over a very sensible galvanometer. To effect this, it was, however, necessary
to pierce through the bulb B, and close the thermo-electric element hermeti-
cally into the opening. This circumstance would not have presented any
difficulty, if the apparatus had only been intended for experiments made under
the ordinary pressure of the atmosphere, but it required a special joint for
experiments made at a high pressure, and my endeavour was to diminish the
number of the joints used, as much as possible. I was also apprehensive that
the indications of the thermometric element might be inexact under the con-
ditions in which it would have to be placed.

270 LATENT HEAT OF STEAM

If it were admitted in the formulae 15 and 16 that the
values of A, Ap and K were constants for all the experiments,
we should find for the corrections, values which scarcely differ
by 0o,01 from the more exact values obtained by taking into
account all the circumstances that I have enumerated.

The weight of the copper calorimeters, with their worms,
the agitator, and interior metallic joints, is:

Calorimeter C 15,615 grras.

C 15,542 „

If we admit the specific heat of copper to be =0*0951
{Annates de Chimie et de Physique, 2me serie, tome lxxiii., p. 37, ">
we shall obtain for the value in water:

for the calorimeter C 1482*8 grms.

C 1475*7 „

The quantity of water introduced into the calorimeters
for each experiment always presents the same volume, measured
by the gauging vessel, H. The weight of the water which fills
the vessel to the marked line has been determined by the most
carefully conducted experiments.

In a first experiment, this weight was found to be equal
to 65,039*4 grms., the temperature of the water being at 12°,06.

In a second experiment, made at a temperature of 12°00,
the weight was found to be 65,046*3 grms. I have adopted this
second value, which differs but little from the first, and presents
greater guarantees for its accuracy.

It was, however, necessary to determine exactly the weight
of the water which filled this gauging vessel, at the different
temperatures observed during the experiments. It is easy to
determine this weight by calculation, when the dilatations of the
water and the iron plate are known. The dilatations of the
water may be ascertained with precision, by means of the con-
cordant experiments of M. Despretz and M. Pierre. {Annates
de Chimie et de Physique, 2me serie, tome lxx, p. 47, and Annates,
3me serie, tome xv., p. 350.)

Hitherto, the coefficient of the linear dilatation of the iron
plate has alone been determined, and I do not think that the
coefficient of cubic dilatation can be deduced from it, so as to
apply to an iron plate vessel of great capacity, and whose walls

AT DIFFEEENT PRESSURES.

271

have only an inconsiderable degree of thickness. I have given,
in another Memoir, the object of which is to determine the
calorific capacity of water at different temperatures, a series
of direct experiments, from which I have obtained for the
co-efficient of the cubic dilatation of vessels of galvanised iron
plate, the number o-=00000305.

The following table gives, in the first column, the capacity
of the gauging vessel at different temperatures; its volume at 0°
being supposed equal to 1.

The second column gives the weight of water enclosed at
different temperatures in the vessel; the water having the
density which it actually manifests at these temperatures.

The third and fourth columns give the weight of the water
at different temperatures increased by the value in water of the
calorimeters in respect to calorific capacity.

Weight of the Water

Tempera-
ture.

Capacity

of the

Gauging Vessel.

Weight

of the Water

enclosed.

increased by the Value in
Water of the Calorimeters.

Differences
for 1°.

Calorimeter C Calorimeter C'

Grms.

1-0000000

65045-0

66527-8

665207

....

1*0000305

65050-4

66533*2

60520- 1

+ 5*4

1-0000610

650550

66537-8

005307

+ 4-6

1-0000915

65058-6

66541-4

66534-3

+ 3*6

1-0001220

65061-1

66543-9

665368

+ 2*5

1-000] 525

65062-5

665453

66538 2

+ 1-4

1-0001830

65003-0

66545-8

665387

+ 0*5

1-0002135

650624

60545-2

66538-1

-0-4

1-0002440

65060-8

665436

665365

-1-6

1-0002745

650592

665420

60534-9

-1-6

1*0003050

65055-5

66538-3

66531-2

-3-7

1 0003355

650516

66534-4

60527-3

-3*9

1-0003660

650463

06529-1

66522-0

-5-3

1-0003965

650407

66523-5

605164

-5-6

1-0004270

65034-4

06517-2

66510-1

-6-3

1-0004575

650260

005088

60501-7

-8-4

The mercurial thermometers of the calorimeters have been
graduated with the greatest care.

A centigrade degree occupies, on the stem of the thermo-
meter of the calorimeter C, a space of 18D*7620; consequently,
1D has the value 0°-053283; whilst, for the calorimeter C, the

272 LATENT HEAT OF STEAM

same quantity occupies 18D,5800; and, therefore, 1D has here
the value 0°-053821.

It is easy to distinguish with certainty the tenth of these

divisions, that is to say, — j^ of a centigrade degree, by means
of the horizontal lenses with which these thermometers are
observed.

I have collected, in the following table, all the experi-
ments that have been made under the ordinary pressure of the
atmosphere.

This table consists of 18 columns.

In column 1 are inscribed the numbers of the succession of
the experiments.

In column 2 the calorimeter is indicated with which the
experiment was made.

Column 3 gives the weight of the water introduced into
the calorimeter, increased by the value in water of this calori-
meter.

Column 4 gives the initial temperature, t0, of the water in
the calorimeter.

Column 5 its final temperature, tv that is to say, its maxi-
mum temperature.

Column 6 the observed elevations of temperature, t — tQ.
These have been corrected, with respect to different causes of
perturbation by the method already described.

Columns 7, 8, 9, and 10 give the elements of these corrections.

Column 1 1 gives the corrected values of t1 — tQ.

Column 12, the weight of the water condensed in the
calorimeter.

Column 13, the temperature of this water on its escape
from the calorimeter, after the application of the small cor-
rection that I have already indicated.

Column 14 gives the elastic force of the steam, deduced
from the observation of the barometer in the external air, the
values being given in millimeters of mercury.

Column 15 gives the same in atmospheres.

Column 16, the temperature of the steam deduced from the
height of the barometer.

Column 17, the temperature of the steam observed on the
thermometers of the boiler.

AT DIFFERENT PRESSURES. 273

The temperatures observed in column 17 are generally a
little higher than those calculated in column 16, this being
necessarily the case, since, in order that the distillation of the
water may proceed with activity, the elastic force of the steam
in the boiler must be a little above the external pressure.

Column 18 gives the total heat yielded by the steam on its
condensation to the condition of liquid water, and its subsequent
refrigeration to 0°.

The six first must be considered as preliminary experiments,
which have purposely been conducted under the most unfavour-
able circumstances.

Thus, in experiment No. 1, the fire under the boiler was
very small; the introduction of steam into the calorimeter C
was continued for 21 minutes.

In experiment No. 4, the introduction likewise lasted 22
minutes.

In experiment No. 3, the fire under the boiler was, on the
contrary, very large, but the distributing cock E-j was shut
and opened ten times during the experiment, in order to see
whether this circumstance would occasion any sensible per-
turbation during the experiment.

In experiment No. 5, the fire was small, but the cock H19
which conveys the steam to the condenser, was completely
closed, so that all the steam of the boiler passed into the calo-
rimeter.

In experiment No. 6, the cock ~RX was continually opened
and closed.

It will be observed that these different circumstances exer-
cised only a very slight influence on the value of the total heat,
having merely rendered it a little smaller.

The 38 other experiments gave values, the extremes of
which are 635*6 and 638*4.

The general mean is 636*67.

274

LATENT HEAT OF STEAM

TABLE I. — Experiments under the

Time

Time in-

Number

Weight of

Mean

during

tervening

of the

Calori-

the Water

Initial

Final

h — h

Tempera-

which the

before the

Experi-

meters.

of the

Tempera-

Observed.

ture of the

Steam

fall of

ments.

Calorimeter.

ture /0.

ture fj.

Air t.

was in-
troduced.

the Tem-
perature.

]

Grms.

66524 0

12-81

21-53

87216

14-61

21'

23'

665204

1340

24-84

11-4409

16-02

66534 4

11 00

22-82

11*8208

1365

66534-4

11-00

21-96

10*9603

11-85

66538-6

6-19

18-28

12*0918

512

66538-1

6-96

1994

129847

625

665235

1300

24-34

11-3414

1615

66523*1

13-09

24-42

1 1-3303

16-19

66523-3

1305

24-44

11-3967

1523

66522-5

13-27

24-62

11-3524

15 30

I9|

20|

66535-4

10-76

22-48

U-7196

1315

665273

11-02

23*07

12*0554

14-21

665262

11-26

23*15

11-8980

1426

10|

66534-4

11-00

22*77

11*7675

14*23

665262

11-27

21-63

10-3627

14 34

66535-2

10-83

22-53

11*7063

12 18

665267

11-14

2380

12-6594

1328

665276

10-91

23-55

12*6453

13-55

665334

11-23

21-36

10*1317

13-70

665340

11-05

21*39

103422

13 70

665262

11-29

24-41

13*1180

13-57

66526-2

11-26

24*39

13*1301

11-93

10i

66535-0

10-88

22*10

11*2214

1230

66527 0

11-04

23-68

12*6372

12*30

66534*9

10-90

22*22

11-3226

12*11

66527- 1

11-05

23*21

12*1585

11*97

66545-8

6-03

18*79

12*7612

4 33

66545-6

6-30

18-01

11*7089

5*67

66538-5

6-51

19-37

12*8594

6*04

66545-3

6-82

19*29

12-4736

6-10

665442

4-13

16-79

12*6573

3-13

66536 9

4-04

21*97

17*9331

3-68

66544*9

464

18-03

13 3873

2-66

665379

4-89

19-67

14-7792

283

66545-6

567

18-16

12*4922

4-87

665378

4-78

21*06

16*2835

5-40

665456

5-64

18*23

12*5908

5 29

665386

5-58

20*10

14-5209

668

66543-3

3-77

16*98

13-2142

412

66537*1

4-04

18-99

149542

4*65

66544-2

4-12

17*19

13 0703

5- 0

66537-8

4-75

20-32

155677

552

66545-0

4-81

18*48

13-6724

5-79

665384

5-19

19*86

14-6689

605

AT DIFFERENT PRESSURES.

275

Ordinary Pressure of the Atmosphere.

Pressure

Temperature of

Heat

Weight of

Tempe-

of the Atmosphere.

the Steam.

Total

by Con-

'i — 'o
Corrected.

the Water
Con-

rature
of this

Heat of

duction

the

during 1'.

densed.

Water.

In milli-
meters.

In Atmo-
spheres.

Cal-
culated.

Observed.

Steam.

Grins.

mm.

Atm.

0*0040

8-7441

949T0

21-50

74652

0-983

99-49

633-3

11-4829

1255*15

24-80

746-55

0-983

99-49

6341

11-8532

128788

2285

767*06

1 009

100-26

100-58

635-2

11 0485

120079

21-98

76594

1008

100-22

100-25

6340

12-1728

131685

770-13

1-013

100-37

10009

6334

13-0710

141345

768-37

1-011

10031

100-18

6354

11-3471

123M0

2371

74643

0-982

99-49

6358

11-3356

1230T0

2338

74571

0-981

9946

6363

11-4284

1240-23

23-53

74129

0-975

99*31

636-4

113744

1233-05

24-16

74082

0*975

99-28

637*6

11*7391

1272-84

22 37

765 19

1-007

100-19

100-44

636-0

12-0794

1308-85

23-00

765*19

1-007

100-19

100-56

636-8

11-9100

1287-65

23-03

765-23

1007

100-19

100-44

6383

11-7908

1275-57

22-85

765-28

1007

100T9

10049

6379

103542

1121-04

2163

765-20

1007

100T9

100-19

6359

11-7613

1274-57

22*22

767-00

1009

100-26

10053

635-9

12-6883

137161

22-38

76703

1-009

10026

100-58

6379

12-6666

1369-23

2261

76712

1-009

100-26

100-58

6379

10-1540

1099-03

20-92

76702

1-009

10026

100-38

635 6

10-3599

1120-99

21-00

767-00

1-009

10026

100*36

635-8

131368

1425-42

23-81

767-09

1009

10026

100-56

636-7

13- 1 702

1425-58

23-00

765-87

1-008

100-22

100-47

6376

11-2514

1213T9

2145

765-72

1-008

100 22

100 63

638-4

12-6904

137668

23-67

765*90

1-008

10022

100-70

636-8

11-3640

123003

2200

765-92

1-008

100-22

10067

6366

12-2085

132187

22-85

76585

1-008

100-22

10058

6372

12-8426

138470

18-79

77010

1-013

100-37

100-40

636 1

11-7818

126675

18-00

76850

1-011

100-32

100-15

6367

12-9476

1395-18

19-37

768-47

1-011

100-32

100-13

637*3

12-5640

135470

1932

768-32

1-011

10031

10020

636 1

12-7499

136922

1603

766- 19

1-008

100-22

635-6

18-0644

195G-04

2237

76624

1-008

10022

,...

6369

13*5057

145509

18-36

767-15

1009

100-26

635-9

149242

1612-69

2019

767-23

1-009

10026

635 9

125778

135720

1889

73576

0968

99*09

6357

163707

177229

2150

735-76

0-968

99-09

6361

12-6746

136338

18*02

735-09

0-967

9907

636-6

14-6091

1575-64

20-19

73509

0*967

99-07

636-9

132922

143145

18-16

74287

0977

99-36

6361

150354

1619-65

19-18

74287

0-977

9936

6368

13-1655

141395

16 86

742 08

0-976

9933

637*3

156584

169T20

2019

742-05

0-976

99 33

636-4

137426

148310

19*03

74053

0-974

99 27

6357

147505

1590 50

1975

740-53

0-974

9927

6368

T 2

276 LATENT HEAT OP STEAM

II. Total heat of Steam under Pressures greater than the
ordinary Pressure of the Atmosphere.

I have so arranged my apparatus, that the experiments on
steam at high pressure might be conducted under circumstances
precisely similar to those occuring where the experiment is
made on steam engendered under the ordinary pressure of the
atmosphere. The pressure of the external air is then replaced
by that of an artificial atmosphere, which may be varied at will.
This atmosphere has a considerable volume and is kept at a
constant temperature, so that its elasticity does not change
sensibly during the course of one experiment.

Air is compressed into the receiver E F by a forcing-pump,
in order to obtain the pressure desired for the experiment, this
pressure being measured by the mercurial manometer. The
ebullition of the water in the boiler is as regular as if effected
under the ordinary pressure ^of the atmosphere. These experi-
ments are conducted exactly in the same manner, and I have
nothing to add on this subject to the remarks I have already
made in the preceding pages. The manometer is observed at
two periods during the course of the experiment; the first
time some moments after the opening of the cock R, and again
some moments after its close. The pressure, under which the
steam is distilled, is taken as the mean of these two obsevations,
which never differ very much from one another.

The corrections, which it is necessary to adopt for the eleva-
tions of teperature observed, are determined in the same manner
as in the experiments made under the pressure of the atmos-
phere. The elements of these corrections are given by the
experiments themselves, as has already been remarked.

The water yielded by the condensation of the steam is pro-
jected with force, at the moment in which the cock rx of the
calorimeter is opened, in order to collect this water in the bulb
O, in which it is to be weighed. The time occupied by the
flowing of the water is very short, and on this account, the
water must experience a smaller loss of heat than when the
experiments were made at the ordinary pressure of the atmos-
phere. I have, however, assumed, as in the last experiments,
that the water lost about 00,2 during its passage, because it is

AT DIFFERENT PRESSURES. 277

now subject to a new cause of diminution of heat. This cause
arises from the steam, while it is condensed in an atmosphere of
compressed air, dissolving a larger quantity of this air than it
is able to hold in solution when it reaches the bulb. A number
of small vesicles of air are disengaged, which render the water
turbid for some instants, and necessarily bring with them a
certain quantity of heat.

The experiments progress very easily, and with remarkable
regularity, up to a pressure of 10 atmospheres, the results pre-
senting the same precision as those yielded by experiments
under the pressure of one atmosphere. Above 10 atmospheres
the experiments become more difficult of prosecution; it is
necessary to keep up a very strong fire under the boiler, the
different parts of the apparatus show much strain, and they
require constant inspection, in order to make it certain that
no flaw exists in any p art.

Every morning, water that had been distilled the evening
before was poured into the boiler, the whole with what had
remained, amounting to about 150 liters. This quantity of
water was found sufficient, by managing the fire properly during
the interval of the experiments, to make from four to six con-
secutive determinations. By making use of distilled water only,
the encrustation of the boiler was avoided, which would other-
wise have taken place in a short time.

After the last experiment of the day, the cock R4 of the air-
receiver was closed, whilst the water was still in a state of ebul-
lition in the boiler, in order that the air compressed within it
might be kept for the next day. The only loss of air was
therefore the small quantity contained in the condenser, and in
the pipes of the apparatus ; this air escaped on the cocks rY
of the calorimeters being opened. The following morning, the
screw-peg was removed from the boiler, and the necessary
quantity of water introduced.

AVhen the pressure of the steam exceeds 10 atmospheres,
the joints of the apparatus become rapidly deranged, and it is
necessary frequently to renew the twist of hemp, steeped in
caoutchouc, which is passed round the stem f of the distribu-
ting cock, because the hemp is rapidly consumed by the high
temperature of the steam. It is also found requisite frequently

278

LATENT HEAT OF STEAM

to repair the joints connecting the distributing cock to the
worms of the calorimeters. These joints are much worn by the
violent and sudden variations of temperature to which they are
exposed.

The experiments admitted of being prosecuted with perfect
success under pressures equal to 14 atmospheres. I even began
a series of experiments under a higher pressure, but the boiler
was seriously injured, many crevices were found in the joinings
of the iron plate, and numerous flaws round the joint L L/ of
the cover ; it was consequently found necessary to suspend the
experiments.

I hoped, however, to be able to obtain determinations
under the pressure of 20 atmospheres, by refitting all the joints

TABLE II

— Experiments under Pressures

WTeight of 1

Time

Time in-

Numbei

the Water

Initial

Final

Mean

during

tervening

Heat

of the

Calori-

of the

Tempe-

h-h

Tempera-

whieh the

before the

by Con-

Experi-

meters.

Calorime-

rature

Observed.

ture of the

Steam

fall of

duction

ments.

ter.

tQ.

'l-

Air r.

was in-
troduced.

the Tem-
perature.

during 1'.

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

(10)

Grms.

66537-0

10-32

23-77

13-4566

11-38

0-00578

665369

10-37

19-72

9-3511

11-65

0-00578

66537-4

10-17

21-79

11-6237

11-32

0-00578

66545-8

607

18-09

12-0180

4-75

0-00600

66538*5

6-21

21-73

15-5178

5-13

000600

66545-8

5-94

18-27

12-3270

4-98

0-00600

66538-5

6-23

20-40

14-1684

5-40

0-00600

66527-1

12-35

24-03

11-6756

13-73

0-00600

66526-6

12-47

24-28

11-7971

14-12

0-00600

66523-0

12-82

24-39

11-5651

15-37

0-00600

66523-0

12-79

24-42

11-6343

15-30

0-00600

66525-2

12-60

24-22

11-6233

15-73

0-00600

66545-4

5-24

18-16

12-9158

4-44

0-00600

66538-5

5-40

19-39

13-9947

5-06

0-00600

66538-7

5-81

21-20

15-3925

5-60

0-00600

66545-7

6T1

17-55

11-4399

5-83

0-00600

66538-4

6-38

19-58

13-2069

6-26

0-00600

66545-6

5-50

16-94

11-4452

4-97

0-00600

66544-9

4-81

16-99

12-1778

4-98

000650

66537-1

4-10

16-76

12-6655

5-10

0-00650

66541-4

301

16-21

11-1974

4-77

0-00650

66537-9

4-85

17-68

12-8271

4-51

0-00650

66534-4

10-99

22-51

11-5238

13-70

0-00620

66538-5

5-40

1906

13-6634

5-25

0-00650

66528-3

12-11

23-85

11-7393

14-01

0-00620

AT DIFFEEENT PEESSUEES.

279

of the boiler ; but the high pressure to which the boiler had
long been exposed, had driven the pegs so deep into the cast-
iron cover, that it was impossible, even with the greatest exer-
tions, to remove them, the greater part breaking before they
could be extracted. The apparatus could not be repaired with-
out very great expense, whilst there would always have been
considerable risk in subjecting it to a more considerable pressure.

The pressure of 14 atmospheres, which was employed in
these experiments, exceeds, however, very considerably the
pressure obtained in a steam-engine, which rarely exceeds 5 or 6
atmospheres.

I have collected in Table II., 73 experiments, made under
pressures varying from 1 to 14 atmospheres.

Greater than that of the Atmosphere.

Pressure of the

Temperature of

Total heat

Steam.

the Steam.

Subtract-

Weight of

Tempe-
rature

Total
Heat of

ing the

/j — 10 the Water

Tempera-

Corrected. Con-

of this

Calculated

Observed

the

ture of

densed.

Water.

In Milli-

In Atmos-

by an Air

on the mer-

Steam.

the

meters.

phere*.

Ther-
mometer.

curial ther-
mometer.

Steam in
column 16.

(11)

(12)

(13)

(14)

(15)

(16)

(17)

(18)

(19)

Grms.

mm.

Atm.

13-5190 1456-27

24-68

1448-

1-905

11925

119-46

642-3

523-0

9-3563 1000-48

19-72

1462-

1-924

119-60

119-53

641-8

522-2

11-6780 j 1253-75

22-36

1582-

2-083

122-17

122-31

642-2

520-0

12-0700 1282-58

18-12

1742-

2-293

125-2

125-30

643-9

518-7

15-6032 1669*45

21-73

1768-

2-327

125-5

125-34

643-6

5181

12-3871 \ 1315-25

18-16

1849-

2-433

127-2

12713

644-8

5176

14-2464 ! 1517-46

20-42

1952

2-568

1290

128-97

645-1

5161

11-6942 1 1247-20

25-27

2285

3-007

134-4

6490

514-6

11-8112 l 1262-70

25-27

2273

2-991

134-2

6475

513-3

11-5746 l 1234-23

24-83

2335

3-072

1351

648-5

513-4

11-6347 1 1238-80

24-39

2325

3-060

1350

649-1

514-1

11-5833

1235-50

23-96

2340

3-080

135-2

647-6

512-4

129977

1376-53

18-76

2365

3113

135-5

135-54

647*0

511-5

14-0530

1488-95

1942

2370

3119

135-7

135-65

6473

511-6

15-4585

1641-90

21-25

2426

3-193

136-4

647-6

511-2

11-4794

1212-75

17-59

2498

3-288

1375

137:49

647-4

509-9

13-2781

140772

19-69

2517

3313

1377

137-71

647-2

5095

11-5038

1211-70

16-84

2588

3-394

138-6

138-53

648-4

509-8

12-2258

1287-05

17-32

2842

3-739

142-0

14211

6492

5072

12-7136

1339-02

17-29

2860

•71

3-764

142-2

142-56

648-9

506-7

11-247.^

1184-84

16-19

2911

•75

3-831

142-5

142-54

647-8

505-3

12-8699

1353-60

17-92

2955

•66

3-889

1434

14341

650-3

506-9

11-5008

1221-20

22-95

3042

•51

4-003

144-3

649-4

5051

13-7108

1445-32

18-74

3049-85

4-013

144-3

144-31

649-7

5054

11-7345

1246-50

24-54

3116-00

4100

145-3

6510

505-7

280

LATENT HEAT OF STEAM

TABLE II.—

Number
of the
Experi-
ments.

Calori-
meters.

Weight of
the Water

of the
Calorime-
ter.

Initial
Tempe-
rature
t0.

Final
Tempe-
rature

h-f0
Observed.

Mean
Tempera-
ture of the

Air r.

Time
during
which the
Steam
was in-
troduced.

Time in-
ter ver in g
before the

fall of
the Tem-
perature.

Heat
by Con-
duction
during 1'.

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

(10)

Grms.
66528-0

12°22

24°35

12-1318

14?28

0?00620

66527-6

2920

0-00620

66537*8

4464

000620

66531-8

2920

0-00620

66533-0

2881

13*

0-00620

66534-3

2500

0-00620

66537-5

9206

/ 2

0-00670

66544-6

1360

0-00670

66542-5

7826

0-00700

66537-7

2507

0-00700

66537-3

1563

0-00700

66543-2

4869

0-00700

66544-5

0073

0-00700

66541-0

0090

0-00805

65542-0

3864

7A
• 2

13£

0-00805

66541-4

2079

0-00805

66544-5

2284

•69

0-00810

66537-9

•3663

•68

0-00810

66537-5

•45

9971

•18

0-00830

66545-5

•75

•3270

•59

0-00830

66538-6

•65

•9179

•82

0-00830

66544-2

•13

•49

•3581

•96

0-00850

66537-5

•07

•5482

•44

0-00850

66538-7

•47

•3808

•35

0-00850

66545-7

•69

•59

9034

•20

0-00850

66537-5

•09

•6249

0-00850

66544-1

•10

•02

•9184

•36

0-00850

66536-4

•86

•71

•8492

•26

0-00860

66543-9

•00

•9531

•83

0-00900

66545-0

•83

•74

•9105

•46

0-00900

66538-2

•06

11)

•57

■5118

•73

0-00900

66543-8

•95

•97

0214

•55

000950

66544-6

•60

•92

•3207

•82

0-00950

66538-2

•98

•1818

•07

0-00950

66537-7

■67

•55

•8788

0-00950

66545-4

■17

•04

•8732

•62

0-00950

66538-6

•15

•31

•1.603

0-00950

66538-6

'76

6275

0-00950

66545-4

•62

4521

0-00950

66538-5

1764

o-oiooo

66545-4

•9042

o-oiooo

66538-4

8384

0-01080

66544-9

2558

0-01080

66538-4

5785

0-01080

66545-5

41 13

8029

0-01100

66544-4

14-

1999

001100

66538-4

1920

o-oiioo

66545-8

5-93

20-04

14-1093

9-70

0-01100

AT DIFFERENT PRESSURES.

281

Continued.

h-t0
Corrected.

(11)

12
12
11
12
12
12
10
14
12
13
14
12
12
13
14
14
12
14
9
12
13
13
14
11
11
14
12
13
12
12
14
14
13
14
14
12
14
13
15
14
13
14
14
15
13
14
16
14

•1253
•2711
•4335
•2756
•2761
•2523
•9170
•2327
•7840
•3174
•1941
•5336
•1007
•0276
•4533
•2382
•2476
•4008
•9838
•3514
•9473
•3434
•5716
•4135
•9120
•6566
•9418
•9066
•9904
•9349
•5565
•0992
•4219
•2353
•9825
•9428
•2713
•6712
•5494
•2028
•9510
•8716
•3140
•6185
•8296
•2716
•2719
•1389

Weight of
the Water
Con-
densed.

(12)

Tempe-
rature
of this

Water.

(13)

Gnu

1291

25-

1307

24-

1208

21-

1305

24-

1299

24-

1296

23-

1144

15-

1499

18-

1338

15-

1402

18-

1490

18-

1313

16-

1269

17-

1360

15-

1518

18-

1491

17-

1280

16-

1509

18-

1038

14-

1292

17-

1459

19-

1390

17-

1522

19-

1190

17-

1241

17'

1530

19-

1347

17-

1449

17-

1353

17-

1347

17-

1520

19-

1457

18-

1387

18-

1471

19-

1552

20-

1337

18-

1478

20-

1413

19"

1614

20-

1465

20-

1440

20-

1537

20'

14 74

19'

1617

21-

1427

21-

1479

22-

1585

23-

1456

20-

12
54
79
13
12
96
59
70
84
36
59
42
32
84
16
52
71
89
77
77
65
14
07
52
20
05
20
73
60
71
57
59
59
20
00
88
11
75
92
22
42
60
75
73
66
13
33
48

Pressure of the
Steam.

In Milli-
meters.

(14)

3128-
3149-
3223-
3323-
3437'
3565'
3883-
3945-
4045-
4067-
4068-
4070-
4115-
4195-
4268-
4350-
4643-
4653-
4821-
5182-
5212-
6127
6287
6298-
6329-
6368-
6401'
6478-
6483-
6702-
6728-
7350-
7416-
7420-
7465-
8056-
8106-
8131-
8138-
8550-
8563-
8925-
8990-
9004-
10141-
10193-
10332-
10354-

In Atmos-
pheres.

(15)

•00

•25

•09

•69

•85

•81

•14

•55

•81

•44

•52

•06

•56

•09

•15

*75

•20

•47

•67

•61

•49

•21

•87

•76

•81

•35

•83

•59

•02

•65

•62

•28

•49

10-

•48

10-

•26

10-

•24

lO-

•41

ll-

•30

11-

•38

11-

•73

11-

•86

11-

•52

lS-

•27

13-

■38

13-

•84

13-

Atm.
T16
T44
•241
•373
•523
•692
T09
T91
•335
•352
353
•357
•415
•520
•616
•724
•109
T23
•344
•818
•858
•062
•273
•287
•328
•380
•423
•524
•530
•819
•853
•671
•759
•764
•822
•600
•666
•699
•708
•250
•267
•744
•830
•848
•344
•412
•595
•625

Temperature of
the Steam.

Calculated
by an Air

Ther-
mometer.
(16)

145
145
146
147
149
150
153
154
155
155
155
155
155
156
157
157
160
160
161
164
164
171
172
172
172
173
173
173
174
175
175
179
179
179
180
183
183
183
183
186
186
1K7
188
188
193
194
194
194

Observed
onthemer
curial ther-
mometer.
(17)

153-62
154-34

155-39

160-74
160-77
162-34
165-06
165-40
173-03
174-17
174-96
174-96
174-92
174-84

Total

Heat of

the

Steam.

(IS)

Total heat
Subtract-
ing the
Tempera-
ture of
the
Steam in
column 16.
(19)

649
649
651
649
652
652
650
650
651
650
652
651
651
652
651
652
653
653
654
653
655
655
655
655
655
656
656
655
656
656
656
662
662
662
662
662
662
662
661
664
664
664
665
6<i4
666
66 J
665
666

504
503
504
502
503
502
496
496
496
494
496
495
495
496
494
494
492
493
492
489
490
483
483
482
482
483
482
482
482
480
480
583
482
483
482
479
479
479
478
478
478
476
4 77
476
472
470
170
471

282 LATENT HEAT OF STEAM

This table presents the same arrangement as Table I., and
we will therefore refer our readers to the explanation given,
with reference to the latter. We will only observe, that the
elastic forces indicated in column 14, are the heights of the
mercury observed on the manometer, and increased by the
height of the barometer.

Table II. contains one column more than Table L; this
column, which is headed with the number 19, comprises the
numbers obtained by subtracting from the total heat inscribed
in column 18, the temperature of the steam given in column
16. We shall require these numbers to compare our expri-
mental determinations with the results deduced from the law of
Southern.

Column 16 comprises the temperature of the steam on the
air thermometer, deduced from the elastic forces inscribed in
column 14. Column 17 gives the temperatures of the steam,
observed directly on the mercurial thermometers, which are
arranged in the boiler. These two kinds of temperature neces-
sarily differ from one another, since the two instruments do not
admit of being compared together ; but on taking into account
the different course of the two instruments, it will be found
that the temperatures observed are generally very sensibly
higher than those deduced from elastic forces. These differences
often exceed those which we have remarked in the experiments
of Table I., and which were made under the pressure of the
atmosphere. This circumstance arises from the fact, that in
experiments at a high pressure, the steam of the boiler must
present an elastic force considerably superior to the pressure of
the air of the receiver, in order to produce a sufficiently
abundant distillation. This difference of pressure is besides too
small in all cases to exercise any detrimental influence on the
correctness of the results.

III. Total heat of Steam under Pressures below that of the

Atmosphere.

The forcing air-pump used in the preceding experiments is,
here replaced by a pneumatic machine, and the air is discharged
into the receiver E F, until the degree of pressure is obtained
which is required for making the determinations. The expe-

AT DIFFERENT PRESSURES. 283

rirnents are conducted in the same manner as when the steam is
engendered under high pressure; but the temperature of the
water proceeding from the condensation of the steam, can no
longer be directly measured, because the water cannot be made
to pass from the calorimeter until air has been introduced into
the apparatus. It is necessary to assume that the temperature
of the condensed water is in a state of equilibrium with that of
the calorimeter, at the moment when the maximum temperature
is observed. The error which may result from this hypothesis
is, however, made apparent, by watching the refrigeration of the
calorimeter during the first ten minutes that succeed the obser-
vation of the maximum, and by comparing this refrigeration
with that which occurs during the following ten minutes. The
difference between the two results observed, gives very nearly
the quantity of heat yielded by the condensed water to the
water of the calorimeter, and it may be assumed that the tempe-
rature of the condensed water is in equilibrium with that of the
surrounding air.

After every couple of experiments, the cock R4 of the air
receiver E F is closed, and the air suffered to re-enter into the
condenser and the boiler; finally, the water condensed in each
of the calorimeters is successively collected in the bulb O, and
its weight determined.

The ebullition of water at low pressures is more irregular
than under high pressures, being attended by starts and inter-
missions, and the thermometers of the boiler indicate notable
variations of temperature according to the greater or less activity
of the distillation. In general, a considerable difference is
manifested in these experiments, between the pressure of the
steam of the boiler and the air of the artificial atmosphere. As
the steam has a very considerable density, it is necessary to
conduct it with much rapidity, in order that the experiment
may not be continued too long. In these more recent experi-
ments, in the place of coke and coal, I have used charcoal, as
being more easy to regulate suitably.

284

LATENT HEAT OF STEAM

•9 1 uam^oo ni ubatS sb ^
'areajg aqi jo a.in}B jaduia j, aqi 2
SuipBjjqng laye '^eajj [i^oj, *"'

rtno^ioooocoxHXOiONHOH^Nifl'^toin

iOiONn«CifliOtCKNNOOO)(NCOHtD«)-*XX(N

moioiOioiOKjininoiOinioiOiOiOioininininiOin

Total

Heat

of the

Steam.

05 g

a 03
-<-> 05
03 *?
05
Pi 05

1-9

05
BQ

-H

«OCO-*rHC<ltX)H(XI(NMCOCO<J5 iCOinNNHcOOOOO
05«50rfcO!0<005'*HO«ON ;N(D0005N^Ol>>C3

1— 1

HfCN^Oxx^OMONNiOaiOinHOOHO^N
i— I 000>NNMOlOO<0«5HONiflCOH^NOfO«0

«*H .

° a

<" Xfl

05 a,

ft" r3

Atmos-
pheres.

i— i

je0!0di0-*N(X)0)i£M-1(OC0i0'*ONt0f0HO0)'*
S -*«a)KK«OINHOOKI>NincOOINWHOOONCOIN

^ooooooooooooooooooooooo

milli-
meters.

Oi-H^cOC<l<NOC\10'^C^COC<ICOt>.*>»iOOJt>.e<JO*>»(— I

• NCOMH^OJ^OiOieOHHWOHOinOHNH'^ffi

a OOMOlNiOr-i^OiPOONOOONNIiXOCOQOHO

oooo^cocoMOosvC^o^in^coOTic^cocoHOiats

Tempe
rature
of this

Water.

:::::::::::::::::::::::

W eight
of the
Water
con-
densed.

.»T5iOOOQOkOeO'*iC*>»iOCOOCOt>.(NOCvaOQDeOCO-*

«iX0H0O05(O(0C0^C0HNHNCi:OO(NO0J<N00C0

±0)pH-*ON^HK«r)*CO'1<KNiOtD«005inC!miNX

h — to

Cor-
rected.

r— 1

HH^NXOOKOtOMOOON^rtNODiOrtaiiflffi

diN^KaeoKNxox^TffOONHTCHXinino

MNOOOOKTj((NWCMHWNK5CONNfOHH(NNH
XlNNCOl^TfOiOWCON-^OlsrtlNTHCC-^^DOOO;^

«HCOINHHH(N(NHHrt050COCOriHNNH(NH

Heat by
Conduc-
tion
during
1'.

OOOOOOOOOOOOOOOOOOOOOOO

cococon«cocoeocort«eo(Nco«co«cocococococo
ooooooooooooooooooooooo
ooooooooooooooooooooooo

ooooooooooooooooooooooo

inter-
vening

before
the fall

of the
Tempe-
rature.

uurmg
which

the
Steam

was
Intro-
duced.

H|N

r-H i— 1 r-H I— 1 rl I— It— li— 1 I— ( I— 1 I— 1 i—H I— li— (r— lr— IrHi-H

Mean
Tempe-
rature
of the
Air t.

f^"

(OTfTfomannotortCincoi- lOo^Oi— <ocoooo<m
fOcoNioinHTfONaiNHN^OHXcjotocoffio

coxoo<oanoeooHn^xcoo(ON(ootoa<ONN

I— 1 I— ( 1— 1 1— 1 1— (

Ob-
served.

0'*(NO)0>XHNX^DN05INtsHNi005'*(N(NCOeO
<MCO«00<MaiiM^i-HOOr-i-OCi^OOO^'<^-^cOC^^

OincoHeocoNxoaHffixoTfiffiin«oNcoT)HM(N

XINHXb.-^OinOJINNMJXONHnNMONX^

NHrtNHHHNNHHHXOPJMiiHNMHNH

Final
Tempe-
rature

MOCOHTttTjicOcOinHTfifHNOXXCiNiNCOMin
OOlrtN^KifHOOflHiiO-ONININtOHHINiON

G>NOffiCON(NHH(NNOiNNOOXX01OiXC!X
i— 1 l-H CN rH i-H i— <<NOJ<MOJ<Nl-Hi-Hr— 1 N M H Hi- Ii-Hi— 1 r— 1 I— I

Initial
Tempe-
rature

NinooHOHininNMinMomacocoooixnoo
x«NTiiffico(00»offloooo>oo)ajXiO'<j<Kin

°o ii><b(0!0OC)00HOx»NWN<o«oii(bi

i-H i-H i-H

of the
Water
of the
Calori-
meter.

NHxinrttoociHijoiOinHMHMrtrtHorfio

xif5NXin>no-*(NTi<inMcoxif3xxiOioxoxio
co^coco^^^ro^coco^^co^ccicO'HH-HHcOTticOTfi
inio»siniOi(3iOi/5iOioi(5ioiniOiniOiOinK5iOini(5ifl

•.iapuiuo[B3

bobbooobooouobuubuobobo

•s^uacuu

)dxg aqj jo jaqum^j

1— 1

HNrti'iO^NXOOHINCOHiOtONXOiOHINtO

AT DIFFERENT PRESSURES. 285

Table III. contains 23 experiments, made under pressures
varying from 0atm* 64, to 0atm* 22.

On comparing together the experiments made under nearly
the same pressure, it will be seen that the total heat yielded by
them, presents somewhat greater differences than those shown
in Tables I. and II., but only one of these differences amounts
to 4*7, that is to say to about yjq of the total quantity measured.

I have not attempted with my large apparatus to make
determinations of the total heat of steam under smaller pressures
than y of an atmosphere, since the ebullition of water under
lower pressures is very irregular^ and it is difficult to direct it
in such a manner, as to obtain results that may be relied upon
with any degree of certainty.

It would, however, be desirable, if the total heat of steam in
a state of saturation, at very low pressures, could be ascertained
with exactness, since this knowledge is of the highest impor-
tance in the study of meteorological phenomena. I have
repeatedly made a large number of observations to determine it,
but many of the methods which I employed require the know-
ledge of certain elements which are still involved in much
uncertainty.

Instead of determining experimentally the quantity of heat
yielded by the condensation of a known weight of steam, at a
determined pressure, to the cold water of a calorimeter, I have
sought to ascertain the quantity of heat, which a known weight
of water, placed in the receiver of a calorimeter, abstracts from
this calorimeter when evaporated at a very low pressure.

I have made use, for this purpose, of a small calorimeter
made of tin-plate, similar in form to that employed by M. Brix,
with the exception that the dimensions of my apparatus are
larger than the former. After measuring the exact weight of the
calorimeter in a scale, it is again weighed after the introduction
of about five grms. of distilled water, into the receiver efg h,
by means of a small pipe 1 1. By this means the weight of the
water to be evaporated is ascertained with precision. The
receiver is then put into communication by means of its tubu-
lure m n, with a flask which serves as an air-receiver, and which
communicates on the one side with a pneumatic machine, and
on the other with a barometrical manometer.

286 LATENT HEAT OF STEAM

The flask serving as an artificial atmosphere is plunged into
a freezing mixture, composed of ice and common salt. When
the apparatus is properly arranged, a determined volume of
water, whose temperature is known, is poured into the calori-
meter, which is again covered, whilst a very sensible small
thermometer is so arranged, that the bulb is placed in the axis
of the calorimeter.

The observation is begun by noting the advance of the
refrigeration, or heating of the water of the calorimeter for five
minutes, whilst the water is being continually agitated. In
order that the refrigeration may proceed with more regularity, the
calorimeter is arranged in the centre of a somewhat larger iron
plate vessel, to preserve it from the action of accidental currents
of air. It is then rapidly emptied by means of the pneumatic
machine, and the elastic force of the internal air is then brought
to a determined value, observed on the barometrical manometer,
which must necessarily be inferior to the elastic force of the
steam, at the temperature of the calorimeter.

The distillation of the water in the receiver efg h of the
calorimeter, begins as soon as the steam condenses in the cooled
flask. The temperature of the calorimeter, the water of which
is continually agitated, is then noted from minute to minute,
and the elastic force indicated by the barometrical manometer
observed at the same time.

The moment, at which the water of the receiver efg h is
completely distilled, is easily ascertained by the sudden stopping
of the thermometer in its fall ; the temperature continues sta-
tionary for a very short time, and then rises slowly, provided
the temperature of the water of the calorimeter is below that of
the surrounding air. The termination of the distillation is
likewise ascertained by the barometrical manometer, owing to
the sudden rise of a small quantity of mercury in the column
communicating with the apparatus.

The ascent or descent of the thermometer is again observed
for 5 minutes. This last observation, combined with that of
the refrigeration for the 5 minutes preceding the distillation,
gives the elements, by which an exact calculation of the cor-
rections for the final temperature may be made, in order to take
into account the loss, or gain of heat experienced by the calori-

AT DIFFERENT PRESSURES. 287

meter from the surrounding air, during the whole course of the
experiment.

Let M be the weight of the water introduced into the calori-
meter, increased by the value in water of the calorimeter itself
and of its appendices.

m the weight of the water to be evaporated;

t0 the initial temperature of the calorimeter ;

fj the final temperature;
then t0 — tx is the decrease in the temperature of the water of
the calorimeter owing to the evaporation of the water in the
receiver.

The quantity (t0 — tj) must be corrected owing to the action
of external causes on the calorimeter. The correction was
effected in the following manner :

A series of preliminary observations was made on the rapi-
dity of the cooling, or the heating presented by the calorimeter,
when filled with water at different temperatures; the external
air preserving the same temperature. A graphic curve has
been constructed from the data yielded by these experiments,
the temperatures of the calorimeter being taken as the abscissa?,
and the velocities of the cooling, that is to say, the decrease of
temperature experienced in a minute, serving as the ordinates.
It has been found that this curve differs but little from a
straight line, and it may even be assumed, without any sensible
error in the valuation of the correction, that it is actually a
straight line.

The direct observation of the cooling of the calorimeter
during the 5 minutes preceding the experiment, and the obser-
vation of the heating during the 5 minutes succeeding it, give
the extreme points of the straight line representing the cooling
in each experiment. It will, therefore, be sufficient to take on
this line the coolings and heatings experienced by the water
every moment of the experiment, indicated by the temperature,
which must be noted from minute to minute. The total cor-
rection to be applied to t0 — tx will be equal to the algebraic
sum e of all these partial refrigerations.

We shall have, on designating by X the quantity of heat
absorbed by 1 gramme of water, when evaporated under the con-
ditions of the experiment : m X = M (t0 — tl + e).

288 LATENT HEAT OF STEAM

In order that X may represent what we have termed the
total heat of steam, it is requisite that the evaporated water m
should be introduced into the calorimeter at the temperature of
0°. It has, however, in reality the initial temperature t0, and
consequently the total heat will be X + t0.

The quantity X + t0 presents perfectly well the quantity
of heat absorbed by the weight m of water at 0°, in order to be
reduced to steam under the conditions in which the evaporation
is effected; but the question here arises: is the steam actually
developed in a state of saturation, and if so, what is the elastic
force or the temperature to which this state of saturation cor-
responds? These are the true difficulties of the subject, and
they do not appear to me to admit of being obviated with any
degree of certainty.

We know the elastic force f of the air of our artificial
atmosphere, this being yielded by the barometrical manometer;
but it is evident, that the steam must have a more considerable
tension in the receiver efg h, since distillation could not other-
wise be effected.

The difference between these two tensions must even be
very great, for in order that the experiment may be conducted
under favourable conditions of exactness, the distillation must
be very rapid and the correction e be limited to a very small
fraction of t0 — tv

A cause of uncertainty of the same nature exists in our
experiments at high pressures, the steam necessarily possessing
in the boiler an excess of pressure above the air of the artificial
atmosphere; but no sensible error can arise from this circum-
stance, since the excess of pressure is always an extremely small
fraction of the total pressure. The case is different with our
actual experiments, in which the difference of pressure may be
a considerable fraction of the total pressure, since the latter is
very small. It is therefore probable that the elastic force f,
observed on the barometrical manometer, is appreciably smaller
than the mean pressure F at which the steam has distilled.

Moreover, the steam that issues from the calorimeter, is not
in a state of saturation. This steam is in a state of saturation
at the moment in which it is developed, and its tension cor-
responds with the temperature t, presented at this moment by

AT DIFFERENT PRESSURES. 289

the water to be evaporated; but this temperature t must neces-
sarily be below that of the surrounding air of the calorimeter.
The steam, before its escape, becomes again heated against the
walls of the receiver of the calorimeter, abstracting from them a
small quantity of heat, by which the result is rendered more
complicated.

Thus (1) we do not know with certainty the temperature 6
to which the saturation of the steam corresponds; this tempe-
rature must be somewhat above that corresponding to an elastic
force of aqueous vapour, equal to the pressure f of the mano-
meter: and (2) the value of X + t0 must be found somewhat too
high, since the steam issues from the calorimeter at a tempe-
rature above 6.

It is evident, that these causes of uncertainty would be con-
siderably diminished by bringing the pressure f only a little
below the elastic force of the aqueous vapour, corresponding at
every moment to the temperature of the water of the calori-
meter ; but the distillation is in that case very slow, the experi-
ment lasts a long time, and from that circumstance alone is no
longer susceptible of exactness.

I do not, however, think that the different circumstances
which I have enumerated can sensibly change the results.
With a view of ascertaining this, I conducted my experiments
under the most varied circumstances; sometimes, for instance,
the distillation of 5 grammes of water was most rapidly effected,
being completed in 3 or 4 minutes, whilst the apparatus was
only subjected to a very low pressure; at another time, I allowed
a higher pressure, and in that case, the distillation of 5 grammes
of water required 12 minutes. The causes of perturbation were
necessarily much more energetic in the former than in the latter
instance; the differences were, however, always very small. I
also varied the initial temperature of the water of the calori-
meter from 9° to 28°.

I have tried to ascertain whether exact results could be
obtained for a very slow distillation, on comparing the rapidity
of the refrigeration of the calorimeter whilst the water was being
evaporated, with that presented by this calorimeter under the
same circumstances when the water was not being distilled;
but it appeared to me, that it was impossible to deduce any

290 LATENT HEAT OF STEAM

certain data from this mode of operation, owing to the great
irregularity of the distillation.

It has already been stated that the same volume of water
was introduced at each experiment into the calorimeter. The
weight of this water varied from 500gr. to 502 gr., between the
limits of the initial temperatures t0 observed during my expe-
riments. To this weight must be added the value in water of
the calorimeter and its appendices. This value, determined
from the weight of the apparatus and the specific heat of the
metal constituting it, was found to be 37gr. 5. I determined
this value by direct experiments, and by observations on the
changes of temperature experienced by the water of the calo-
rimeter on the addition of a determinate quantity of hot water,
whose temperature was known with certainty, or by the fusion
of a known weight of ice.

I also found as the mean of a great number of experiments,
that the value in water of the calorimeter was 42gr. 3. I have
assumed for this value the number 40 '0, which is evidently the
mean between the value calculated, and that which had been
found by direct observation.

I have collected in Table IV. all the experiments made
according to this process. No special explanation is requisite
for its comprehension, as the elements of which it is composed
have been sufficiently explained in the preceding pages.

It is necessary to observe that the experiments included in
the following table, were purposely made under the most varied
circumstances, which were in many cases unfavourable to the
exactness of the results. Smaller variations would certainly
have been found for the total heat, if the experiments had
always been conducted under similar circumstances.

I do not think that we should deviate very considerably
from the truth, by assuming that the total heat of the steam
developed in a state of saturation at a pressure of 9mm*16, which
corresponds to the temperature of 10°, is equal to 610 units.
The total heat for water at 0°, would be only some units below
this.

AT DIFFERENT PRESSURES.

291

«3 -i

08
H

1— 1

K5C)ML'5iSii:OH0005NCOClO^«NflNHHN

HQOiflfrtHrtLinon'tH^inociHHM^Oi

OOOi-hi-Hi— ioOp-^OOi-hi-hi— ip— ii— (Oi-Hf— ii— ii-Hi— 1

Heat

absorbed by
the evapo-
rated water,

i—t

rtN^OOMrtOONifXN^^HOrJUONLlO

Duration

of the

Distillation.

„■— 1 NCiOOOlHONHXNOi'X^'^Tti^o^cO

l-H 1— 1 1— 1 l-H i-H f— 1 r-i

Temperature
at which
Steam pre-
sents the

Elastic Force,

0O NONQONOOHN^HMinJl^OoilOMiMO
i—l 1— 1 rH i-H 1— ( i-H

1 1 +++++++++++++++++++

Mean pres-
sure at which
distillation
is ellected.

g fW^NMNOiONHOHOiJsNXBXOOWOON
l-H H I-H i-H i-H i-H

"^-3

1 <D

e >-i

XPJ^lfllflNNOXL'JNtHNTllNOL'JOHrtL'J

'^i-H(M<Mi-H-H<C5t>.eOlMin-^t,Oi-HO^OC^Ot>.-^<(NO

observed.

O^HNNiflrtt^rtONCOONCOXXNClHO

»«X»HiflX^X001«HOL'5C5M5iNNNH

0-o xMrtrtrt^flooTiimNoooooooiflts
0 0 0 us 0 0 0 0 0 0 0 0 in 0 in li in 0 0 if) in 10

Final
Temperature

of the
Calorimeter

iftT(i«001if)NOXf)ONOHT)iOinNOtONOin
0»-0 rtHOClXNOOOXOOO^MHTl<^NOf

MONOOOcOOeO^^ifXO^ONNXXXHNN

Initial
Temperature

of the
Calorimeter,

to-

i-Ht>.--ft>.l>.l>.'!t*(Ni-HOp-H00C^p-Hi0Tj<f-H0SOOt>»O
NHfi'ilHNNNCgcOliNNOinXOHTtufjH

0)(NNinOt3XCiOOOHHHene0WT(<i<<0NX

Weight of
the Water

of the

Calorimeter,

MOxx^tntnooooooixx^Tj'cflcnwHoc)

C<>Ji— '•— '1— 11-H1— ii— *i— ip— ir- IphOOOOOOOOOOCS

Uooinoooioooirjinoi.'SirtinoifSinoooin

Weight of

the Water

evaporated,

o;oot^.oc^i>«oooocjc<ic^i— ii-HOOioe4?cc<ioi>.

-.nXNNONN^NL'ji.'STCNNOOfflCHOXO
CcOi-hi— (i-h(Mi-hi— (C^C^(Mi-HtNC^C^C^C^i-HF-i01i-HO(M

jjooininininininoinininininininoininininin

Number
of the
Experi-
ments.

HMrt^inoNXOiOHNrti'inoNxaoHrM

U 2

292 LATENT HEAT OF STEAM

I have endeavoured to obtain the latent heat of steam satu-
rated at low pressures, by another method, which will, I hope,
enable me to arrive at this result with much certainty, without
its being open to the same objections advanced against the first
process. But this method, which I have described at the end
of my memoir on hygrometry, (Annates de Chimie et de Physique,
3me serie, tome XV, p. 227,) requires a knowledge of many
data that are still involved in considerable uncertainty. It is
necessary, for instance, to know the calorific edacity of the air,
and the quantity of heat absorbed by the air during its dilata-
tion. I have, therefore, considered it necessary to determine
these two elements by means of new experiments, and it is only
when these are concluded that I shall be able to calculate the
determinations of the latent heat of steam.

On glancing over the Tables I., II., III., and IV., it will be
easy to ascertain whether the experiment verifies either of the
two laws, which so equally share the authority of physicists.

According to the law of Watt, the quantity of heat neces-
sary to change a kilogramme of liquid water at 0° into steam in
a state of saturation, is the same under all pressures. These
quantities of heat are set down in column 18 of the tables, and
the numbers representing them must therefore remain constant,
or, at any rate, only manifest accidental variations arising from
errors in the experiments. These numbers increase with the
pressure in a perfectly regular manner from 610, which is the
value of the total heat of steam at a pressure of y^-g of an
atmosphere, to 666, which expresses the same quantity under
the pressure of 13*6 atmospheres.

The law of Southern maintains, on the contrary, that the
heat obtained by subtracting from the total heat the sensible
heat indicated by the thermometer, or that which is ordinarily
termed, the latent heat of evaporation, remains constant at all
pressures. Column 20 of Tables II. and III., gives the num-
bers obtained by subtracting from the total heat set down in
column 18, the temperature (column 19) which would be
indicated by an air thermometer plunged into the steam, this
temperature being calculated from the elastic forces of steam, as
set down in column 14.

If the law of Southern be correct, the numbers of column

AT DIFFERENT PRESSURES. 293

20, in Tables II. and III, must remain constant; but these
numbers diminish gradually from 560, which corresponds to the
pressure of ^ of an atmosphere, to 471, corresponding to the
pressure of 13*6 atmospheres, or even from 600 to 471, if we
admit the value 610 — 10, which corresponds to the temperature
of 10° deduced from Table IV.

The law of Southern deviates therefore still more than the
law of Watt, from the numerical results yielded by direct
experiments.

It would now be better to seek the true law, connecting
together the total quantities of heat contained in steam in a
state of saturation, but I do not believe, that this object can at
present be effected with any degree of success from the want of
several elements, the knowledge of which, appears to me, to be
absolutely necessary to the solution of the problem. It seems
essential that we should know the law regulating the densities
of aqueous vapour in a state of saturation or non-saturation, at
different pressures and different temperatures. These densities
are generally calculated by supposing, that for a constant tem-
perature the densities of steam in a state of saturation, or non-
saturation, may be calculated according to Mariotte's law, and
that at the same pressure, but different temperatures, the
volumes of non-saturated steam may be determined by admit-
ting that steam dilates for every degree of temperature, be its
density what it may, by the same fraction of its volume from
zero, as the atmospheric air, during the same interval of tempe-
rature, when the density of the latter is equal to the pressure of
one single atmosphere.

The different researches that I have published, render it
very probable that these hypotheses deviate considerably from
the truth; it is, therefore, greatly to be desired that these rela-
tions may be established with certainty by means of direct
experiments.

The observations included in this memoir give the quanti-
ties of heat yielded by a kilogramme of steam in a state of satu-
ration at different pressures, and therefore at different tempera-
tures, when reduced to the condition of water at zero. But it
also appears essential that the quantity of heat should be known,
which is yielded under similar circumstances by a kilogramme

294 LATENT HEAT OF STEAM

of steam at different temperatures and different pressures, when in
a state of non-saturation. This will be the object of a future
memoir, to which I purpose subjoining new observations on the
specific heat of permanent gases, and on the heat disengaged by
compression.

These remarks appear to be absolutely necessary in order to
explain, in a rational manner, the physical laws of elastic forces
of aqueous vapour at different pressures, and those laws, which
connect together the total quantities of heat possessed by
the steam under determined circumstances. I purpose revert-
ing to this subject, whenever I may succeed in solving experi-
mentally the problems I have just enounced.

But, whatever the law may be, which connects the total
quantities of heat with the temperatures, the extent of my expe-
riments lead me to conclude that it may always be numerically
developed, according to the ascending powers of t, and to sup-
pose that it gives A - A + B T + C T2 + D T3 + . . . A, B,
C, D . . . being constant co-efficients.

If we assume A = A, we have the law of Watt, the incor-
rectness of which we have seen demonstrated in the preceding
experiments. I have tried whether my experiments might be
represented with sufficient precision by means of the formula
with two terms : X = A + B T.

We have admitted, according to the experiments of Table IV.,
that the total heat, contained in steam in a state of saturation,
for the temperature + 10°, was . . . 610 units.

Table III. gives as the total quantity of heat,
contained in steam in a state of saturation, at 63° 625 „

According to Table I., the total heat of steam
in a state of saturation, at the temperature 100°,
is 637 „

Finally we see at the end of Table II., that
the total heat of steam at 195° of an air-thermo-
meter, is 666 „

I have calculated the two constants A and B of the formula
by means of the two last observations, that is to say, by means
of the values of the total heat at the temperatures of 100° and
195°, and from these data, in which I place perfect confidence, I
have found :

AT DIFFERENT PRESSURES. 295

A s= 606-5

B c- 0*305 ; so that the numerical
formula is X — 606-5 + 0-305, T.

This formula gives very exactly the total heat for the tem-
peratures of + 10° and + 63°, thus for T =- + 10°
we have ....... X == 609*6

We have assumed from our experiments . X = 610*0

For T = + 63° the formula gives . . . X = 625*2
Our experiments have yielded . . . X == 625*0
We shall besides see that the numbers, yielded by the for-
mula for the different temperatures at which the experiments of
Tables II. and III. were made, never present greater differences
from the number found experimentally, than may be ascribed to
the probable errors of the observations. We may, therefore,
admit that our numerical formula represents all our experimen-
tal determinations in a satisfactory manner, and that the expe-
riments are not susceptible of sufficient precision to render it
expedient to have recourse to a formula with three terms.

We will therefore provisionally adopt the formula X = 606*5
+ 0*305 T, as expressing the numerical law of the phenomenon,
until a more thorough knowledge of the properties of steam
shall enable us to establish its true physical law.

According to this formula, the total heat, incorporated in a
kilog. of saturated steam at the temperature T, is equal to the
quantity of heat yielded by a kilogramme of saturated steam at
0°, in passing into the condition of liquid water at 0°, increased
by the product 0*305 T.

The fraction 0*305 is therefore the special calorific capacity
of steam, different from the calorific capacities of gases at a con-
stant volume, or at a constant pressure, although intimately
related to them. It is the quantity of heat required to raise
the temperature of a kilogramme of steam 1° in temperature,
whilst at the same time, this steam is compressed so as to keep
it in a state of saturation.

I have calculated, by means of this formula, the following
Table, in which are comprised the total heats incorporated in
saturated steam for every 10° of increasing temperature.

296 LATENT HEAT OF STEAM AT DIFFERENT PRESSURES.

Temperature
of the

Corresponding '.

Elastic Force.

Saturated

Total Heat.

Steam.

In Millimeters.

In Atmospheres.

mm*

Atm.

0-006

606

0-012

609

0-023

612

0-042

615

0-072

618

0121

621

148

0-196

624

233

0-306

627

354

0-466

630

525

0-691

633

100

760

1-000

637

110

1075

1-415

640

120

1491

1-962

643

130

2030

•28

2-671

646

140

2717

3-576

649

150

3581

4-712

652

160

4651

6-120

655

170

5961

7-844

658

180

7546

9-929

661

190

9442

12-425

664

200

11688

15-380

667

210

14324

18-848

670

220

17390

22-882

673

230

20926-40

27*535

676-6

Same Table in Fahrenheit Degrees of

Heat and

English Inches.

Temperature of the
Saturated Steam.

Corresponding Elastic Force.

Total Heat.
(Latent heat + sensible

(Vapour at the point

of condensation.)

In Inches.

In Atmospheres.

heat above 32°.)

32° F.

0-1811

0-006

1123-70° F.

3606

012

1129

6846

023

1134

2421

042

1140

104

1618

072

1145

122

6212

121

1151

140

8578

196

1156

■64

158

1767

306

1162

176

9621

466

1167

194

6869

691

1173

212

9212

000

1178

230

3374

415

1184

248

7116

962

1189

266

9321

671

1194

284

106

9930

576

1200

302

140

9930

712

1205

320

183

1342

120

1211

338

234

7105

844

1216

356

297

1013

929

1222

374

371

7590

425

1227

392

460

1943

380

1233

410

560

9673

848

1238

428

684

6584

882

1244

446

823-8723

27-535

1249-88

297

VII.

REPORT.

ON THE ARTIFICIAL FORMATION OF ALKALOIDS.
By M. E. Kopp.

There are few questions more deserving of the attention of
chemists and pharmaceutists, and of being made the object of
their researches, than the formation of alkaloids.

In a chemical point of view, the alkaloids are remarkable
for their composition, for their special properties, both physical
and chemical, and for the interesting reactions to which many
of them give rise, when exposed to the influence of different
agents. Considered medically, the organic bases are distin-
guished by their energetic properties. They constitute at the
same time the most violent and sudden poisons, and the most
valuable and heroic remedies. Hence their use in medicine is
very extensive, while their price is often exorbitant.

It is more than probable that by studying the mode of their
formation, we may in course of time be led to prepare artifi-
cially, and at little expence, morphine, quinine, strychnine, &c.,
so as to render these medicines more available to the poor, and
to free ourselves from the necessity of heavy payments to
foreign countries.

Moreover, there is no impossibility in the artificial formation
of the alkaloids.

During the last few years the chemist has formed a tolerably
large number of artificial organic bases, similar in all respects
to the natural alkaloids; and without doubt the methods
adopted in our laboratories^ must be analogous to those em-
ployed by nature herself.

298 ON THE ARTIFICIAL FORMATION

The question regarding the mode of formation of the alka-
loids is naturally divisible into two heads; the former relating
to the formation of the natural organic bases, the latter to the
modes of preparing the artificial alkaloids.

The former point is beyond all doubt the most important,
but unfortunately our ideas on this point are extremely obscure.

Our knowledge respecting the formation of the natural
alkaloids is so imperfect, that it may be plainly asserted that
nothing has yet been effected, and that we are ignorant of even
the first elements of this investigation.

Can we, in point of fact, tell the period at which the alka-
loids first present themselves in the juice of plants? Do we
know the substances from which they originate ? or the circum-
stances favouring their development? or the influence of oxy-
gen, water, &c. on their formation?

If then our knowledge respecting this department of science
be so imperfect, how ought we to proceed, in order to acquire
any information regarding the mode in which the natural
organic bases are produced?

To attain this object, three different courses are open to us.

I. The Study of Chemical Botany.

This would certainly be of immense service in the solution of
the question. It would lead us to examine the composition of
plants, especially of those yielding these remarkable principles,
in the different phases of their growth and decay.

We should especially select the periods of germination, of
the development of the leaves, of that of the flower, and of the
maturity of the fruit. We should thus arrive at a knowledge of the
principles which are contained in plants in the different phases
of their vegetative life; Ave should know what principles are first
formed, which afterwards disappear, and what new ones result
from them ; and we should be thus led to form conjectures on,
if not actually to verify, the nature of the relations existing
between the first and the last. An investigation of this nature,
extending to all the most important plants, would undoubtedly
be rich in new and interesting results.

Such a work would fall within the domain of the pharma-
ceutist, who, being perfectly familiar with the ordinary modes of

OF ALKALOIDS.

299

their extraction, could, without any great difficulty, isolate pure
and frequently crystallizable principles; if their elementary
analysis, and their exact description were beyond the ordinary
means at his disposal, any chemist would willingly carry on the
examination of the substances thus isolated, and in this manner
contribute his share to the advancement of his own science.

The following Table contains all the natural alkaloids yet
discovered.

They may be divided into three classes:

lo Those which have been submitted to ultimate analysis.

2. Those which have not yet been analysed, but whose
existence seems well established.

3. Those whose existence is problematical.

Cinchonine ....

Quinine

Aricine

Brncine

Strychnine ....

Morphine ....

Narcotine ....

Codeine

Pseudo-morphine

Atropine

Berbeerine ....

Bebeerine ....

Cafeinel

Theine )

Chelidonine

Cinchovatirie

Conine

Corydaline ....

Harmaline ....

Jervine

Menispermine

Nicotine

Piperine

Sabadilline ....

Staphysai'ne....

Theobromine

C20H12NO
C19Hn NO (Laurent)
C20H12NOo
[C19 Hu N 02 (Laurent)
C20H12NO3
fC2,H13N04
[C23 H13 N 04 (Laurent)
C^ H12 N 04

[CMHttNO,

[C34 H19 N06 (Laurent)

C46 H27 N 014

C36H21N06

C24H13N012

C34H23N06

C4<> H26 N 016

C34H20NO6

C16 H10 N4 03 (Payen)

C8 H5 N2 02

C40 H20 N3 06

C23H14N04
jC16H17N
IC16 H15 N (Laurent)

C34 H22 N °10

C,tH13NO

^30 H23 N 02

C18H12N02

C10 H7 N
C34II19NOG
C20H14NO6
C3o IL3 N O.,

c;h;n3o2"

300

ON THE ARTIFICIAL FORMATION

II.

Delphine

Sanguinarine

Pereirine

Emetine

Porphyroxine

Jama'icine

Solanine

Sepeerine

Surinamine

Glaucine

Cissampeline

Cicutine

Aconitine

Thebaine

Macrophylline

Xanthopicrine

Veratrine

Saracine

Chelerythryne

Digitaline (?)

III.

Hederine

Apirine

Cynapine

Euphorbine

Buxine

Esenbeckine

Fumarine

Carapine

Eupatorine

Guaranine

Crotonine

Bases derived from Escholzia Californica.

1. A base yielding red salts.

2. A colourless acrid base, soluble in ether.

3. A colourless acrid base, soluble in water, and assuming a
violet tint when treated with sulphuric acid.

II. The Study of the Modes of Preparation of the
Artificial Alkaloids.

As we have already observed, it is probable that some of
the modes of procedure employed by the chemist in the labora-
tory are analogous to those actually occurring in nature.

Attention should be especially paid to those methods which
most nearly approximate to those employed by nature, and do
not require either a very high temperature, or the influence of
very energetic reagents; such, for instance, as give origin to
thiosinamine, furfurine, &c.

The comparative and rational examination of these processes
appears likely to indicate the course we must follow, in order
to prepare artificially the natural alkaloids.

The following are the artificial bases at present known:

Melamine
Ammeline

Ammelide

Aniline
Leukol i
Quinoleine

C3H3N3

C6 H5 N5 02

/C^Hg^OJLiebig)
1C3 H2 N2 02 (Gerhardt)

C12 H7 N

C6 H7 N (Gerhardt)

OF ALKALOIDS. 301

Napbtalidine C20 H9 N

Seminaphtalidine) n tt at

Naphtidine J ^10^5^

Theosinnamine .... .... .... .... C4 H4 N S

Sinanine .... .... .... C4 H3 N

Sinepoline C7 H6 N O

Urea CH2NO

Lophine C23 H8 N

Amarine C21 H9 N

Furfurine CI5 H6 N 03

Toluidine C14 H9 N

Narcogenine C36H19NO10

Cotarnine C26 H13 N 06

Guanine C10 H5 N5 02

Nicotine C10 Hr N

Sugar of gelatine C4H5N04

The bases derived from harmaline, namely:

1. Porphyrharmine, resulting from the action of alcohol on
its granules.

2. Leucoharmine, obtained from the chromate of harmaline.

3. Chrysoharmine, yielded by the action of nitric acid on
sulphate of harmaline ; and finally animine, odorine, and pyrrhol.

To these we may add,

C6 H9 P. The phosphorized alkaloid of M. Paul Thenard.

C2 H3 As. Cakodyl and its oxide.

Cakoplatyl, and the platinised bases of M. Reiset.
But their composition is very different from that of the ordi-
nary alkaloids.

III. The Study of the Products of Decomposition of the
Natural Alkaloids under the influence of different Re-
agents, and under different circumstances.

It is a recognized fact in chemistry, that the manner in
which a substance undergoes decomposition, and the products it
yields, frequently enable us to draw conclusions regarding the
mode of its formation, and the course to be pursued in forming it.

But the decomposition must not be accompanied with much
violence, if we wish to establish the relations existing between
the primary body and the derived products, and the equation
illustrating the change that occurs. It unfortunately happens
that most of the investigations that have been undertaken, in

302 ON THE ARTIFICIAL FORMATION

relation to the changes which the alkaloids undergo, have in no
way come up to these conditions, and consequently possess only
a moderate degree of interest.

Such, for instance, are those of Brandes, Leber, Jonas, and
Andre, on the action of chlorine on quinine and cinchonine, by
which there are produced a series of imperfectly characterised
bodies to which the names thalleiochine, russiochine, and mela-
nochine have been applied ; those of M. Rousseau on the trans-
formation of strychnine into strychnic acid by chlorate of potash
and sulphuric acid; and the researches of M. Marchand on cin-
chonetine, quinetine, morphetine, &c. — red bodies arising from
the action of peroxide of lead acidulated with sulphuric acid, on
the sulphates of cinchonine, quinine, morphine, &c.

In the same way the reactions which give origin to quino-
leine, are too energetic to afford any information on the
natural composition of the body, which furnishes this artificial
base.

The same is possibly the case with regard to the action of
nitric acid on brucine. As, however, the reaction takes place
in the cold, the disengagement of nitrous ether might perhaps
indicate that in brucine there exists a grouping of the elements
analogous to that occurring in the composition of alcohol, and
hence we might draw the conclusion that it may perhaps be
advantageous to employ alcohol in the artificial formation of
brucine.

The researches of MM. Wohler and Blyth on narcotine, are
more of the nature we have indicated, as permitting us to see
into the intimate constitution of a body.

But the (still unpublished) memoir of MM. Redtenbacher,
Rochleder, and Wertheim, promises more important results on
the nature and composition of the alkaloids, than any of the
memoirs to which I have previously alluded.

In fact, these able chemists have announced the following
highly important facts.

Piperine should be considered as a neutral combination of a
nitrogenous acid with aniline, formed of 2 eq. of the acid and
1 eq. of aniline.

They have even succeeded in reproducing piperine synthe-
tically by directly combining the acid with the base ; narcotine

OF ALKALOIDS. 303

has been separated into a non-nitrogenous acid combined with a
peculiar base.

Narcogenine is a combination of this same base with half an
equivalent of the acid.

These combinations of the alkaloids are constituted with
perfect regularity, containing 1 eq. of water inseparable from
the ammonia and the organic alkalies.

It must be apparent that most of the alkaloids, having very
high equivalents, thus admit of separation into more simple
elements, which, in all probability, will be common to many of
them.

The confirmation of this fact would be an important step in
the true progress towards our knowledge of the artificial prepara-
tion of the alkaloids in general.

Formation of Artificial Alkaloids.

With the view of methodically arranging the modes of pre-
paring the artificial alkaloids, we have united them into groups
which embrace in every instance those bases whose formation
presents any points of analogy.

But as the same alkaloid may belong equally to several
groups, and as, further, it may sometimes be produced under
special circumstances, on which it would be difficult to gene-
ralize, each base has been examined specially, and the different
reactions, under which it may be produced, are pointed out.

* FIRST GROUP.
Transformation of the Ammoniacal Compounds.

1. Thiosinamine, a combination of ammonia with the essence

of black mustard.

2. Urea, a metamorphosis of cyanate of ammonia.

3. Furf urine, a metamorphosis under the influence of potash

on furfurolamide.

4. Amarine, a metamorphosis of hydrobenzamide under similar

circumstances.

5. Melamine, Ammeline, and Ammelide, products of the dry

distillation of sulphocyanide of ammonium.

6. Lophine, derived from hydrobenzamide.

304 ON THE ARTIFICIAL FORMATION

7. The alkali produced by the action of ammonia on cacothe-

line.

8. Aniline, resulting from the action of heat on phenate of am-

monia.

1. Thiosinamine, C4 H4 N S.

On mixing essence of mustard or horse-radish with an excess
of a concentrated aqueous solution of ammonia in a closed vessel,
the essence is gradually transformed into a beautiful crystalline
mass which is thiosinamine:

C8H5NS + NH3 = 2(C4H4NS)

Ess. of mustard. Ammonia. Thiosinamine.

This alkaloid presents the most simple formation of any.

2. Urea, CH2NO.

An aqueous solution of a combination of hydrated cyanic
acid with ammonia, becomes transformed on the least elevation
of temperature, or by concentration even at an ordinary tempe-
rature, into urea :

C2NH02 + NH3 = 2(CH2NO)

Hydrated cyanic acid. Urea.

But urea is further the result of the alteration that may be
effected (by oxidation) in uric acid, and in different substances
derived from it.

Thus, for instance, urea may be obtained :

a. By oxidizing uric acid with peroxide of lead (Pb 02),

when allantoine and oxalic acid are also simultaneously
produced :

2 at. uric acid .... C10 H4 N4 06 \ ( C4 H3 N2 03 .... allantoin.

2 „ peroxide of lead .... Pb2 04 > = < C2 H4 N2 02 .... 2 at. urea.
3„ water H3 03 J [ C4 Pb2 08 .... oxalate of lead.

b. By oxidizing uric acid with nitric acid, when at the same

we obtain alloxantin :

2 at. uric acid ..

.. C10H4N4O6)

r C2 H4 N2 02

.... 2 at. urea.

5 „ water

.. ... H505 \ =

1 „ oxygen

I C8 H5 N2 O10

.... alloxantin.

OF ALKALOIDS. 305

c. By oxidizing alloxan with peroxide of lead, when carbonic

and oxalic acid are at the same time produced :

C4 H2 N 05 + Pb 02 = C 02 + C2 04 Pb + C H2 N O

Alloxan. Urea.

d. By boiling alloxan with hydrated baryta, when at the same

time mesoxalic acid is formed :

C4 H2 N 05 + Ba O, HO = C H2 N O + C3 06 H Ba2
Alloxan. Urea.

C3 06 H Ba2 = (C3 04, Ba2 O + H O)

Hydrated mesoxalate of baryta.

e. By treating allantoin with peroxide of lead :
C4H3N203 + Pb02 + HO = 2(C H2 N O) + C204Pb

Allantoin. Urea. Oxalate of lead.

f. On decomposing a solution of oxaluric acid by ebullition :
C3 H2 N 08 + H O = C H2 N O + C2 H 04

Oxaluric acid. Urea. Oxalic acid.

Although urea does not exhibit an alkaline reaction, it
deserves to be classified amongst the best characterized of the
most distinctive alkaloids.

It not only combines with acids, frequently forming crystal-
lisable salts, but also with true salts, such as nitrate of silver,
chloride of mercury, and chloride of sodium (according to the
researches of MM. Werther and Piria); and further it is capable
of forming double salts.

In a large number of cases urea forms compounds analogous
to those to which ammonia gives rise.

3. Furfur ine, C15 H6 N 03.

Furfurine is undoubtedly one of the most remarkable of the
artificial bases.

In its composition and properties it approximates the most
to the well characterised natural alkaloids; the reaction, on
which its origin depends, is not only one of high interest, but it
is susceptible of generalization, and we do not fear to affirm
that some day or other we shall be indebted to it for the dis-

306 ON THE ARTIFICIAL FORMATION

covery of several new alkaloids. In point of fact, there exists a
very close analogy, in respect to the mode of formation, between
furf urine, amarine, and melamine.

All these bases result from a modification impressed on com-
pounds derived from ammonia (amides, imides) by a dilute
boiling solution of caustic potash, The elementary composition
of the original matter has not undergone any alteration, but its
chemical character has been changed ; previously it was neutral,
now it has become strongly alkaline.

In the preparation of formic acid from starch, peroxide of
manganese, and sulphuric acid, there is formed an oleaginous
matter, which was especially noticed by Mr. Stenhouse, and has
received the name of furfurol from Mr. Fownes.

Its composition is represented by the formula C15 H6 06,
or more simply by the formula C5 H2 02

This oil, when brought in contact with ammonia, becomes
transformed into a solid, slightly crystalline substance, insoluble
in water :

CI5H606 + NH3 = C15H6N03 + H3 03

Furfurol. Ammonia. Fulfurolamide. Water.

This new compound possesses all the characters of an acid;
by the action of acids on it, an ammoniacal salt is formed, and
furfurol is reproduced.

But when boiled with weak caustic potash, furfurolamide is
converted into an energetic alkaloid, furf urine, having the same
composition and the same equivalent number. The only change
that has taken place is in the arrangement of the molecules :

C15H6N03 = C15H6N03

Furfurolamide. Furfurine.

4. Amarine, C21 H9 N.

This alkaloid is formed in exactly the same manner as fur-
furine. When 2 eq. of ammonia react on 3 eq. of oxide of
benzene (hydride of benzoile) water and hydrobenzamide are
formed :

3(C,H30) + NH3 = C21H9N + H3 03

Hydride of benzoile. Ammonia. Hydrobenzamide. Water.

OF ALKALOIDS. 307

Hydrobenzamide is decomposed by acids into ammonia and
hydride of benzoile, but when treated with a weak boiling solu-
tion of potash, it becomes converted into amarine, which forms
well marked salts with acids, and has the same composition and
equivalent number as hydrobenzamide.

5. Melamine, Ammeline, and Ammelide.

The consideration of the formation, nature, and composition
of these three alkaloids presents several points of extreme deli-
cacy, not only on account of the inherent difficulties of the sub-
ject and the varying statements of different chemists regarding
it, but further because it has been made the object of a very
warm controversy between MM. Liebig, Gerhardt, and
Vcelkel.

Without entering at length into these discussions, let us suc-
cessively examine the different views regarding the formation of
the bases in question, and endeavour to apply to their elucida-
tion the new ideas suggested to us by the discovery of furfurine
and amarine.

By the dry distillation of sulphocyanide of ammonium, there
are given off sulphide of carbon, sulphuretted hydrogen, and
ammonia (and consequently the combinations of these bodies),
while there remains a grayish residue to which M. Liebig has
given the name of melam :

2 (C2 N S2, H4 N) = C3 H3 N3 + C S2 + 2 H S + N H3

Sulphocyanide of ammonium. Melam.

M. Liebig has assigned to melam the formula C12 H9 Nn.

The melam thus obtained cannot be a pure substance.

According to M. Gerhardt it is a mixture of melamine and
mellon ; M. Voelkel, at first, regarded it as impure poliene :
C4N4H4.

Poliene is a substance isomeric with melamine, but neutral
or very faintly basic, insoluble in water and in a weak solution
of caustic potash.

But in his subsequent attempts to purify his poliene, M.
Voelkel finding that the only constant product was melamine,
concluded that these two substances were identical; and he re-
gards melam as a mixture of melamine with glaucene, C4 H N3,
ammelene C6 H5 N3 (?), and albene, C12 H9 N10 03 (?).

X 2

308 ON THE ARTIFICIAL FORMATION

It is evident that none of these opinions is exactly correct.

In the first place impure melara cannot contain oxygen, since
sulphocyanide of ammonium is devoid of that element; and
melam cannot contain any considerable quantity of melamine,
in consequence of the marked differences in the properties of
those bodies.

Let us now explain the way in which we conceive the
reaction to take place, basing our ideas on the fact demonstrated
by Vcelkel, that melamine, and the body contained in the
melam, can lose ammonia by heat and be converted into glau-
cene.

CMH„NM H9N3 = C12H3N9

4 at. melamine. 3 at. ammonia. 3 at. glaucene.

On distilling sulphocyanide of ammonium, there remains as a
residue a substance which, when pure, is expressed by the for-
mula:

C6 H6 N6 = 2 at. C3 H3 N3
and which constitutes pure melam (poliene?).

But too high a temperature may alter a certain quantity,
and liberating some ammonia may given rise to a corresponding
quantity of glaucene.

(Mellon, which may contain crude melam, must arise from
a more advanced decomposition which especially takes place on
the walls of the retort).

Hence it may be easily understood that crude melam may
give as the result of analysis, the formula C12 H9 Nu; for, in
point of fact :

C12H12N12 - H3N = C^N,,

Pure melam. Ammonia. Crude melam.

Hence crude melam must be a mixture of pure melam with
a certain quantity of glaucene.

We regard melam as analogous with furfurolamide and
hydrobenzamide.

On submitting it to prolonged ebullition with a weak solu-
tion of caustic potash, it gradually undergoes a transformation ;
from being insoluble and amorphous, it becomes soluble and
assumes a crystalline state, and the octohedra, which it finally
deposits in the alkaline fluid, consist of a new substance, isome-

OF ALKALOIDS. 309

ric with that from which it was derived, but possessing also the
characters of a tolerably strong alkaloid.

In short the melam is converted into melamine.

Under the prolonged influence of alkalies, and by the action
of acids, melamine becomes successively transformed into amme-
line, ammelide, and lastly into hydrated cyanuric acid. These
changes are induced by the successive loss of one eq. of ammonia
which is replaced by two eq. of water. They are expressed by
the following equations, which at the same time indicate the
relations existing between these different bodies :

C6 H6 N6 + 2 H O = H3 N + C6 H5 N5 02

Melamine. Ammeline.

C6 H6 N6 + 4 H O = 2H3N + C6 H4 N4 O,

Ammelide.

C6 H6 N6 + 6 H O = 3 H3 N + C6 H3 N3 06

Hydrated cyanuric acid.

Hence we see that melamine, ammeline, and ammelide may
be regarded as arising from different cyanurates of ammonia,
the mono, bi, and tribasic forms, having lost for each atom of
base one atom of water*.

C6 H3 N3 06, H3 N — 2 H O = C6 H4 N4 04

Cyanurate of ammonia. Ammelide.

C6 H3 N3 06, 2 (H, N) - 4 H O = C6 H5 N5 02

Di-cyanurate of ammonia. Ammeline.

C6 H3 N3 0„ 3 (H, N) - 6 H O = C6 H6 N6

Tri-cyanurate of ammonia. Melamine.

We have adopted Gerhardt's formula for ammelide, for the
following reasons :

1. Because the combination of ammelide with oxide of
silver, the most distinctive combination that this body is capable
of forming, corresponds better with the formula of M. Gerhardt
than with that of M. Liebig.

2. Because the formula C6 H4 N4 04 explains in a better
and more simple manner the relations subsisting between the
four bodies mentioned.

* Gerhardt, Precis de Chimie organique.

310 ON THE ARTIFICIAL FORMATION

3. Because Liebig's formula leads us to admit the existence
of a body possessing the singular property of being only capable
of combining with 2 at. of acids, 2 at. of oxides, and 2 at. of
salts, a circumstance altogether exceptional, and which does not
occur in any other body of the class of alkaloids.

For the more complete explanation of this last fact, let us
express the bases in question by the notation of equivalents
which has latterly been exclusively adopted by Liebig :

Melamine will be expressed by . C6 H6 N6

Its combination with acids by . C6 H6 N6 + A

Its combination with nitrate of silver

by . •. C6H6N6+ N05,AgO

Ammeline will be expressed by . C6 H5 N5 02

Its combination with acids by . C6 H5 N5 02 + A
Its combination with nitrate of silver

by • • C6H5N502 +N05,AgO

Ammelide will be expressed by the

formula C6 H4i N4^ 03

But as a formula cannot contain half-equivalents, we must
double it; when we have:

Ammelide = C12 H9 N9 06

Its combination with acids

will be C12 H9 N9 02 + 2 A

And the combination with

nitrate of silver . . . C12 H6 N9 06 + 2(N05, AgO).

Ammelide may also be developed under the following cir-
cumstances :

On exposing nitrate of ammeline to the action of heat,
water and protoxide of nitrogen are simultaneously liberated.

C6 H5 N5 02 + N H 06 = C6 H4 JST4 04 + 2 N 0+ 2 H O.

Nitrate of Ammeline. Ammelide.

By the dry distillation of urea,
C8 H16 Ns 08 = 2 C 02 + 4 H3 N + C6 H4 N4 04

l»i at. urea. Ammelide.

OF ALKALOIDS. 311

MM. Laurent and Gerhardt have further announced that
ammelide occurs in the residue of the calcination of hydro-
persulphocyanic acid. These able chemists meant to imply
that ammelide was one of the products of the action of potash
on this residue.

M. Liebig has very rationally observed, that a body not
containing oxygen, cannot, when decomposed by heat, yield, as
a product, a compound containing oxygen.

6. Lophine, C23 H8 N.

On heating hydrobenzamide (C21 H9 N), ammonia and a
very fluid and odorous oil are first disengaged. When ammonia
ceases to be developed, the residue left in the retort must be
poured into a mortar, and after being exhausted with boiling
ether, must be placed in boiling alcohol, into which fragments
of caustic potash must be gradually added until perfect solution
is effected.

The lophine is deposited as the fluid cools.

This base is also developed in the dry distillation of the
mixture obtained by acting on the essence of bitter almonds
with sulphide of ammonium.

7. The Alkaloid derived from Cacotheline.

It has been shown by MM. Gerhardt and Laurent, that
when nitric acid acts on brucine, nitrous ether is liberated,
and a peculiar yellow body is formed, to which M. Laurent has
given the name of cacotheline.

According to this illustrious chemist the reaction is expressed
by the following equation.

C46 H26 N2 08 + 3 (N 06 H) = 2 (C21 Hu N2 O10)

Brucine. Nitric acid. Cacotheline.

+ C4 H5 N 04 + 2 H O

Nitrous ether.

Cacotheline submitted to the action of ammonia is trans-
formed into a new alkaloid, which apparently includes the
nitrous vapour.

Its composition has not yet been established.

312 ON THE ARTIFICIAL FORMATION

8. Aniline, C12 H7 N.

In consequence of its stability, its characteristic properties,
the numerous combinations into which it enters, and the highly
interesting products to which it gives rise, we may regard
aniline as the most important of the artificial bases.

In numerous circumstances its behaviour is similar to that
of ammonia, and in many cases the study of aniline will faci-
litate that of the ammoniacal compounds.

This alkaloid is also produced under extremely various
circumstances, and there are many different ways in which it
may be formed.

The following are the principal ways in which it may be
produced.

1. On decomposing phenate of ammonia by heat, the phenol
(C6 H3 O,) combines with ammonia, and the combination
heated to 392°F. in a closed tube over a lamp, gives rise to
aniline and water.

C12 H602,H3Nr2HO + C12 H7 N.

Aniline is then a phenamide, in the same way as melamine,
ammeline, and ammelide are cyanuramides.

2. On applying heat to anthranilic acid, it becomes at once
decomposed into carbonic acid and aniline. Anthranilic acid is
then isomeric with bicarbonate of aniline.

C14 H,N04=2C02+ C12 H7 N

Anthranilic acid. Aniline.

3. By the decomposition of isatine when acted on, at a high
temperature, by hydrate of potash, carbonate of potash is
formed, and hydrogen liberated.

C16H5N04+4(KO, HO) = C12H7N + 4(KO,C02) + Ha

Isatine. Potash. Aniline.

4. By the decomposition of nitro-benzoene, when acted on
at a high temperature by carbonate of lime.

C14 H7 N 04 = 2 C 02 + C12 H7 N.

5. By the decomposition, under similar circumstances, of
salicylamide.

C14 H5 04, N H2 = 2 C 02 + C12 H7 N

Salicylamide. Aniline.

OF ALKALOIDS. 313

6. By the decomposition of nitrobenzide by sulphuretted
hydrogen.

Cia H5 N 04 + 6 H S = C12 H7 N + 4 H O + 6 S

Nitrobenzide . Aniline .

7. In the preparation of M. Mitscherlich's azobenzide, that
is to say, in distilling nitrobenzide with an alcoholic solution of
potash, there are formed not only aniline and azobenzide but
also oxalic acid and water.

| C19 H5 N azobenzide.

2 (C H5 N 04) + 2 (C H3 O) = ^/o^t^acid.
Nitrobenzide. Alcohol. | 4 H2 02%vater.

SECOND GROUP.

Reduction by sulphuretted hydrogen or sulphide of am-
monium of nitrogenous combinations, formed from the hydro-
carbons, by the substitution of 1 or 2 eq. of nitrous vapour,
(N2 04) for 1 or 2 at. of hydrogen.

1. Aniline, by the reduction of nitrobenzide.

2. Xitraniline „ „ binitrobenzide.

3. Toluidine „ „ nitrobenzoene.

4. Naphtalidine „ „ nitronaphtalide.

5. Seminaphtalidine „ binitronaphtalide.

This mode of forming alkaloids promises a rich future harvest.

We already possess a very considerable number of hydro-
carburets, and further so many means of producing them, that
science will never be at a loss in furnishing them.

Moreover most of the carburets of hydrogen are attackable
by nitric acid, and we shall certainly be able to substitute N2 02
for H2 of the hydrocarburet.

Having once obtained the compound, all we have to do is to
dissolve it in alcohol, and to submit it to the action of sulphu-
retted hydrogen, or of an alcoholic solution of sulphide of am-
monia. Reduction will ensue, and an alkaloid will be found
amongst the new products.

The organic bases formed by this method are generally very
well characterised, crystallizable, have an alkaline reaction, and
readily form crystalline and distinct salts.

314 ON THE ARTIFICIAL FORMATION

They scarcely ever contain oxygen, unless they include the
elements of nitrous vapour, (nitraniline).

As the process of formation is absolutely the same for all
the different bases of this group, and since, further, (except in
the case of aniline,) we do not know of any other method of
preparing them, we shall place before our readers the equations
illustrating their formation.

It must be remembered that nitrobenzide and binitrobenzide
are produced by the action of nitric acid on benzol, (C12 H6);
and nitrobenzoene or nitrotoluide by the action of the same acid
on benzoene or toluol (C14 H8); and nitronaphtalide (nitro-
naphtalese) and binitronaphtalide, (nitronaphtalese) in the same
way, from naphthaline (C20 H8).

Reducing by means of sulphuretted hydrogen we have,

C12 H5 N 04 + 6 H S = C12 H7 N + 4 H O + 6 S

Nitrobenzide. Aniline.

C12 H8N208 + 6HS = C12(H6N02)N + 8HO + 6S

Binitrobenzide. Nitraniline

C14H7N04+6HS^C14H9N + 4HO + 6S

Nitrobenzoine. Toluidine.

C20 H7 N 04 + 6 H S = C20 H9 N + 4 H O + 6 S

Nitronaphtalase. Naphtalidine.

C20 H6 N2 08 + 12 H S = 2 (C10 H5 N) + 8 H O + 12 S

Nitronaphtalise. Seminaphtalide.

Note. Nitrous ether, which has for its formula, C4 H5 O,
N 03, or C4 H5 N 04, may be regarded as a hydrocarburet,
C4 H6, in which an equivalent of hydrogen has been replaced
by an equivalent of nitrous vapour, N 04. If such an arrange-
ment occurred, the action of the sulphuretted hydrogen would,
conformably to analogy, be expressed by the following equation :

C4H5N04 + 6HS=C4H7N + 4HO + 6S

Nitrous ether. New alkaloid.

We should obtain the compound C4 H7 N, which very probably
possesses the characters of an alkaloid,

These are the considerations which have induced me to
examine the action of sulphuretted hydrogen on nitrous and
nitric ethers.

OF ALKALOIDS. 315

I. Reduction of Nitrous Ether by Sulphuretted Hydrogen.

The nitrous ether was prepared in the manner that I des-
cribed some years ago in my Chemical Thesis, submitted to the
Faculty of Sciences of Strasburgh. It consists in making
equal volumes of alcohol and nitric acid react on each other
in contact with iron filings or fragments of copper.

The reaction always proceeds quietly, even when operating
on considerable quantities, and is almost entirely accomplished
without the external application of heat.

The vapour of nitrous ether is first passed through a flask
filled with water, for the purpose of washing it; then through a
long tube containing chloride of calcium, for the purpose of
drying it ; and it is finally condensed in a small globular receiver
placed in a powerful freezing mixture.

Nitrous ether obtained in this manner is perfectly pure and
free from aldehyde.

In fact, on decomposing it by an alcoholic solution of potash,
(a decomposition which only takes place slowly,) it remains
perfectly colourless. The slightest trace of aldehyde would have
given rise to the production of a brown tint.

In this process it is evident that the copper serves to reduce
the nitric acid to the state of nitrous acid, by abstracting oxy-
gen from it, and it is the nitrous acid thus developed which
reacts on the alcohol.

Nitrous ether is very readily reduced by sulphide of ammo-
nium. The two bodies are hardly in contact before there is
an energetic reaction, and in the course of one or two minutes
the reduction is complete.

In order to study the products of this reaction, I have
adopted two different modes of procedure.

a. An aqueous solution of recently prepared sulphide of
ammonium, is introduced into a pretty strong flask with a
ground glass stopper; above there is carefully poured a layer
of nitrous ether. The flask being well closed, the whole is
thoroughly shaken. An energetic reaction instantly occurs,
accompanied with the disengagement of heat, and the contents
of the flask momentarily assume a brick-red colour, in conse-
quence of the formation of per-sulphide of ammonium. This
is immediately followed by decoloration of the fluid and by an
abundant precipitation of sulphur.

316 ON THE ARTIFICIAL FORMATION

By successive additions of nitrous ether and sulphide of
ammonium, we may readily achieve the complete decomposition
of these two bodies.

The precipitated sulphur having been separated by filtra-
tion, the liquid is carefully distilled; with the exception of a
minute quantity of sulphur, no residue is left in the retort.

The distilled fluid possesses a strong ammoniacal odour. It is
neutralized with hydrochloric acid, and again submitted to
partial distillation.

The distilled fluid possesses a weak alcoholic, and very
slightly alliaceous taste ; when heated to the boiling-point the
vapour becomes inflammable.

The remaining fluid yields an abundant crystallization of
hydrochlorate of ammonia; on concentrating a portion to dry-
ness, it was impossible to detect anything except that salt.

Hence nitrous ether reduced by sulphuretted hydrogen, gives
origin to alcohol and ammonia.

C4H5N04 + 6HS=3C2H30 + 2HO + H3N+6S

Nitrous ether. Alcohol.

b. Nitrous ether is dissolved in absolute alcohol, containing
a little ammoniacal gas, and a stream of sulphuretted hydrogen is
passed through the mixture. A precipitation of sulphur ensues;
the distilled fluid yields no residue, the alcoholic fluid in the
receiver is ammoniacal, but at the same time possesses an odour
of mercaptan.

II. Reduction of Nitric Ether by Sulphuretted Hydrogen.

The nitric ether was prepared according to the method de-
scribed by Millon, was washed with plenty of water, dehydrated
and rectified.

Its reduction proceeds much more slowly than that of nitrous
ether, and as it is far less volatile, the reaction can be much
more easily watched.

On passing a current of sulphuretted hydrogen through
nitric ether, diluted with ammoniated alcohol, the mixture
begins to boil at between 95° and 104° F. It deepens in colour,
and we soon observe minute needles of sulphur deposited at the
bottom of the vessel.

OF ALKALOIDS. 317

They rapidly increase in quantity, and if the nitric ether
has not been diluted with alcohol, the whole assumes the cha-
racter of a half solid magma.

The deposited sulphur is, at intervals, removed by filtration,
and the fluid again submitted to the action of sulphuretted
hydrogen.

This part of the operation being concluded, we find, that on
distilling the alcoholic fluid, no residue is left in the retort, with
the exception of a trace of nitrate of ammonia, which is evi-
dently dependent on the regular decomposition of nitric ether
by an alkaline base, and a little hyposulphite of ammonia, the
latter being the result of the action of the air on an alcoholic
solution of hydrosulphuret of ammonia.

In fact, in an imperfectly closed flask containing the last-
named substance in solution, we observe that in the course of a
few days the fluid becomes colourless, and deposits very thin,
magnificent, hexagonal lamina?, forming a considerable surface
of very pure hyposulphite of ammonia.

The distilled alcoholic fluid contains ammonia and mercaptan.

The ammonia having been saturated with hydrochloric acid,
the fluid, rectified by a fresh distillation, precipitates the salts of
lead with a yellow, and the chloride of mercury with a white
tint ; the precipitate dissolves in boiling alcohol, and reappears
on cooling, in the form of brilliant scales.

Nitric ether reduced by sulphide of ammonium, without
the intervention of alcohol, also yields mercaptan.

THIRD GROUP.

Distillation of Organic Substances with very Concentrated
Hydrate of Potash.

In this group we should place two alkaloids which are ordi-
narily grouped amongst the natural alkaloids; these are, conine,
Cl6 H15 X; and nicotine, C10 H7 N.

These two bases do not naturally exist in the plants from
which we obtain them, but result from the action of hydrate of
potash on particular substances or principles, of which that

318 ON THE ARTIFICIAL FORMATION

occurring in tobacco is nicotianine, while that in the Conium
maculatum is still unknown.

This reaction is probably similar to that of potash on isatine,
which gives rise to aniline.

The alkaloids beloning to this group are :

1. Quinoleine C18 H7 N, resulting from the action of hy-
drated potash, on quinine, cinchonine, and strychnine.

2. Aniline, resulting from the action of potash on isatine.

3. Monochloraniline, derived from chlorisatine.

4. Bichloraniline, derived from bichlorisatine.

5. Monobromaline, derived from bromisatine.

6. Bibromaniline, derived from bibromisatine.

7. Conine, derived from an unknown compound.

8. Nicotine, derived from nicotianine.

1. Quinoleine (Leukol), C18 H7 N.

It is only quite recently that M. Gerhardt, to whom science
is indebted for the discovery of quinoline, has ascertained that
it is identical with a base, leukol, which was previously known,
and which is obtained from coal-tar. It occurs together with
aniline (kyanol).

It is a fact not devoid of interest that while quinoleine is
thus withdrawn from the natural alkaloids already described,
aniline may also be obtained from pipeline which approximates
closely to the alkaloids.

The reactions, giving rise to the formation of quinoleine, are
easily expressed by the following equations :

C19 Hn N02= C18 H7N + C02 + H4

Quinine. Quinoleine.

C19 Hn N O + H O = C18 H7 N + C 02 + H5

Cinchonine. Quinoleine.

C22 H12 N 02 + 6 H O = C18 H7 N + 4 C 02 + Hn

Strychnine. Quinoleine.

In the dry distillation of coal, the quinoleine is probably
formed in a manner analogous to that in which aniline is pro-

OF ALKALOIDS. 319

duced by the decomposition of a salt corresponding with phenate
of ammonia. The following will be the reaction that ensues :

Cis H6 02 + X EJ3 = C18 H7 N + 2 H O.
2. Aniline and its derivatives.

Although we have already given the formula for the pro-
duction of aniline from isatine and potash, we shall here repeat
it in order to shew the identity between its preparation and
that of its derivation by substitution.

The chlorine and bromine of chlor- and bromisatine occur
as integral portions of chlor- and bromaniline :

Cie H5 N04 + 4 (K O, H 0)= C12 H7 N + 4 (K O, C 02) + 2 II

Isatine. Hyd. potash. Aniline. Carb. potash.

C16 (H4 CI) N 04 + 4 (K O, H O) = C12 (H6 CI) N

Chlorisatine. Chloraniline.

+ 4 (K O, C 02) + 2 H
CI6 (H, C1.2) N 04 + 4 (K 0,HO)= C12 (H5 Cl2) N

Bichlorisatine. Bichloraniline.

+ 4 (K O, C 02) + 2 H
C16 (H4 Br) N 04 + 4 (K 0,H0)= C12 (II6 Br) N

Bromisatine. Bromaniline.

+ 4 (K O, C 02) + 2 H
CI6 (H, Br2) N 04 + 4 (K 0,H0)= C12 (H5 Br2) N

Bibromisatine. Bibromaniline.

+ 4 (K O, C 02) + 2 H.

FOURTH GROUP.
Dry Distillation of Organic Matters.

In addition to the well-known alkaloids this group contains
some other organic bases which are only imperfectly known ;
as, for instance, pyrrhol, odorine, and animine, noticed by
Unverdorben.

We know scarcely anything of their properties, and still less
of their nature and composition.

Dry distillation yields the following alkaloids :
1 . Aniline, quinoleinc, and pyrrhol by the dry distillation of coal.

320 ON THE ARTIFICIAL FORMATION

2. Odorine and animine by the distillation of animal matters ;

they ought to be included in Dippel's animal oil.

3. Nicotine by the dry distillation of tobacco.

4. Lophine by the distillation of hydrobenzamicle and of the

sulphurous products of hydride of benzoile.

FIFTH GROUP.

Decomposition of a Nitrogenous Acid.

This group yields two extremely interesting and remarkable
examples: one is the decomposition, by the dry method, of
anthranilic acid into carbonic acid and aniline :

C14 H7N04=:2C02+ C12 H7 N

Anthranilic acid. Aniline.

For the second we are indebted to the recent investigations of
M. Dessaignes; it is the transformation of hippuric acid into
benzoic acid and sugar of gelatine.

Both acids and bases effect this transformation. In conse-
quence of the great analogy between urea and sugar of gelatine
we feel compelled to place the latter amongst the alkaloids.

SIXTH GROUP.
Desulphuration of a Sulphuretted Alkaloid.

Under this head we must place the transformation of thiosi-
namine into sinamine.

Thiosinamine, C4 H4 N S, when treated with oxide of lead,
reacts as if it were an organic combination of 2 at. of sulphu-
retted hydrogen with sinamine. Water and sulphide of lead
are formed, and the new alkaloid, sinamine, is liberated :

C4 H4 N S + Pb2 O = Pb2 S + H O + C4 H3 N

Thiosinamine. Sinamine.

SEVENTH GROUP.

Desulphuration of a Sulphuretted Essential Oil.

The hydrated oxides of lead and mercury, when in contact
with the essential oil of mustard, break it up into bisulphide of
carbon and into a peculiar alkaloid, sinapoline, whose formula
is C7 H6 N 0.

OF ALKALOIDS.

321

C8 H5 N S2

Essence of mustard.

+ HO = C7 H6 N O + C S2

Sinapoline. Bisulphide of carbon.

EIGHTH GROUP.
Alteration of Natural Alkaloids.

a. Alteration of Narcotine by Oxidation.

The excellent investigations of MM. Wohler and Blyth on
narcotine, which throw so much light on the most intimate con-
stitution of that alkaloid, have brought to our knowledge new
artificial organic bases, one of which, narcogenine, appears to
form an integral constituent of narcotine.

On boiling a solution of narcotine in hydrochloric acid with
bichloride of platinum, this latter loses some of its chlorine,
which decomposes the water, and the resulting compounds are
protochloride of platinum, 1 eq. of opianic acid, and 1 eq. of
cotarnine.

C46 H25 N 014

Narcotine

Water

Bichloride of platinum 4 Pt Cl2

H202

4 Pt CI Protochloride of platinum.
4 H CI Hydrochloric acid.

Opianic acid.

Cotarnine.

^20 "10 ^10

C26 H13 N 06

The two last substances may also be obtained on oxidising
narcotine with a mixture of sulphuric acid and peroxide of

"ancoc.

r c20 H10 o10 ....

.... Opianic acid.

C46 H25 N 0I4 + 04 =

<^ C26 H13 N 06

.... Cotarnine.

Narcotine.

I H2 02

.... Water.

But during the oxidation of narcotine, there is also formed
another base to which the name of narcogenine has been given.
The following formula fully explains the nature of its formation :

2 eq. of narcogenine = 2 (C36 H19 N O10) = C72 H38 N2 O20

^72 -""38 ^2 ^20

C46 H25 N 014
C26 H13 N 06

Narcotine.
Cotarnine.

Narcogenine.

Hence it follows that in its nascent condition, the cotarnine
combines with the still undecomposed narcotine to form 2 eq. of

narcogenine.

We may yet further regard the oxidation of the narcotine
to take place as follows :

^36 Hl9 ^ ^10 ••

. Narcogenine.

cio H5 05 ..

.. Opianic acid.

H 0 ...

. Water.

322 ON THE ARTIFICIAL FORMATION OF ALKALOIDS.

C46 H25 N 014 + 02 =

Narcotine.

The formulae adopted for narcotine and narcogenine are
those of Dr. Blyth; those for cotarnine and opianic acid are the
formulae corrected by M. Gerhardt, to whom, conjointly with
M. Laurent, we owe the possession of formulae representing in
equations the reactions we have described.

b. Derivatives of Cinchonine by substitution.

It has been shown by M. Laurent that on allowing chlorine
and bromine to react on hydrochlorate of cinchonine we obtain
certain new alkaloids; namely, chlorocinchonine and bromocin-
chonine, which are derived from the cinchonine by the substitu-
tion of chlorine and bromine in place of the hydrogen of that
organic base.

Amongst the artificial alkaloids it yet remains to notice
guanine, C10 H5 N5 0.2, which should almost be regarded as a
natural alkaloid, since it is formed by processes occurring in
nature, with which we are still unacquainted, and independently
of the aid of chemical manipulations.

It probably arises from the decomposition of uric acid under
the influence of reducing substances.

Such are the reactions by which the artificial alkaloids are
produced : many of them, especially these last, present an
altogether special character which forbids us to hope for many
analogous instances; but those earlier described, admit of gene-
ralization, and the views regarding them may be extended to a
large class of bodies.

We may entertain a hope that amongst the new alkaloids
thus produced, especially if we conduct our researches on fami-
lies characterised by the same number of atoms of carbon, we
may discover bases identical with the natural alkaloids.

323

VIII.

MEMOIR.

ON THE INTIMATE CONNECTION EXISTING BETWEEN THE
PSEUDO-VOLCANIC PHENOMENA OF ICELAND.

By R. Bunsen.

A careful study of the relations under which the innumerable
systems of Thermal springs and Fumeroles are manifested in
Iceland, must convince every attentive observer that an inti-
mate connection exists between them and the active volcanoes
of the island. The latter, upheaved on fissures, intersect the
whole island in a parallel system of longitudinal lines, whose
north-north-eastern extremity corresponds perfectly with the
parallel expansion of the principal valleys and elevated ridges
of the main system, and with the numerous volcanic fissures
and dykes with which the country abounds ; whilst the fume-
roles and thermal systems appear to be likewise connected with
this main direction. Following this line, independently of the
course of the valleys and elevated ridges from which they
proceed, they may often, as at Krisuvik or in the neighbour-
hood of Krafla and Leiznukr, be traced in lines passing trans-
versely over mountains and valleys. This circumstance, and
the common occurrence of solfataras, suffiones, thermal springs,
and geysers, and still more the intimate connection existing
between the phenomena of decomposition to which they give
rise, lead, even on a superficial observation, to the conclusion
that all these phenomena are merely to be regarded as modified
expressions of one and the same fundamental cause.

Y 2

324 PSEUDO-VOLCANIC PHENOMENA

In order to attain to a deeper insight into these intimately
connected phenomena, it will, in the first place, be necessary to
throw a glance on the hitherto imperfectly understood geognostic
and hydrographic character of Iceland, and by these means to ac-
quire a clear idea of the relations on which depends the remark-
able development of the phenomena of the Icelandic springs.

The geological character of the volcanic system of Iceland
is most strongly marked in the relations existing between the
palagonite-mountain and the plutonic masses which have pene-
trated through it, and still continue to penetrate it during the
eruption of volcanoes, still active, in the form of currents and
strata of lava.

The palagonite-tuff* forms the most ancient member of this
series of formations. Its formation immediately preceded the
elevation of trachyte and of clinkstone, which last passes into
the former, and both of which, owing to their relations of super-
position and expansion, play a much less important part than is
ascribed to them by Krug of Nidda in his Treatise on Iceland.
The penetration of the older trap, a doleritic rock which is trans-
formed into the most various kinds of amygdaloid, characterizes
the third and most important period of elevation, during which
this rock was upheaved in large veins, and frequently spread
itself laterally in far extended strata through the tufFaceous
masses. We may regard as appertaining to the fourth period,
the olivine and basaltic elevations which have penetrated the
trap formations in systems of veins of different ages, and
which are especially manifested in the Island of Yidhey, in the
Bay of Reykjaviker and on the Esja, as well as at innumerable
other points of the northern and southern part of the island.
The series of plutonic elevations closes with the fifth period of
the older and more recent lavas.

The structural relations of the trap system which predomi-
nates over the whole island, exercise, in their relation to
palagonite-tufF, the most marked influence on the hydrographic
condition of the island. The general character of the stratifica-
tion of this class of rocks is manifested by a tendency to
assume a horizontal direction and a flattened superficies, inde-
pendently of the innumerable elevations and dislocations occa-

* Named from Palagonia in Sicily.

OF ICELAND. 325

sioned by local causes ; so that the trap-strata on the declivities
of the abruptty sloping valleys generally terminate in hori-
zontal or gently inclining lines between the tuff-strata. This
structure is most intimately connected with the general level
of the country, which rises in a flat ascending arch, until it at-
tains an elevation of 689 meters above the level of the sea, near
the middle of the island, on the Sprengisandz, the point where
the waters diverge to the north and south, This gently -rising
elevated plateau, covered by innumerable moving glaciers,
forms the basis of the grand ice-berg, called by the Icelanders,
the Jokull, and which terminates similarly in a flat superficies,
manifesting, with the exception of occasional sharp projections
from its base, the structure in plutonic strata analogous to the
above.

Inaccessible fields of snow cover the summits of the moun-
tains, and reveal, at great distances, the limits of the region of
glaciers, which penetrate with their huge masses of ice for a
length of many miles, even to the lower range of the plateaux,
and may be traced by the bluish reflection of their dazzling
masses in the glacier ice. It is owing to these ice-bergs, which
cover almost a tenth part of the island, that Iceland, taking
into account its climatic relation, is characterized by so remark-
able an abundance of atmospheric deposition, and it is to the
same cause we must refer the singular development of the
phenomena of springs, which is intimately connected with the
peculiar structural relations of the palagonite rock. Vast
masses of water break through the fissures and arches of the
glaciers, or rush in cascades down the icy walls of the moun-
tain-slopes, not unfrequently converting a district of many
miles into a bottomless mass of moving mud, in which the
streams accumulate before they can form for themselves a
well-defined and regular bed for their waters. Innumerable
inland seas, vast marshes and swamps, which make this barren
and desolate country appear even more terrible to the eye of
the traveller, are the consequence of such overflowings, dif-
fusing a mass of waters over the elevated plateau of this island,
which, finding its way into the deep declivities along the
gently-inclining strata of rocks, serves to nourish the various
systems of springs.

326 PSEUDO-VOLCANIC PHENOMENA

The dykes and fissures corresponding with this volcanic line
of elevation must necessarily interrupt the course of these sub-
terranean waters, and cause them to diverge into those deep
ravines where a process of heating and evaporation must un-
avoidably be induced from the action of the heat of the volcanic
soil. The water, elevated by the combined force of elastic
vapour and hydrostatic pressure, then breaks forth in thermal
springs, whose frequent north-eastern direction necessarily
stands in the closest connection with the above-named general
geognostic relations of the island. That it is from meteoro-
logical depositions that water is conveyed to the thermal
springs, and a connection established, in consequence, between
atmospheric and volcanic phenomena, is proved by the most
undoubted facts. In the first place, nitrogen, which rises from
the thermal springs, either pure or mixed with other gases,
affords certain evidence of the atmospheric origin of these
waters, since this gas> which cannot, by any possibility, be
counted amongst the direct products of volcanic activity, has
never yet been found, as far as can be ascertained by any ap-
proximative calculation, in a larger proportion to the mass of
thermal waters than that corresponding with the solubility of
the gases of the atmosphere in water. Another proof of the
atmospheric origin of these springs is afforded by the small
quantity of ammoniacal salts and organic extractive matters
which are found to exist in the water, even where the latter is
in further connection with the atmospheric air than that afforded
by the narrow mouths of the springs. These observations do
not, however, in any way exclude the possibility that the great
quantity of water contained in the palagonite of the tuff-strata
may take an active part in the development of gases by which
the suffiones and geysers are characterized, and which, on
occasion of the great eruptions of the volcanic islands of
Iceland, force a passage for themselves through the craters
in huge columns of ashes bearing the form of colossal pines.

A simple calculation shows that palagonite, whose density is
2*43, and whose combined water amounts to 17 per cent., should
generate a quantity of vapour on being heated; which at 0° C.
and Bar. 0*m76 occupies 512*7 times the volume of the original
rock, and which must necessarily find vent where the ux>heaved

OF ICELAND. 327

fluid masses of lava come in contact with that rock at some
great subterranean depth. The degree, however, in which
these aqueous contents contribute to the formation of thermal
springs involves points of consideration that are not included
in the sphere of experimental investigations.

One of the most important questions associated with the
Icelandic thermal springs has reference to the nature of their
composition.

If it be true that the great atmospheric process of distilla-
tion gives rise to the flow of water to the springs, we may
expect that their mineral constituents likewise admit of being
explained by a reciprocal reaction of the originally pure water
and of the gases dissolved in it on the rock constituting the
base of the spring. Geology has, unfortunately, limited itself
exclusively to hypothesis regarding this subject, or to the
exposition of mere possibilities without investigating it experi-
mentally with the circumstantiality it merits. Iceland presents
the best field for such an inquiry in its remarkable thermal
springs, and the innumerable geysers and suffiones which afford
an exhibition of the most striking decompositions, and furnish
us with a key for the experimental solution of many, I might
almost say all, the questions which have hitherto been directed
to these mysterious phenomena.

The Icelandic mineral springs, to which belong all the
systems of geysers and suffiones, are distinguished from all
others in Europe by the proportionally large quantity of
silica which they contain; and if we except the acidulous
springs which are confined to the western part of the island, the
so-called beer -springs (olkilder,) of the natives, we may divide
the springs of Iceland into two main groups, according to their
chemical properties, one of which would comprise the acid and
the other the alkaline silica springs. The former belong to the
actual solfataras, (the Ndmar of the Icelanders,) and owe their
slight acid reaction more commonly to the presence of a small
quantity of ammonia-alum or soda and potash-alum, than to
their inconsiderable traces of free sulphuric or muriatic acids.
They contain, besides, sulphates and chlorides of calcium, mag-
nesium, sodium, potassium, and iron, also silica and sulphurous
acid, or in the place of the latter, sulphuretted hydrogen

328 PSEUDO-VOLCANIC PHENOMENA

gas. They are especially characterized by depositions of
gypsum and sulphur. Periodic springs of eruption are of very
rare occurrence. The alkaline springs are the most preva-
lent, and constitute the periodic bubbling springs (geysers),
as -well as those designated in Iceland by the name of Huer,
and which are generally warm, or even boiling. Their ex-
tremely slight alkaline reaction is owing to the presence of
alkaline sulphides and the carbonates of soda and potash, which
serve as solvents for the silica, and give rise to the silica-tuff
formations by which these springs are so strongly characterized.
Alkaline sulphates and hydro-chlorates are the salts most
commonly found in these waters; they also generally exhibit
traces of magnesia.

The process from which all these mineral waters derive
their constituents is so much the more interesting from the
detailed manner in which it may be traced. It appears in a
series of very remarkable decompositions, experienced by the
palagonite under the influence of the volcanic gases.

The following is the formula which I determined for this
rock from experiments which have been given in detail in
another Memoir*:

Mg3]

LSL+2

K 3 | | Al

Si + 9H.

* The experiments to which allusion is made yielded the following results

Oxygen Oxygen of the

* ° Isomorphous Bodies.
Silica 37*417 19-4280 19*4280 4

94670

Sesqui- oxide of iron 14*175 4*2480)

Alumina 11*165 52190>

Lime 8*766 2*4928]

Magnesia 6'036 2*33821

Potash 0*685 01163 [ 5*1146 1

Soda 0*652 01673 )

Water 17152 15*2490 15-2490 3

Insoluble residue 4-108

100-156
If we assume that the oxygen of the isomorphous bases is as 4 : 2 : 1 : 3, and

OF ICELAND.

329

The volcanic gases of Iceland, which act upon this rock
and are acted upon by it, only differ from those of European
volcanoes by containing a less considerable quantity of carbonic
acid. They consist, as is generally the case, of sulphurous acid,
sulphuretted hydrogen, carbonic acid, and hydro-chloric acid.

that of the whole amount 49-2586 of the oxygen, JL. belongs to the lime and
its isomorphous bases, -2- to the alumina and sesqui-oxide of iron, -JL. to the
water, and'-A_ to the silica ; and if, further, we divide the quantity of oxygen thus
calculated among the separate isomorphous bases, making it proportional to the
respective quantities of oxygen found, we shall obtain the following formula,
which does not differ more considerably from the results of the experiment
than might be expected from an isomorphous mixture.

Silica 37-947

Sesqui-oxide of iron .... .... .... 14-751

Alumina ... .... .... .... .... 11619

Lime 8442

Magnesia .... .... .... .... .... 5*813

Potash 0*659

Soda (M;28

Water 16621

Residue 4*108

100588

This leads to the formula

Si., + 2

Fe
Al

Si + H.

Paiagonite approaches, therefore, very nearly to Oltrelite, which was
examined by Damour, and of which the formula is —

Mn,

SL, + 2AlSi+3H.

Its place in a system, will, therefore, be next to that body. It only
differs in its proportion of water from the formula of several of the Skapolitres
investigated by E. Th. Wolf. The general formulae of these three minerals are
in very close relation to each other, as will be seen by what follows :

Skapolitre

Oltrelite

Paiagonite

R3 Si2 + 2R Si.

R3Si, + 2RSi + 3H.
.. R3Si2 + 2RSi + 911.

330 PSEUDO-VOLCANIC PHENOMENA

The latter, both here and at Etna, constitutes a far less im-
portant feature than at Vesuvius.

The chemical relations of nitrogen, ammonia, and their
compounds, teach us with certainty, that although these sub-
stances are scarcely ever absent from volcanic exhalations; they
are foreign to the actual force of plutonic activity. They
undoubtedly belong originally to the atmosphere, or to organic
nature, their occurrence being due to the water which holds
them in solution and conveys them from the air to these sub-
terranean depths.

The great abundance of the sublimations of muriate of
ammonia observed in the neighbourhood of Vesuvius and Etna,
serves rather to support, than to invalidate this view. It
is further confirmed, in a very striking manner, by the phe-
nomenon that occurred in the eruption of Hecla in 1846,
when the lava current streamed forth from the deepest of the
four newly formed craters, spreading itself over the plain of
Thjorsa. In July, 1846, (only a few months subsequent to
the eruption of the volcano) when I was sojourning in that
district, the lower portion of the lava stream appeared studded
over with smoking fumeroles, in which so large a quantity
of beautifully crystallized muriate of ammonia was undergoing
a process of sublimation, that, notwithstanding the incessant
torrents of rain, hundreds of pounds of this valuable salt might
have been collected. On surveying the stream from the
summit of Hecla, it was easy to perceive that the formation
of muriate of ammonia was limited to the zone, in which
meadow lands were overflowed by the lava. Higher up, where
even the last traces of a stunted cryptogamic vegetation dis-
appear, the formation of this salt likewise ceased. The large
fumeroles of the back of the crater, and even of the four new
craters, yielded only sulphur, muriatic and sulphurous acids,
without exhibiting the slightest trace of ammoniacal products.
When we consider, that, according to Boussingault, an acre of
meadow land contains so much as 32 pounds of nitrogen, cor-
responding to about 122 pounds of muriate of ammonia, we
shall hardly be disposed to ascribe these nitrogenous products
of sublimation in the lava-currents to any other circumstance
than the vegetation which has been destroyed by the action of

OF ICELAND. 331

fire. The frequent occurrence in Southern Italy of tuff decom-
posed by acid vapours containing muriate of ammonia, likewise
confirms the hypothesis regarding the atmospheric origin of this
salt. For the same body of air which can annually convey to
a piece of meadow-land a quantity of ammonia corresponding
to these large nitrogenous contents, must at least be capable of
depositing an equal quantity of this alkali on tuff-beds saturated
by acid water; which may be actually observed in some rare
instances both in Southern Italy and Sicily.

Without entering more particularly into the question of the
origin of the acid gases of volcanoes, the object of this treatise
will be sufficiently attained, if we consider merely the remark-
able phenomena of decomposition exhibited by the substance of
palagonite under their action.

On treating pulverised palagonite with a large quantity of a
solution of sulphurous acid, it is dissolved in the cold, and forms
a fluid, coloured yellow-brown by salts of sesqui-oxide of iron.
On heating it, the sesqui-oxide of iron yields its oxygen to the
sulphurous acid. Sulphuric acid and protoxide of iron are pro-
duced in the proportion of one atom of the former to two atoms
of the latter. The oxidation of sulphurous acid is likewise
effected in part at the surface of the fumeroles, by the action
of the atmosphere, or at the subterranean depths, by the atmo-
spheric oxygen diffused in the spring-water. The sulphuric
acid thus generated, is diffused among the constituents of the
palagonite, which are liberated, together with a portion of
silicic acid, and appear as sulphates in solution. This process
represents the first stage of the fumerole action, which is
manifested in the Namer or Solfatara of Krisuvik and Rey-
kjatilid, on the grandest possible scale. Exhalations of sul-
phurous acid, sulphuretted hydrogen, sulphurous and aqueous
vapours, here burst in the wildest confusion from the hot soil,
consisting of palagonite-tuff, and spread themselves far over
the steaming sulphurous fields, that are constantly in the act
of being formed by the reciprocal decomposition of palagonite
and this gas. The erupted gases and vapours assume the
most different character on these plains, whose deceptive sul-
phurous and aluminous soil must be traversed with caution by
the traveller who would avoid the danger of being drowned

332

PSEUDO-VOLCANIC PHExVOMENA

in the hot mud. On the declivities of the mountains, where
firmer trap, or lava rocks oppose an obstacle to their further
advance, they burst foaming and hissing, in the form of vast
columns of vapour, from the fissures and clefts of the rocks,
giving rise to sounds like thunder when they strike the bottom
over subterranean cavities. But where the spring-system in-
clines more towards the valleys and the loose tuff soils, the
traveller continually meets with pools of boiling mud, in which
a horrible bluish-black argillaceous paste rises in huge bubbles,
which on bursting, often throw the boiling mud to a height of
upwards of fifteen feet, accumulating its masses in crater-like
ledges round the basin of the spring. These phenomena con-
stitute a picture of the wildest devastation, only to be surpassed
in horror by the dread waste of the dark rocky masses by
which the scene is enclosed.

If the activity of the chemical decompositions witnessed in
this remarkable theatre of volcanic activity, closed with the
formation of sulphuric acid and the solution of palagonite, to
which the latter gives rise, the relation existing among the
earthy bases found in the acid siliceous springs would neces-
sarily be the same as that existing in the constituents of palago-
nite itself; but experience shows that such is not the fact.

I here give the composition of a water, which I took, in
August, 1846, from one of the largest boiling mud-cauldrons of

Composition of Water, in Ten Thousand Parts.

Sulphate of lime....

1-2712

Sulphate of magnesia

1-0662

Sulphate of oxide of ammonium .... .... 0*7333

Sulphate of alumina

03261

Sulphate of soda....

02674

Sulphate of potash

01363

Silica

0-4171

Alumina*

0-0537

Sulphuretted hydrogen

0 0820

"Water ....

9995-6467

10000-0000
* These salts are all regarded here as anhydrous; the free alumina, the
small quantity of which brings it within the limits of the errors incidental to
the experiment, may have been dissolved in excess by the alum of the water.
The water contains, besides, a scarcely appreciable trace of protoxide of iron
muriatic acid and organic matter.

OF ICELAND. 333

the Reykjahlider solfatera, between the north-eastern declivity
of the Namarfjall and the Burfell lava stream.

If we reckon the bases of the salts, which scarcely amount
to ^q-q part of the water*, as 100, and compare the preceding
numbers with the relation of the bases in palagonite, we shall
discover a great difference.

Relation of Bases.

In In the Water of

Palaconite+. the Suffiones.

Oxide of iron 3675 O'OO

Alumina 25'50 1227

Lime 20*25 42 52

Magnesia 11'39 29'42

Soda 344 951

Potash 2-67 5*98

100-00 10000

A glance at this composition will suffice to show, that the
sulphate of iron, produced by the action of the sulphurous acid
on palagonite, does not occur in the water; that the alumina of
the water exists in a much smaller relation than that which cor-
responds with the composition of palagonite; that the gypsum
stands in a less considerable relation to the remaining bases of
the water than is required by the composition of palagonite;
and, finally, that the relation of the magnesia, soda, and potash,
taking into account the errors incidental to the experiment and
the oscillations to which these bases are subject as isomorphous
bodies, perfectly express the relation of these constituents in
palagonite.

These facts prove, in the most decided manner, that the
activity of the chemical decompositions induced by the sul-
phurous acid does not end in the solution of palagonite. It
still remains to determine how this activity is manifested in a
series of actions by which the quantity of protoxide of iron
formed, together with a portion of the alumina and lime, are
again removed from the solution. The separation of these
constituents depends on a very remarkable relation of the sub-

* We may thus see how inexact and exaggerated are the data yielded, even
by the most recent travellers, of the composition of tliese fumerole waters.
These waters have so little taste that they might, in case of necessity, be used
for drinking after thoy had been cooled and filtered.

t Liebig's Annals, bd. Ixi. II 3, s. 2/3.

334 PSEUDO-VOLCANIC PHENOMENA

stance of palagonite, which indicates a mode of origin of many
many of the older rocks that has hitherto remained wholly un-
heeded, and which may throw a new light on some metamorphic
structures, whose origin must otherwise be considered to belong
to the most mysterious problems of geology.

One mio-ht indeed be inclined to ascribe the total absence of
oxide of iron in the natural solutions of palagonite to its preci-
pitation by the free or carbonated alkalies which, under special
circumstances, proceed, as I am about to show, from the decom-
position of this mineral ; but an explanation of this nature appears
wholly inadmissible, as the alumina must have been precipitated
by alkalies before the protoxide of iron, or simultaneously with
it; facts that are alike incompatible with the presence of this
substance in most suffione waters. The true cause of the
phenomenon is different. Thus I have found that palagonite
possesses the property, on being digested in a neutral solution
of sulphate of iron, of precipitating the protoxide of iron, as a
hydrate or rather as a silicate, with the formation of sulphate
of lime. Free sulphurous acid, therefore, originally dissolves
the oxide of iron of the tuff as a proto-salt, together with a
portion of its remaining constituents; but when the solutions
have been neutralised by their passage through the rock, the
oxide of iron is again deposited as hydrated protoxide, on being
brought further in contact with the rock, or as hydrated sesqui-
oxide when oxygen is present. The decomposed palagonite is
thus converted into alternate and irregularly penetrating beds
of white ferruginous and of coloured ferruginous fumerole
clay. The limits of these deposits indicate the strata where
the first action of acid solutions has passed into the second one
of neutral solutions. The solfataras of Krisuvik and Eey-
kjahlid are rich in instructive examples of this nature. Thus,
for instance, we observe on the north-eastern wall of the
Namarfjall at Reykjahlid in the neighbourhood of the largest of
the cauldrons of boiling mud, a small ravine, formed by the
offshot of the waters and by the jets of steam issuing from the
fumeroles, in which these depositions are disclosed to a consi-
derable depth, exhibiting in the clearest manner the pheno-
menon of alternating colours. One is astonished at observing
the great similarity existing between the external phenomena of

OF ICELAND. 335

these metamorphic depositions of clay still in the act of forma-
tion, and certain structures of the keuper formation.

Thousands of years hence, the geologist, who explores
these regions, when the last traces of the now active fumeroles
have vanished, and the clay formations have become conso-
lidated into marl-like rocks by the silica with which they are
saturated, may suppose, from the differently stratified petro-
graphic and chemical character of these strata, that he was
looking at floetz strata formed by deposition from water, as in
the case of many structures of the secondary period. Not-
withstanding that this metamorphosed rock originally con-
sisted of a homogeneous mass of palagonite, the alternating
composition of whose interior is owing to a secondary transport
of its constituents.

The same action which is exercised by palagonite on the
neutral solutions of sulphate of protoxide of iron occurs also in
the sulphates of alumina and sesqui-oxide of iron. Both of these
bases owe to this their precipitation from neutral solutions,
with the formation of gypsum; so that not only is the alumina
removed from the suffione waters, but it is transported from
one part to another within the domain of these decomposi-
tions. A great variety and inequality must consequently be
induced in the composition of the argillaceous strata, affected
as it is by the incessant changes in the place of eruption of
gases, by the special relations of water-currents, and by the
saturations sroing on in the interior of the rock.

We see from these considerations, that sulphate of lime is a
main product of these reactions, appearing in the second as
well as the first stage of the decomposition of the palagonite.
Palagonite exercises no further influence upon this or any other
of the soluble products of decomposition. The inconsiderable
solubility of this salt combined with its great capacity for crys-
tallization gives rise to its constant separation under very re-
markable relations.

The clay of the fumeroles becomes filled up with deposi-
tions of this substance. At the surface, especially, where the
deposition is favoured by slow evaporation, innumerable crystals
of gypsum, often an inch in diameter, may frequently be ob-
served loosely surrounded by the argillaceous mass, as at the

336 PSEUDO-VOLCANIC PHENOMENA

solfataras on the Leirnukr. These crystals being very pure, but
not apparently very fully developed.

At the mountain ledge of the Namarfyall and at Krisuvik,
this gypsum is found to penetrate the argillaceous masses in
connected strata and floor-like depositions, which not unfre-
quently project as small rocks where the loose soil has been
carried away by the action of the water. These depositions
are sometimes sparry, corresponding in their exterior very
perfectly with the strata of gypsum so frequently met with
in the marl and clay formations of the trias.

Their deposition is owing to a fact that has not hitherto
been sufficiently regarded in the explanation of geological
phenomena, viz.: that substances crystallizing from solutions
are more readily deposited on a surface identical with their
own (although at a considerable distance from the limits of
their solubility), than on substances different from themselves.
These depositions of gypsum increase, therefore, in these
formations in the same manner as we observe small crystals
to enlarge in a solution, without any deposit being formed
on the sides of the vessel; much salt being removed from
the solution (not by a change of temperature, but owing
to the cohesive force emanating from the crystal,) so that
no further deposit can be made on the particles of bodies
of a different nature. The process of crystallization here
comes within the domain of mechanical forces, since it causes,
by the expansive growth of the layers of gypsum, the
upheaval of the moistened clay deposit, or compresses it
towards the exterior as the first-named masses increase in
quantity*.

* The attention of geologists has hitherto been almost exclusively directed
to the metamorphism of rocks from the action of fire. The metamorphic
transformations effected by the action of gas and water at low temperatures, as
we still see them exemplified on a small scale in the fumeroles, must, however,
have played a no less important part in the more ancient plutonic disturbances,
and exercised an immeasurable degree of influence in the formation of the sub-
stance, constituting the accumulated masses of the strata of the secondary
period. I have endeavoured, in the present treatise, to bring prominently
forward some indications and relations that may, perhaps, lead the geologist in
the right path for investigating these structures. Everything seems to indicate
that we are justified not merely from observations, but more from experimental
investigations, in referring the metamorphoses of rocks to hydatothermic and

OF ICELAND. 337

On a more attentive consideration of these phenomena, we
can scarcely avoid the conviction that the origin of a portion
of the vast deposits of gypsum, which so frequently characterise
the marly-argillaceous strata of the later floetz series, and in
which the total absence of calcareous conchylia points to the
action of acid vapours, is due to a chemically-identical, but
perhaps geologically-different, action. A careful investigation
of the varying composition of these marly deposits, having
special reference to their mutual relations to one another and
to the strata containing oxide of iron and deposits of gypsum,
would, therefore, prove of the greatest interest in studying
the history of such floetz-beds. I propose, on a future occasion,
to enter more fully into the relations of the tertiary clay forma-
tions to basaltic, doleritic, and trachytic tuff, since these fume-
role actions throw great light on the subject.

Before I quit the phenomena of springs and turn to those
embraced by the alkaline waters and the geyser-formations to
which they give rise, I must not omit to notice some few pro-
ducts which, although less prevalent, nevertheless belong to the
characteristic phenomena manifested by the activity of the sol-
fataras in Iceland. Among these, I reckon plumose alum, iron
pyrites, sulphide of copper, sulphate of copper, and, as the most
important of all, sulphur itself. The formation of plumose alum
is limited to the surface of the fumerole clay, such soluble salts
being only found at periods when a dry atmosphere favours
their efflorescence on the ground. Such depositions do not
depend on a spontaneous evaporation alone of the suffione
water, which is so poor in alumina, or even on the evaporation
induced by the heat of the volcanic soil, but must rather be
referred to the sulphurous gases that penetrate through the
hot soil; and they usually indicate the portion of the smoking
mud-crust which the traveller must tread with extreme caution,
in order to avoid the danger of being burnt. It may be

pyrocaustic — or where these occupy the same scene of action — to hydatocaustic
formations. I do not know, however, whether the time has yet arrived when
we may introduce these denominations into the nomenclature of science. Such
distinctive appellations certainly remain devoid of application until the test of
experiment has decided the question in all its bearings. And geological che-
mistry is, unfortunately, still far from having attained to this object.

338 PSEUDO-VOLCANIC PHENOMENA

plainly seen that the remarkable property possessed by pala-
gonite to precipitate neutral solutions of alumina, must limit
the formation of alum to those localities where a constant flow
of acid, more particularly sulphurous acid, is kept up by the
action of the fumeroles. Sulphuretted hydrogen gas, which
accompanies the gaseous exhalations of these suffiones, expe-
riences, where in contact with the porous fumerole clay, the
same kind of contact combustion which has lately been so
ably described by Dumas, at the expense of the oxygen of
the atmosphere; and of which the principal product is
sulphuric acid. Every fall of rain which dissolves the salt
and carries it to the inner strata of the palagonite, not
only withdraws it, but actually gives rise to a second process
of decomposition within the rock, in which the alumina is
precipitated by the palagonite, which, in its turn, passes
through all the different phases of decomposition that, as I
have already remarked, begin by the neutralization of the
solution and terminate in the precipitation of the alumina and
the oxide of iron. Hence it follows, that the phenomena of
the alum formations which, in dry weather, run through their
whole course in the space of a few days, and as rapidly dis-
appear in rain and a humid state of the atmosphere, bear no
relation to the small quantity of alumina contained in the
water of the suffiones.

The main attendant of the suffiones, and the one by which
they are actually converted into solfataras, is sulphur, which
is found deposited in very large quantities on the Namar of
Krisuvik, of which we have already frequently spoken; and
to a still more considerable extent in the neighbourhood of
the Krafla. It appears to be essentially due to the reciprocal
reaction of sulphurous acid and sulphuretted hydrogen gas.
At all events, these gases which, as is well known, mutually
decompose each other, depositing sulphur, constantly accom-
pany sulphur sublimations, exhibiting mutually antagonistic
conditions. It must be admitted that I have endeavoured un-
successfully to detect by the most sensitive test paper the pre-
sence of sulphuretted hydrogen in the fumeroles of the four
recent craters of Hecla, which are so rich in sulphurous acid
and sulphur. I was equally unsuccessful in discovering traces of

OF ICELAND. 339

sulphuretted hydrogen gas by the use of the same means, when
I investigated the gas that streamed from the fissures and
crevices of the innermost and highest of the craters of Hecla,
which have been upheaved during the late eruption. The inves-
tigation was rendered peculiarly difficult from the column
of gas that issued from the uppermost and innermost walls of
the crater. A phenomenon may, however, be noticed in the
immediate effects of the eruption, which indicates the presence
of some inconsiderable traces of this gas, which has escaped
undecomposed with the great excess of the liberated sulphurous
acid.

On bringing a lighted cigar near the evolving gas the
well-known phenomenon of contact-combustion is at once
manifested, which was first observed at the fumeroles of the
solfataras of Naples, and may be considered as exercising the
most sensitive reaction on sulphuretted hydrogen gas*. A
dense cloud of vapour is instantaneously observed to emanate
from the burning body, diffusing itself far over the ravine and
the plain of the fumeroles. This phenomenon could be called
forth at Hecla even where there was no smell of sulphurous
acid, and where paper, saturated in acetate of lead, exhibited
no trace of sulphuretted hydrogen after having been exposed
for many hours to the action of the vapours. It is, more-
over, a very common occurrence in Iceland to find that the
water of the boiling cauldrons, although giving indications
of sulphurous acid only, is permeated by gases impregnated
by a large quantity of sulphuretted hydrogen — a circum-
stance which proves that the last gas is decomposed on
its solution in the fluid containing sulphurous acid, and that
this decomposition is accompanied by the deposition of sul-
phur. Where these gases appear in contact with aqueous
vapour, thick crystalline crusts of sulphur may be observed
deposited round the mouths and over the top of the fumeroles.

* Almost all the hot springs and fumeroles of Iceland exhibit this phe-
nomenon. Even the light vapour which rises from the clear bluish-green
water of the basin of the large crater is converted, on the approximation of a
burning body, into a dense cloud of vapour, which envelopes the whole mirror
of the water and spreads itself in all directions, from the point of contact, as if
from a centre of combustion.

340 PSEUDO-VOLCANIC PHENOMENA

Their deposition is apparently owing to a mechanical trans-
portation of the flowers of sulphur by the vapour escaping,
and this formation may not inaptly be compared to the depo-
sitions of dust and soot, which, when finely disintegrated, are
mechanically borne to great distances by the currents of the
atmosphere.

Another and far less considerable part of the sulphur is
deposited in the form of a delicate white powder, by which
the argillaceous strata seem not unfrequently to be cemented
in a manner. This deposition appears to be owing to the
decomposition of the sulphuretted hydrogen by atmospheric
oxygen, or it may also depend on vapour of sulphur that
accompanies the stream of aqueous vapour. This latter kind
of sulphurous deposition can be exhibited by a very simple
experiment, for when water containing flowers of sulphur is
boiled, a considerable quantity of this body may be observed
to pass into the receiver in the form of a delicate white
powder. I have only found sulphur in a state of fusion in
the two largest of the recent craters of Hecla, the upper walls
of which manifested, at different parts of their superficies, a
temperature far exceeding 1 00° C, this being especially the
case in the vicinity of the many fissures that run in a north-
eastern direction parallel with the back of the crater. The
observation was made in July 1846, and, consequently, some
months after the last great eruption.

Another product of fumerole activity deserving, in a high
degree, the attention of geologists, is iron pyrites, which very
frequently fills up different portions of the clay deposits pro-
duced by the decomposition of palagonite, appearing in the
form of small crystals, often very beautifully developed. Its
formative process, which can be most plainly observed, throws
considerable light on the formation of the iron pyrites, which
occurs so frequently in the older argillaceous marl beds, espe-
cially in the trias, for although the two processes may originate
in very different geological conditions, there can be no doubt
that they are of the same chemical nature.

These formations really depend upon a very remarkable
decomposition experienced by the substance of the palagonite
under the action of sulphuretted hydrogen. In consequence

OF ICELAND. 341

of this, proto-sulphide of iron and alkaline metallic sulphides
are produced. The palagonite is converted by the former into
a black mass, which occasionally imparts a bluish-black colour
to the clay of the boiling mud-pools, and is not unfrequently
manifested in the spring by the deposition of a sandy-black
powder. The alkaline metallic sulphides, on the contrary,
are dissolved by the boiling water, and converted, wherever
they come in contact with sulphur, into poly-sulphides. No
one who is familiar with analytic investigations can be igno-
rant of the facility with which these bodies dissolve slight
traces of proto-sulphide of iron with a green tint, and then
again are able to deposit it under peculiar circumstances.
It will, therefore, be easily understood how the higher oxide
of iron, converted into proto-sulphide of iron, by sulphuretted
hydrogen, with the separation of sulphur, should be dissolved
by the simultaneously formed alkaline poly-sulphides; and
abstracting from the latter an atom of sulphur should again
be precipitated in crystals of bisulphide of iron, or iron
pyrites. The mode in which the iron pyrites is formed fully
confirms this view. The coloration by oxide of iron stands
in an inverse ratio to the quantity of the crystals formed,
the former increasing in depth as the latter diminish in
quantity, the pyrites being only found in its greatest possible
development where the oxide of the iron had wholly vanished
from the clay.

I shall postpone for another occasion the consideration of
the conclusions which might be drawn from these processes
regarding the origin of the older clay-formations. It is easy
to perceive that these pyrites-formations have nothing in
common with those that have been produced by the decom-
position of sulphates by the action of organic matter, since
they occur as well in the highest crater of Mount Hecla, where
all idea of the co-operation of organic substances must fall to
the ground, as in the geysers of Reykir and in the solfataras
of Krisuvik and Reykjahlid.

I must not omit to mention, that some compounds of copper,
arising from secondary actions, likewise appear as separate pro-
ducts of fumerole activity, as, for instance, sulphide of copper,
krisuvigite, and sulphate of copper. The formation of these

342 PSEUDO-VOLCANIC PHENOMENA

substances is, however, too simple to require any separate
notice; although we would observe that the diffusion of copper
in the great Icelandic volcanic system seems to receive some
elucidation from the occurrence of pure copper in the trap
formations of the island of Faroe.

On considering the numerous chemical processes, whose
foci are the solfataras and fumeroles of Iceland, we cannot
fail to see that it is the great abundance of volcanic gases,
especially sulphurous acid, which, together with the reaction of
palagonite, constitute the main character of these phenomena.
Where these gases are no longer prominently manifested, or
where sulphurous acid is almost wholly absent, the scene sud-
denly changes. The observer finds himself at once transported
to a totally different field of pseudo-volcanic phenomena, as it is
represented by the innumerable thermal and geyser systems. The
connection existing among these phenomena and those we have
been considering, is not less simple than it is easy of compre-
hension. Here, too, the relation of the palagonite substance to
the composition of the waters of the springs constitutes the
starting point, from whence the observer, aided by experiments,
may advance, step by step, from the most inconsiderably mani-
fested chemical actions, till he arrives at a knowledge of that
wonderful mechanism by which the grand activity of the violent
sources of eruption in Iceland is maintained. As it will be
necessary to direct our attention to some definite locality, I
would select the Great Geyser, as the best known of all the
intermitting eruptive springs of Iceland.

The thermal group belonging to this celebrated spring lies
on the outskirts of the great glacier plain which constitutes the
elevated plateau in the centre of the island, almost exactly
south-west of the highest point of Hecla, and only distant
from it about 20 geographical miles in a direct line. The
height of the geysers over Reykjavik (at the residence of
Counsellor of Justice, Thorstensan,) amounted, according to
the barometrical measures made on two consecutive days, to
112*8 met. and 107*2 met. Their main direction runs about
N. 17° E. and is therefore almost parallel with the chain of
Hecla and with the general direction of the fissures. The
oldest rock forming the base of the springs is also here com-

OF ICELAND. 343

posed of palagonite tuff, penetrated lengthwise by a vein of
clinkstone, running from the western margin of the springs.
Here and there a few boiling and vapour springs burst from
the clinkstone, at a height of about 55 met. above the Great
Geyser. The main focus of thermal activity is, however,
situated in a loose palagonite tuff at the foot of the opening in
the clinkstone. This rock, on the north-western side of the
geyser-cone, where the strata are broken through by a jet of
water, is covered above by the siliceous deposits from the
spring, whilst below it becomes transformed into the variegated
fumerole clay, of which we have already spoken as a product of
the decomposition of palagonite.

The characteristic phenomena of decomposition which mark
the acid siliceous waters, are never wholly absent from the
class of alkaline springs. Here, too, where the fumerole clay
is freed from siliceous incrustations, a foaming pool may occa-
sionally be met with, the dark viscid mud of which rises in
huge bubbles, or a steaming bed covered with crystals of
gypsum and alum, or, finally, a deposit of sulphur superposed
on the clay, or even on the siliceous incrustations. But these
phenomena, which depend on the occurrence of small quantities
of sulphurous acid, sink into insignificance, or I might almost
say, entirely vanish before the stupendous phenomena developed
by the action of carbonic acid, sulphuretted hydrogen, and
heated water, on the substance of the palagonite. In the
mutual reaction of these four substances are combined all the
conditions required by nature to convert, in the course of cen-
turies, simple boiling springs into geysers, whose clear, vapoury,
and foaming columns of waters, shall burst from the summits of
their self-created siliceous tuff-craters, either continuously, or at
periods extending from a few minutes to hours, or days. These
geysers and the other alkaline siliceous springs of Iceland, do not
evince that dreary character of wild devastation which is mani-
fested by the fumeroles and solfataras, with their boiling mud-
pools and their steaming fields of sulphur. The eye rests
rather with pleasure on the white ledges and stalactites, which
rise now in the form of small conical craters, then in long
furrows and vast basins, and again in round openings of the
most regular configuration.

344 PSEUDO-VOLCANIC PHENOMENA

It will be necessary here to pause a few moments for the con-
sideration of these incrustations. Their structure is simple and
easy of explanation. It will be seen from the composition of
the water of the geysers, as given in a subsequent part of this
paper, that the silica is dissolved in the water by alkaline car-
bonates and in the form of a hydrate. No trace of silica is pre-
cipitated on the cooling of the water, and it is only after the
evaporation of the latter that silica is deposited in the form of
a thin film on the moistened sides of the vessel where evapora-
tion to dryness takes place, whilst the fluid itself is not rendered
turbid by hydrated silica until the process of concentration is
far advanced. This apparently trivial circumstance is of the
greatest importance in the formation of the geysers. It will be
evident that the basin of the spring, in which the constantly
renewed water affords only a very small field for the process
of evaporation, must remain free from siliceous formations,
whilst the margins projecting beyond the level of the water,
will readily become covered by a siliceous incrustation owing
to the rapid and easy drying of the moisture attracted by a
capillary force. Further on, where the water spreads itself
over the surface surrounding the spring, the incrustations in-
crease in proportion as the surface of evaporation expands. As
the basin of the spring has no part in this incrustation, it
becomes converted into a deep tube as it is gradually enclosed
by a hillock of siliceous tuff, combining, when it has reached
a certain height, all the requirements necessary to convert it
into a geyser. If such a tube be narrow and be filled with
tolerable rapidity by a column of water strongly heated from
below by the volcanic soil, a continuous geyser must neces-
sarily be produced, as we find them in so many parts of
Iceland. For it will easily be understood that a spring, which
originally did not possess a higher temperature at its mouth
than that which would correspond to the pressure of the atmo-
sphere, may easily, when it has been surmounted by a tube,
formed by gradual incrustation, attain at its base a temperature
of upwards of 100° C, owing to the pressure of the fluid
resting in the tube. The mass of water rising in such a spring,
which is constantly renewed from below, and possessing in the
natural shaft of the spring a temperature of 100°, must imme-

OF ICELAND. 345

diately, on its escape from the mouth of the tube, experience a
diminution of temperature corresponding to the diminished
pressure of the atmosphere, by which the whole excess of heat
above 100° will be expended in the formation of vapour. The
water raised by the expansive force of these vapours, and
mixed in a white foam, forces itself from the mouth of the
spring foaming and hissing in one continuous gush. Iceland
abounds in springs of this kind; although the Reykholter valley
presents, perhaps, the largest number to be found in any one
6pot. I shall not, however, enter into any more circumstantial
description of them, since the subject is only one of secondary
interest. When the geyser tube, formed by the process of
incrustation, is sufficiently wide to afford a considerable degree
of cooling to the water at its surface, and the jet which is
heated upwards of 100° C, falls but slowly back to the base
of the wide funnel, we find all the requirements necessary for
converting the spring into a periodically recurring geyser,
which bursts suddenly forth by the action of the developed
force of the vapour, and then forthwith falls back to a state of
long continued repose. The Great Geyser is the most remark-
able of these springs, which have been regarded as natural
accumulators of the force of vapour.

Immediately after an eruption, the water which fills the
tube to the height of 1 or 2 met., gradually rises during several
hours to the margin of the basin, whence it flows over the cone
in the form of a small cascade.

It may easily be shown by experiment, that the column of
fluid filling the tube, is constantly being heated by water
entering it from below, whilst it experiences a constant cooling
and evaporation above on the broad surface of the basin.
Evaporation is likewise effected within the funnel itself by
means of a current which rises and falls in its upper portion;
driving a column of heated water up the centre of the funnel,
it spreads itself over the surface of the basin towards the edges,
and after the evaporation has been completed flows back to the
funnel.

The direction of this current may be observed by throwing
a few shreds of paper into the middle of the basin of the
geyser, as they will be driven to the margin of the upper

346 PSEUDO-VOLCANIC PHENOMENA

surface, and then again be carried back to the bottom of the
tube.

The changes of temperature experienced by the different
strata of the column of water when subjected to the alter-
nately cooling and heating influence during the time interven-
ing between the eruptions, have been made the subject of a
series of thermometrical measurements, conducted conjointly
by M. Descloizeaux and myself at the Great Geyser. The
following is a portion of the results obtained*: —

6th o

f July.

8h 20m p. m.

Height from
the Bottom.

Temperature.

0*3 met.

....

.... ...

123°-6 C.

4-8 „

....

122°-7 „

96 „

....

113°-0 „

144 „

....

.... ....

85°'8 „

192 „

....

82°-6 „

7th of July.

2h 55m p. m.

03 „

....

.... ....

127°'5 „

50 „

....

.... ....

123°0 „

9-85 „

....

.... ....

120°'4 „

1475 „

....

106°-4 „

1955 „

....

85°'2 „

7th of July.

7h 58™ p. m.

0-3 „

....

.... ....

126°' 5 „

5-0 „

....

985 „

....

.... ....

121°'8 „

1475 „

...

....

110°0 „

19-55 „

....

84°-7 „

From these results it follows:

1. That the temperature of the column of the geyser
decreases from below upwards, as had already been shewn by
Lottin and Robert.

2. That, setting aside small disturbances, the temperature
goes on increasing regularly at all points of the column from
the time of the last eruption.

3. That the temperature in the unmoved column of water

* A full report of these observations is given by M. Descloizeaux, in the
Annates de Chim. el de Phys., 3 se'rie, t. xix.

OF ICELAND. 347

did not, at any period of time up to a few minutes before the
great eruption, reach the boiling-point that corresponds to the
atmospheric and aqueous pressure at the point of observation.

4. That it is at mid-height in the funnel of the geyser where
the temperature approaches nearest to the boiling-point corres-
ponding to the pressure of the column of water, and that it
approaches nearer to this point in proportion to the approxima-
tion of the period of a great eruption.

The curves of Plate II., fig. 1, are a graphical exhibition of
these relations. Curve 1 represents the temperatures at which
the column of water would be brought into a state of ebullition
throughout its whole length, the irregular lines 2, 3, and 4,
the temperatures observed of the column of water at the
respective periods of 10 minutes, of 5 hours, 31 minutes, and
of 23 hours, 13 minutes, before a great eruption.

If we now consider the period which immediately precedes
an eruption, we shall find that only a very slight impulse
is necessary to bring a large portion of the column of water
suddenly into a state of ebullition, and, as we shall soon see,
even to produce an eruption. Every cause that tends to raise
this column of water only a few meters, must necessarily be
attended by this result. If, for instance, we assume this
elevation to be equal to 2m, the column of fluid pressing on the
point a (fig. 2) will be shortened by the height ab. The
temperature a of the stratum of water lying under a pressure
diminished by ab is now about b c, or 1° higher than the
corresponding boiling-point of the water. This excess of 1° is
immediately expended in the formation of vapour, generating
in the present case, as may be proved by an easy calculation,
a stratum of vapour nearly equally high with the stratum of
water lm in height. By this diminution in the superincumbent
water a new and deeper portion of the column of water is raised
above the boiling-point; a new formation of vapour then takes
place, which again occasions a shortening in the pressing liquid
strata, and so on, until the boiling has descended from the
middle to near the bottom of the funnel of the geyser, provided
always that no other circumstances have more speedily put an
end to this process.

It appears from these considerations that the column of

348 PSEUDO-VOLCANIC PHENOMENA

water in the funnel of the geyser extending to a certain dis-
tance below the middle, is suddenly brought into a state of
ebullition, and further, as may be shown by an easy method of
computation, that the mechanical force developed by this sud-
denly established process of vaporization is more than sufficient
to raise the huge mass of the waters of the geyser to that
astounding elevation which imparts so grand and imposing a
character to these beautiful phenomena of eruption. The
amount of this force may easily be ascertained by calculating
from the temperatures of the preceding experiments, and the
known latent and specific heat of the aqueous vapour, the height
of the column of vapour, which would be developed by the
ascent to the mouth of the geyser of a section of the column
of water. If we designate the height of such a column of water
in the funnel of the geyser by h; its mean temperature
expressed in centesimal degrees by t\ the latent heat of the
aqueous vapour by w; the density of the latter compared with
that of the water by s; and the co-efficient of expansion of the
vapour by d; we shall find that the excess of heat of the water
above the boiling-point under the pressure of one atmosphere
is £-100. But the height, h, of the section of the column
of water, which at the mouth of the geyser, that is to say,
under the pressure of one atmosphere, would be converted into
vapour by the quantity of heat, £-100, would be to the whole
height of the water column, h, as (£—100): w. A column of

water of the height — ^ ^ would therefore be evaporated

at the mean temperature t, if the water were under the
pressure of one atmosphere. Hence, it directly follows that
the height, H, of the column of vapour sought at 100° and
0m 76, will be

H-- ^(^-•100)(1 + 100^)
w s

On applying this formula to the value of the numbers found
by observation, we obtain the remarkable result that, in the
period of time immediately preceding an eruption, a column
of water only 12m in length, which projects from 5m to 17m
above the base of the tube, generates, for the diagonal section
of the geyser, a column of vapour 638m 8 in height (assumed

OF ICELAND. 349

to be at 100°, and under the pressure of one atmosphere).
This column being developed continuously from the upheaved
mass of water, as the lower strata reach the mouth of the
geyser. The whole column of the geyser, reckoned from the
point where its temperature amounts to 100° C. down to the
base, is capable, according to a calculation of this kind, of
generating a similar column of vapour, 104 lm in height*.
We can easily understand the reason that this enormous force
should not be expended in one single jet of eruption, when
we remember that the jets of water erupted in the air are
continually falling back into the tube of the geyser, and
interrupting, at different moments, the force of the upheav-
ing column of vapour, which is condensed in the cooled water
as it falls, until the temperature of the latter again reaches
the boiling-point, and has consequently regained the power of
being again propelled upwards. At the same time the water
may be seen flowing from the basin back into the funnel,
between the different separate ascents of the water. Occa-
sionally, the water even appears as if it were forcibly drawn
back.

The condensation in question, with the consequent restora-
tion of heat to the water of the geyser, explains, at the same
time, the fact of the great eruptions continuing frequently for
a period of more than five minutes.

We now see the causes to which the column of water owes the
slight elevation which imparts the first impulse to the eruption.
The greater portion of the Icelandic thermal springs exhibit
the peculiar but easily explained property of giving rise
periodically, at certain points in the water of the thermal
basins, to a number of large bubbles of vapour, which become
suddenly condensed on rising into an upper and cooler stratum.
This invariably occasions a slight detonation, accompanied by
a hemispherical elevation, and an instantaneously succeeding
depression of the surface of the water. Even the Great Geyser
is characterised by a periodic succession of these detonations

* I have assumed, for the sake of greater simplicity, that the temperature
of the water in the tube of the geyser does not increase in a uniform curve, but
in broken lines. On the first assumption the calculation would, of course,
yield a somewhat higher column of vapour.

350 PSEUDO-VOLCANIC PHENOMENA

of vapour, beginning about four or five hours after a great
eruption, and continuing, at intervals of from one to two hours,
until the next eruption; immediately preceding which they
occur in rapid succession and extreme violence. The cause
of the periodicity of these detonations admits of an easy ex-
planation; for it will be readily understood that when a
stratum of water under the continuous influence of the heat
of the volcanic soil, is brought into ebullition in the con-
ducting channels of the tube of the geyser, and the vapour
formed by its ascent into the higher and cooler masses of water
is again condensed, the temperature of this boiling stratum is
so far lowered by the formation of vapour, that it requires
some length of time before it can be heated again to the
boiling-point after the condensation of the vapours that have
ascended in the water. The periodic elevation of the masses
of water effected in the geyser by these means appears rarely
to exceed 1 or 2 metres, if we may judge by the average height
of the mass of water which is forced from the mouth of the
tube in the form of a conical elevation. A glance at the
graphic representation given in fig. 2 will, however, show
that such an upheaval as this is insufficient to raise any
stratum of water to an elevation where it could be brought
to a state of ebullition, (in consequence of the diminution of
pressure existing there,) until the mass of water would, by
gradual heating, be brought to the temperature of the broken
line, 2; and such is actually observed to be the case a few
minutes before an eruption. All the other upheavals preceding
this period would only be able to drive partially the lower
heated masses of water by a sudden impulse into the upper
part of the tube of the geyser, where these masses are brought
into a state of ebullition, owing to the diminution of the
pressure. These smaller eruptions may, therefore, be regarded
as abortive beginnings of the larger, being unable, to propagate
themselves beyond a short distance from the point of origin of
the vapour, from the low temperature of the column of water.

On considering all the phenomena presented by the erup-
tions of the geysers, we cannot for a moment doubt that the
main seat of the mechanical force, by which the mass of water
is thrown up and converted into boiling foam, is actually

OF ICELAND. 351

situated in the funnel of the geyser. This view is very beau-
tifully confirmed by observing the manner in which the water
is set into motion during an eruption.

In order to obtain a clear knowledge of these internal pro-
cesses, I have frequently sunk stones weighing several hundred
grammes, and marked by different bands, suspending them by
thin threads at different depths in the geyser funnels filled
with water. The only stones that were thrown from the
spring to heights of upwards of 100 feet proved to be those
suspended near the surface, whilst those that had been sunk
to greater depths never exhibited themselves. Stones weighing
many kilogrammes, placed between the basin of the geyser
were, however, carried back into the tube of the geyser with
the mass of water returning from the basin in the intervals
between the separate jets, and these were again erupted. This
fact, which may appear strange at first sight, is in perfect
harmony with the formation of the jets of eruption within the
tube. The mixture of vapour and water, of which these spouts
or jets consist, must necessarily move with constantly increasing
rapidity as the expansion and development of the vapour increase
at the mouth of the geyser, so that the moving fluid may thus
be able to carry with it, near the mouth of the basin, heavy
objects which it would be unable to upheave at any greater
depths. Such experiments clearly show, that we must regard
the tube of the spring as the actual focus of the mecha-
nical forces that maintain the action of these periodical erup-
tions, and further, that whatever communication may exist
by means of lateral channels between this main thermal stratum
and the heated masses of water within the ground, the water
must still be able to develope considerable masses of vapour
and convey them to the apparatus of eruption, owing to the
diminution of pressure consequent on this phenomenon. A
necessary result of these circumstances is, that the eruptions
gain in force, in proportion as they lose in regularity of their
intermission and duration. The co-operation of this secondary
development of vapour is shown by the remarkable fact, that
the spouts of water move in a rotatory direction during great
eruptions; a phenomenon that can only be explained by a
lateral influx of vapour. The rotatory motion cannot certainly

352 PSEUDO-VOLCANIC PHENOMENA

be observed through the dense clouds of vapour that envelop
the erupted masses of water, but I accidentally discovered it
in an experiment originally instituted with a view of measuring
the pressure at the base of the geyser column during a great
eruption, by means of a small maximum manometer*, which
was suspended by a line in the middle of the spring, and sunk
very nearly to the bottom. A number of stones, which I sunk
suspended to fine threads, that were secured at various parts
of the circumference of the funnel, separated from the threads
without being again erupted, however, with the water. The
threads themselves, which were suspended round the circum-
ference of the tube of the geyser, and parallel with the line at
a distance of 1*5 metres, were whirled into an inextricable coil
round the line of the manometer, which could only arise from
a gyratory motion in the column of water. This also is un-
doubtedly the cause of that tangential expansion of the water-
spouts, by which they are frequently spread far beyond the
margin of the tuff-basin, imparting a form to the eruption
somewhat similar to the species of fire-work termed a bouquet
de feu.

On considering the conditions of the activity of these
eruptive springs, we shall scarcely be led to expect any special
regularity in the magnitude, duration, and succession of the
eruptions and the detonations of vapourf . The supply of water
to the springs, which varies with the state of weather, and
the evaporation at the surface, which depends on the tempera-
ture and intensity of the currents of air, must both be most
intimately associated with the changing play of these phe-
nomena:]:. If the activity of the eruptions depends, as is sup-
posed by those who live in the neighbourhood of the geysers,

* This experiment proved unsuccessful in measuring the amount of pressure,
although it enabled us to determine the temperature at the base of the geyser
during the eruption. At Strokkr I was, however, successful in carrying out
this experiment, but omit any detailed account of it, owing to the minute
description of the apparatus employed, which would be necessary.

t I omit all further account of this subject, as M. Descloizeaux has given the
numerical results of our combined observations on the periodicity of these phe-
nomena during our sojourn at the geyser from the 3rd to the 14th of July.

t The temperature of the water at the surface of the basin of the geyser
varied, during the time of the observations, from 76° C to 89° C.

OF ICELAND. 353

on meteorological influences, an investigation into these con-
ditions, although they are not as yet established as facts, and
are unfortunately not within the scope of the researches of the
passing traveller, would undoubtedly tend to confirm the theory
whose principles I have attempted to develop.

If we compare this theory with the old hypothesis of the
geysers, we can scarcely comprehend how the latter could
maintain so long a place in the domain of science, since it is
impossible to disregard the evidence contradicting it, furnished
by the facts observed at every eruption. The idea involved
in this hypothesis of subterranean cauldrons of vapour, sup-
posed to be alternately filled with vapour and with water, is
totally irreconcilable with the simple observation, that the
masses of water propelled during the eruption beyond the
margin of the basin, correspond perfectly with the depression
of level that immediately follows, and, consequently, the sup-
posed retreat of the water into an imaginary subterranean caul-
dron of vapour can have no reality. It would seem almost
superfluous to mention another circumstance which shows the
utter fallacy of this hypothesis, if it were not to prove more
fully, that the lower part of the geyser frequently takes no part
in the violent commotion manifested during an eruption at its
upper portion. I have succeeded, during an eruption that
attained a height of upwards of 43*3 metres, in keeping a
thermometrograph uninjured in the lower part of the geyser
tube by a manometer, and observing by this means, imme-
diately before the eruption, a temperature lower by 9° C. than
that which would correspond with the boiling-point at this
spot. An observation that would involve an absolute absurd-
ity, if, in accordance with the old hypothesis, we assume the
eruption to be occasioned by boiling water being forced into
the tube of the geyser from the depths below it.

The Strukkr, the greatest source of eruption in Iceland,
next to the Geyser, is only a few hundred paces distant from it.
The form of this spring presents peculiarities which must exer-
cise a special influence on its mechanical results. Fig. 3,
Plate II., gives a section of the tube of this spring as observed
by the soundings instituted by M. Descloizeaux and myself.
The tube is only 13m55 in depth, and not cylindrical as in the

2 A

354 TSEUDO-VOLCANIC PHENOMENA

case of the Great Geyser, but rather funnel-shaped, so that
the diameter at the mouth is 2m4, whilst it is only 0m26 at
a depth of 8m3. The water, which rises to about 3m, or 4m5
from the mouth, has no means of escape, and is only expelled
by the action of the eruption. As the column of water acces-
sible to the sounding line is in a constant state of violent
ebullition, the different temperatures must remain constant,
and correspond with the pressure upon each stratum. In the
following measurements, which were made by M. Descloizeaux
and myself with every care and precaution, the time intervening
since the last eruption has not been taken into account.

Height from the Bottom.

0"00 meters

0 35

295

4-65

620

8-80

Height of Water-column

Temperature.

corresponding with
Pressure observed.

1150° C

(Eru

ption)

113-9°,,

1142° „

15-1

H3-7°„

156

109-3° „

13-6

99-0° „

10-2

From these experiments it follows: —

1. That the boiling-point of the strata of water in the
upper and broader portion of the tube of the Strokkr corre-
sponds precisely, as is required by the theory already developed,
with the pressure acting on it.

2. That in the lower and narrower part of the tube, from
the bottom to a height of 4m6, the temperature remains the
same (excepting very small oscillations).

3. That the temperature, amounting to 114° C., which
remains invariable over a space of about 4m6, corresponds to
the boiling-point of a mass of water under a pressure of a
column of water 16m5 in height.

4. That, if we assume the lower and narrower part of the
the channel to be filled with vapour to a height of 4m65, as
was the case at the time these measurements of the pressure
were made, the weight of water acting on this vapour will
actually amount to about 16m5, or, at any rate, 15m5.

5. That during an eruption ranging to a height of 48*7 met.
the temperature at the bottom of the narrow canal rose to
115° C.

OF ICELAND. 355

We cannot, therefore, doubt, that the lower and narrower
part of the funnel of the Strokkr is filled by a column of
steam, which regulates the uniform temperature of this part
at different elevations, whilst the water, upheaved into the
upper part of the funnel by the force of this vapour, is con-
stantly kept in a state of ebullition.

The seat of the activity by which the great periodic erup-
tions are regulated, must be situated at depths inaccessible to
direct experiment. This may easily be conjectured from the
fact, that if the passage for the vapour be filled up with stones
and earth, an eruption takes place from 20 to 30 minutes after-
wards, by which all the impediments to the passage are
expelled, accompanied by jets of muddy water, and when these
have been wholly erupted, spouts of clear water are ejected,
which not unfrequently measure upwards of 57 met. in height.
It will be easily understood, that these eruptions breaking forth
from that part of the steam channel which is alone accessible
to our measurements, might be brought into a state of periodical
activity, by a process similar to the mechanism of the Geyser.

Besides these periodic jets, whose prototype must be sought
in the Great Geyser, and in the Strokkr Geyser, there is
another species of intermittent thermal springs, whose activity
is not manifested by the sudden occurrence of subterranean
detonations of vapour, and whose eruptions are not charac-
terised by sudden and rapid expulsions of boiling water. The
Litli Geyser (Little Geyser) which belongs to the thermal group
of Reykir, furnishes one of the most remarkable examples of
the species. This spring rises in a palagonite rock, and is the
highest but one toward the north-eastern mountain wall, at the
foot of which rest the extensive siliceous tuff deposits of
Reykir. A conical tuff elevation is here observed, whose small
thermal crater is enclosed by stones. The boiling spout shoots
forth periodically from among these stones, which are frequently
not covered with the water, and exhibit, moreover, a very incon-
siderable development of aqueous vapour. The eruptions were
repeated very regularly at intervals of 3h 45m on the 24th
and 25th of June, 1846, when I had the opportunity of
observing the spring; the main eruption which occurred
between 9 and 10 each morning, far surpassed the others in

2 A 2

356 PSEUDO-VOLCANIC PHENOMENA

magnitude and beauty. The approach of an eruption is made
known by a gradual increase in the development of gas, and by
a subterranean splashing sound. Boiling foam bursts forth
with the vapours, and continues to rise and fall at slow inter-
vals, until at length, after about 10 minutes, when the eruption
has attained its maximum intensity, it rises in vertically and
laterally spouting jets to a height of 30 or 40 feet. The water-
spouts then decrease in height and circumference with the same
gradations that marked their development, until the spring,
at the end of about 10 minutes, returns to its former repose.
This phenomenon is certainly inferior in magnitude to that of
the eruptions of the Great Geyser in which a jet of boiling
water, upwards of 28 feet in circumference, and 100 feet in
height, sends its far projecting point of foam against the clear
sky — but in beauty it is scarcely inferior to its colossal rival.
The deafening hissing and roaring that accompany the eruption
of the jets of water from the crater of the springs, in which one
may distinctly hear the rushing of the masses of water, which
fall down in torrents and are beaten to foam by the force of the
vapours — the dazzling rainbows, which are formed with varying
brilliancy from the reflection of the solar rays amid the pearly
drops of the rushing cascades, and as rapidly disappear beneath
the overwhelming mass of vapours — the dense and spherical
vapoury clouds, which rising from the waterspouts, are made
the sport of the winds as they stand forth in bold relief from
the dark wall of rock behind them — the faint halo round the
head of the colossal shadow which the traveller sees flitting
across the clouds, and which, invisible to his companions, is
seen by him alone. All these combine to excite in the mind of
the traveller an indescribable impression of sublime grandeur.

We shall readily perceive that this spring cannot be main-
tained in a state of activity by the same causes that give rise to
the eruption of the Great Geyser, if we consider the duration of
its eruptions, their slow but regular increase and decrease, and
the great regularity of their periodicity. All these phenomena
appear, on the contrary, to be in perfect harmony with the
hypothesis first advanced by Mackenzie, and since more widely
extended by many others, regarding the existence of a sub-
terranean cauldron or focus of vapour, which, as I have endea-

OF ICELAND. 357

voured, I hope successfully to show, has been very incorrectly
used in explanation of the eruptions of the Great Geyser.
This hypothesis would appear to be applicable to the Little
Geyser and the great number of similar springs found in Ice-
land. As, however, it would be foreign to the object of the
present memoir to enter into the consideration of views, which
it must be assumed are already known to the reader, and the
importance of which will not be allowed until they can be
proved by direct experiment, I will turn at once to another
form of siliceous springs, which designate, as it were, the last
stage of these formations.

I have already shown how the incrusting thermal springs of
high temperature provide themselves in the course of centuries
with a geyser apparatus, and thus necessarily pass into the con-
dition of a continuous and intermittent source of eruption.
The history of the development of these phenomena is not,
however, terminated by this occurrence. The formations of
siliceous tuff advance continually, until the apparatus of the
geyser and the surrounding ground attain a height that must
put an end to the eruptive activity of the spring, owing to the
change effected in the relation of the height of the column of
water and the heat evolved from the ground. As soon as the
supply of heat from below, and the cooling at the surface, are so
far in equilibrium, that the temperature of the mass of water is
not anywhere able to reach the boiling-point, the action of the
spring ceases spontaneously. Large reservoirs of tuff filled
with hot stagnating or running water are then formed, whose
depth and configuration depend upon the accidental condition
of the margins, stalactitic formations, and the sinking in of
the surface ; all of which again depend to a great extent on the
masses of palagonite that are continually brought to the surface
in the form of soluble salts and silica by the water from sub-
terranean depths.

Where Buch springs broke forth from deep points, they
entirely disappear from their original positions, owing to the
constant increase of the hydrostatic pressure to which they are
subject, or only exhibit traces of their former activity in the
disintegrated cavities of their vast siliceous tuff deposits, in
which thermal activity is dearly, or wholly extinct. Among

358 PSEUDO-VOLCANIC PHENOMENA

the many instances that might be advanced in illustration of
such changes, I will only mention the district of the Great
Geyser,

The vast deposits of siliceous tuff with scarcely a trace of
hot springs, which skirt the heights of the Bjarnafell, indi-
cate an earlier condition of this stupendous thermal activity,
that must have passed through the last period of development
characterising these phenomena before they wholly disappear
from the scene of their activity. Further downwards, but above
the thermal region which is now in full activity, there are
several reservoirs filled with hot water, in the depths of which
the older geyser mouths may still be seen through the con-
tinually increasing deposits of silica which have accumulated
in the course of ages. These springs, which recur in many
parts of Iceland, and are especially remarkable at Reykir,
are characterised by extreme beauty. In the depths of the
clear unruffled blue waters of these basins, from which rises
a light vapour, the dark outlines of what once formed the
mouth of a geyser may be faintly traced amid the fantastic
forms of the white stalactite walls. Nowhere can the beau-
tiful greenish blue tint of water be seen in greater purity than
in these springs.

A few remarks on the causes from which they are derived
will hardly be superfluous.

Chemically pure water is not colourless, as is usually sup-
posed, but naturally possesses a pure bluish tint, which is only
rendered visible to the eye when the light penetrates through
a stratum of water of considerable depth. That such is the
fact may easily be shown by taking a glass tube, 2 inches wide
and 2 metres long, which has been blackened internally with
lamp-black and wax to within half an inch of the end, the
latter being closed by a cork. Throw a few pieces of white
porcelain into this tube, which, after being filled with chemi-
cally pure water, must be set vertically on a white plate, and
looking through the column of water (of two metres) at the
pieces of porcelain, which can only be illumined from below by
white light, we shall observe that the objects will, under these
circumstances, acquire a pure blue tint, the intensity of which
will diminish in proportion as the column of water is shortened,

OF ICELAND. 359

so that the shade of colour becomes at length too faint to be
perceived. This blue coloration may also be recognised when
a white object is illuminated through the column of water by-
sunlight, and seen at the bottom of the tube through a small
lateral opening in the black coating. The blue tint so fre-
quently observed in water cannot, therefore, be regarded as in
any way strange. The question rather arises, why this blue colour
is not seen everywhere, and why it should not occur in many
seas? why, for instance, the lakes of Switzerland, the waters of
the geysers in Iceland, and in the South Sea Islands, should
exhibit every shade of green, whilst the waters of the Medi-
terranean and Adriatic are occasionally of so deep a blue as
to vie with indigo? These questions are easily answered, since
clearness and depth are the primary if not the sole require-
ments for imparting to water its natural colour. Where these
fail, the blue tint will likewise be wanting. The smallest
quantity of coloured elements which the water may take
from the sand or mud of its bottom, the smallest quantity of
humus held in solution, the reflection of a dark and strongly
coloured bottom, are all sufficient to disguise or alter the colour
of water. It is well known, that the yellowish red colour of
the waters which traverse the lower group of the trias form-
ations depends upon hydrated oxide of iron, contained in the
mud of the variegated sandstone. From a similar cause, the
vast glacier streams of Iceland, which, in these desolate
regions where there are neither roads nor bridges, the traveller
finds, to his discomfort, that he must ford — are rendered opaque
and milk-white from the detritus of dark volcanic rocks,
which, crushed into a white powder by the overwhelming
mass of the descending glaciers, are carried to the sea, in
the form of white mud and sand, and again deposited there in
vast deltas.

In like manner, the natural colour of the small lakes in the
marshy districts of Northern Germany is concealed by the
black tint imparted by the dissolved humus derived from the
turf. These waters often appear brownish or black, like the
water in most of the craters of the Eifel and Auvergne, where
the sombre volcanic rocks obstruct the reflection of the incident
light. It will, therefore, easily be understood that it is only

360 PSEUDO-VOLCANIC PHENOMENA

where these disturbing influences do not exist that the colour
of the water will be seen in all its beauty. Amongst the
places at which this requirement is most completely fulfilled,
we may especially instance the Blue Grotto at Capri, in the
Gulf of Naples. The sea is there most remarkably clear to a
very great depth, so that the smallest objects may be distinctly
seen on the light bottom at a depth of several hundred feet-
All the light that enters the grotto, the entrance of which is
only a few feet above the level of the sea, in the precipitous
rock opening upon the surface of the water, must penetrate
the whole depth of the sea, probably several hundred feet,
before it can be reflected into the grotto from the clear bottom.
The light acquires, by these means, so deep a blue coloration
from the vast strata of water through which it has passed,
that the dark walls of the cavern are illumined by a pure
blue radiance, and the most differently coloured objects below
the surface of the water are made to appear bright blue.

An equally remarkable example of this fact presents itself
in the glaciers of Iceland as well as in those of Switzerland,
which shows that water does not lose its original colour even
when in a solid condition. At the distance of many miles, the
eye may distinguish, on the flat heights of the " Jokull," the
boundaries that separate the bluish ice of the glaciers from
the white inaccessible plains of snow that rise to the summit
of these mountains. On a closer examination of these glaciers,
one is surprised to observe the purity and transparency of the
ice, which often appears to be wholly free in large masses from
vesicles of air and foreign admixtures, whilst its vast fissures
and cavities are coloured all shades, from the lightest to the
darkest blue, according to the thickness of the strata through
which the light has penetrated.

The blue tint of the cloudless and vapoury atmosphere is
probably dependent on similar phenomena, if we are justified
in concluding, from the colour of solid and fluid water, that
aqueous vapour has a similar colour. On considering all these
facts, we can scarcely doubt for a moment that the blue colour
of water is a peculiar and not accidental characteristic of that
substance. This natural colour of water will also afford us an
easy explanation of a light green tint which is even more

OF ICELAND. 361

strongly manifested in the crystal-like siliceous springs of
Iceland than in the lakes of Switzerland; for the yellow colour
derived from traces of hydrated oxide of iron, in the siliceous
sinter walls surrounding the water, blends with the original blue
to produce the same greenish tint, which, in the Swiss lakes, is
derived from the yellow bottom; — the most different rocks
experiencing a superficial decomposition from the continued
action of water, and becoming tinged with yellow by the for-
mation of hydrated oxide of iron. Hence it will be easily
conceived that the blue, which continues to increase in intensity
with the increased depth of the strata of water, may obliterate
the action of this yellow reflection, and thus either weaken, or
wholly destroy, this greenish tint. The green grotto on the
shores of Capri affords a most striking proof of this fact. The
green colour, which is produced by the reflection, at an in-
considerable depth of water, from the yellowish limestone con-
stituting the bottom and the walls of the grotto, illuminated by
the light from without, wholly disappears in the enormous
depths of the water of the blue grotto; there a pure blue
colour takes the place of the green, observed in the shallower
cavern, although the water and rocks are the same in both cases.

We have already shown, that these siliceous thermal
springs, having so high a temperature, combine all the require-
ments necessary for passing through the various phases which
characterize geyser formations. In order, therefore, fully to
understand these phenomena it only remains to revert to the
origin and the mode of formation of the alkaline siliceous
springs, to which the deposits of siliceous tuff owe their origin.
It will be necessary to start from some definite point, and I
will, therefore, make choice of the water of the Great Geyser,
whose composition corresponds in every essential point with that
of all the other siliceous tuff springs of Iceland.

The analysis was made in my own laboratory by Dr. Sand-
berger, from a specimen which I took from the basin of the
geyser immediately after an eruption in the beginning of July,
1846. As Dr. Sandberger will himself give a more detailed
account of his experiments, I will limit myself to a mention
of the results of his analysis, the accuracy of which I had
Beveral opportunities of testing. I subjoin the analysis lately

362 PSEUDO-VOLCANIC PHENOMENA

made by Damour of the same water, in order that my readers
may compare the two.

Water of the Great Geyser.

Sandberger. Damour.

Silica 0-5097 0*5190

Carbonate of soda 0*1939 0*2567

Carbonate of ammonia* .... 0*0083

Sulphate of soda 0*1070 0*1342

Sulphate of potash 0*0475 0-0180

Sulphate of magnesia .... 0*0042 0*0091

Chloride of sodium 0*2521 0*2379

Sulphide of sodium 0*0088 0*0088

Carbonic acid 0*0557 0*0468

Water 998*7695

1000*0000 1000*0000

In a letter to Berzelius, dated 3rd of November, 1846, and
which has been partly published in the German Journals, I
have given a short notice of the results of my experiments and
observations on the origin of the Icelandic thermal springs,
from which I make the following extract:

" An investigation of the phenomena of the springs at the
place of their origin readily leads the observer to the conviction,
that the same chemical activity, by which the scene of these
stupendous natural phenomena is characterised, must have
played an important part in the formative process of older
rocks. As a general fact it may be admitted that none of
the rocks of which the island is formed, resists the decom-
posing action of the thermal waters. Wherever these waters
break forth, whether in a liquid or vaporous form, whether
in the most ancient structures of the island, as in palagonite
tuff, or in clinkstone and trachyte, the next in order of suc-
cession,— in the older trap, which has been thrown out in veins
after the clinkstone period, and has spread through the tuff,
forming vast strata, — in the basaltic eruptions which followed
the last-named period, — or, finally, in the most recent lava

The ammonia was obtained by adding recently heated hydrate of potash,
and distilling the water into a receiver containing hydrochloric acid ; the last
having previously been purified from all admixture with ammonia, by being
distilled with chloride of platinum. The portion of water analyzed was
brought from Iceland in a glass tube hermetically closed.

OF ICELAND. 363

eruptions, a decomposition may everywhere be observed, which,
agreeing in its fundamental type, has probably, under the co-
operation of secondary influences, given rise to all the various
products which are found in the craters of Hecla and KrafTa, as
well as in the vicinity of the Great Geyser and the different
sufEones.

" The different siliceous compounds formed at the bottom of
the springs are converted by the action of the thermal water
into acid and basic silicates. The former are dissolved by the
water, whilst the latter constitute an insoluble stratum of clay,
whose gradual transition into the original rock may be clearly
traced at certain points, but more especially at the confines of
the fumerole activity. The distribution and size of these
secondary argillaceous strata stand in direct relation to the
magnitude of the thermal phenomena, and with the tem-
perature of the water, which, at a considerable depth, is not
unfrequently upwards of 100° C. The soluble silicates which
proceed from these fissures are brought to the surface with the
thermal water, giving rise, where it is left to evaporate freely
to the formation of siliceous sinters and opals.

" Two gases, namely, sulphuretted hydrogen and sulphurous
acid, occur together with sulphur, as the important accom-
paniments of these vapour and boiling springs; and impart a
somewhat different character to these simple processes, by
calling forth a series of secondary products."

Subsequently to the period when the above was written,
M. Damour has made known (in the Annales de Chimie et de
Physique, 1847,) that heated mesotype is partially dissolved
by boiling water, silicate of soda being formed; and he thus
furnishes us with a very interesting addition to the above
remarks. Amongst the rocks which I have indicated as subject
to this decomposition, palagonite occupies the principal place.
The geognostic relations of the Great Geyser (the largest
source of siliceous tuff in Iceland), which lies at the foot of an
erupted trachytic clinkstone, appear to indicate the existence
of a special connexion between these rocks and the siliceous
tuff deposits, and such has been supposed by Damour to be
the case, but, independently of the circumstances, that other
springs of this kind, as those lying on the north-west of Ilecla

364 PSEUDO-VOLCANIC PHENOMENA

at the foot of the Kandakamba, standing in precisely similar
relations to the clinkstone, are by no means characterised by
their siliceous tuff formations, it may be also determined with
certainty from the secondary decompositions of the bottom of
the spring, that it consists of palagonite tuff, which exercises
the most essential influence on the formation of the geysers.

In order to show the truth of these facts, it will be neces-
sary to enter somewhat more fully into the consideration of the
decompositions experienced in palagonite by the action of hot
water, carbonic acid, and sulphuretted hydrogen. On digesting
pulverised palagonite in a strong and closely corked glass vessel
with distilled water for some hours, at a heat of 100° or 106° C,
silicic acid, potash, and soda are dissolved. 1000 grammes of
water after 12 hours' digestion yield, in this manner, a solution
containing the following proportions: —

Silica 0-03716

Soda 0-00824

Potash 0-00162

0-01702 gram.

On allowing water, saturated with carbonic acid, to act on
the pulverised mineral, all the constituents, with the exception
of alumina and oxide of iron, will be dissolved in the form of
bi-carbonates. 1000 grammes of this water, after four hours'
digestion, yielded the following constituents: —

Silica

.... 009544

Bi-carbonate of lime

.... 016893

„ magnesia

.... 005333

„ soda

.... 006299

„ potash....

.... 0-00189

0-38268 gram.

On heating pulverised palagonite in a similar manner, for
ten hours, in water saturated with sulphuretted hydrogen,
sulphide of iron was formed, and the solution contained, for
1000 grammes, the following constituents: —

Silica 0-1175

Sulphide of calcium .... .... .... 0*2748

„ magnesium .... .... .... 0*0727

„ sodium .... .... .... 00438

„ potassium .... .... .... 0*0410

0*5498 gram.

OF ICELAXD. 365

We see from the relations existing; among these salts them-
selves and with the silica, that the constituents of palagonite
take very different parts in the decomposition which is induced
by hot water, carbonic acid, and sulphuretted hydrogen respec-
tively; whilst, as we have already seen, this mineral is entirely
dissolved in hydrochloric and sulphurous acids, except a small
quantity of silica left as a residue. The alkaline siliceous springs,
in which there is a smaller quantity of this volcanic gas, assume,
consequently, a very different character from the waters of the
suffiones ; since it is evident, that the composition of the water,
and the nature of the argillaceous deposits produced from these
actions, must stand in a definite relation to the greater or
smaller resistance opposed by the separate constituents of pala-
gonite to the action of the weaker volcanic acids, that is to
say, to the water, carbonic acid, and sulphuretted hydrogen
gas. I must reserve the further development of the relation in
which the clay of the geyser stands to the constituents of the
water of the geyser; since the necessary data can only be
attained by an examination of the residuary matters left by
palagonite, after the long-continued action of water, carbonic
acid, and sulphuretted hydrogen.

I think I may, however, pass over the last relations, since
the results already obtained by a decomposition of palagonite are
sufficient to give an explanation of those processes, to which the
constituents of the alkaline siliceous springs owe their origin;
and which, moreover, manifest themselves as the original causes
of all the phenomena presented in the different phases of the
development of geyser formations.

When the alkaline silicates, removed by the heated water
from the palagonite, are brought into contact with carbonic,
hydrochloric, and sulphurous acids, (the latter of which is
formed by the oxidation of the sulphurous acid through the
oxide of iron in the palagonite,) these alkalies must be converted
into carbonates, sulphates, and chlorides, whilst the silicic acid
remains dissolved in the alkaline carbonates and in the water,
and is partially separated from them, by evaporation, as siliceous
tuff, — a fact already observed by Black, in 1792.

The action of the carbonic acid is not, however, limited to
the alkalies taken up by the water, but is directly extended,

366 PSEUDO-VOLCANIC PHENOMENA

as may be seen in our second analysis, to the substance of the
palagonite, since it not only gives rise to a solution of silica
in water and alkaline carbonates, but also forms acid carbonates
of magnesia and lime. The fact that only traces of the first
named of these earths are found in the geyser water, may be
easily explained from the circumstance, that the acid carbonate
of lime is decomposed by boiling into carbonic acid and an
insoluble neutral salt, or is decomposed in the same manner
by the alkaline silicates, whilst an alkaline carbonate is formed.
Magnesian salts, on the other hand, whose solution, when very
much diluted, undergoes, as is well known, only an incomplete
precipitation, must, therefore, in accordance with the analysis,
occur in traces in the water of the geysers.

The products of sulphuretted hydrogen on palagonite, which,
as I have before observed, furnish the key to an explanation of
the formation of pyrites in the clay of the geysers and fume-
roles, experience a similar decomposition by carbonic acid.
Sulphuretted hydrogen, which is never absent from the thermal
limits of the geysers, escapes as a gas, whilst carbonate of lime
and, to a certain degree, carbonate of magnesia are precipitated.
The ultimate product of these reactions is silicic acid dissolved
in water and alkaline carbonates, to which alkaline sulphides, as
the constant accompaniments of these siliceous springs, are
added, when the carbonic acid forms a less strongly character-
ised constituent. We here meet with a simple explanation of
the origin of siliceous sinters.

The relation of the potash to the soda in the geyser water
is totally different from that which exists in palagonite. The
quantity of the former is most strikingly inferior to that of
the latter in almost all these waters. Although, in the above
analysis, the quantity of the constituents sought is too small,
and the duration of the decomposition of palagonite much too
short, to lead to an accurate determination of the relation of
the bases to silicic acid, we yet can find that the soda is
dissolved from palagonite by pure water, or carbonic acid, in a
much larger proportion than the potash; and even in a much
larger relation, by the carbonic acid, than corresponds to the
composition of the geyser water. We may, therefore, easily
understand that when the three volcanic gases are simul-

OF ICELAND. 367

taneously exercising a decomposing influence, these alkalies
may easily be dissolved to the relative extent in which they
occur in the waters of the geyser.

I have already remarked that all the volcanic rocks of
Iceland experience, under the influence of water and volcanic
gases, a similar although much slower decomposition than pala-
gonite. This is most strikingly observed in the volcanic incor-
porations of the palagonite tuff, which are gradually, although
slowly and with difficulty affected by the decomposing action
of the fumeroles. On comparing the quantity of the argil-
laceous layer produced by the decomposition of the palagonite
tuff with the extent of the contiguous siliceous tuff formations,
it will be evident that they stand in a mutually controlling
relation to one another, and that it is especially palagonite
which constitutes the peculiarity of Icelandic thermal pheno-
mena. I think, I may even venture to assert, that geyser and
siliceous tuff springs (all other conditions being the same) are
connected with the occurrence of palagonite tuff, or with struc-
tures similar to palagonite.

If I may be permitted, I will conclude the present memoir
(in which I have collected some of the data of a larger work on
which I am at present engaged) by extracting a few general
considerations from the letter to Berzelius to which I have
already alluded.

A careful study of the actions of fumeroles, as it may be
observed on so large a scale in the fissures which maintain
the connexion existing between the focus of volcanic forces
and atmospheric phenomena in Iceland, leads us imperceptibly
back to the theatre of that great volcanic catastrophe, by which
the trap was upheaved through the tuff and clinkstone rocks,
and distributed in vast strata through them.

I will here only speak of a few facts which may be classed
amongst the most common in Iceland. Where the clinkstone
and older trap break through the tuff, and still more, where the
older trap penetrates in veins through the clinkstone, the pene-
trated rock exhibits a fused and disintegrated appearance, by
which it assumes characteristics resembling those of obsidian or
pitch-stone. On tracing the horizontal strata which have been
erupted from these veins, at their points of contact with the

368 PSEUDO-VOLCANIC PHENOMENA

tuff, we soon lose sight of all appearance of fusion or the action
of fire. On examining the tuff more carefully, we are astonished
to discover that the main body of it is composed of an easily
fusible hydrated silicate, which has preserved its original cha-
racter and its normal proportion of water, notwithstanding its
immediate contact with vast strata of trap which have pene-
trated in a o'lowino: and fused condition. Still more singular
are the innumerable vesicular cavities in the amygdaloid alter-
nating with the tuff and trap, which are lined and not un fre-
quently completely filled up with quartz, chalcedony, calc-spar,
zeolites and other hydrated silicates.

How could these hydrated compounds have arisen from
plutonic mineral masses, closely invested by glowing hot trap,
or how could they be preserved unchanged ?

These questions are explained by the phenomena of the
present active fumeroles. So porous a structure as tuff must
necessarily be penetrated by moisture throughout its entire
mass, and thus be filled with strata and currents of water,
which must have given rise to the phenomena, which appear
at first sight so inexplicable, since the great catastrophe of
the elevation of trap was of a subsequent date. Where the
hot fluid trap came in contact with an inconsiderable cooling
surface, as at the mouth of the vein, the water has yielded and
been expelled by the action of the constantly renewed stream
of fire, which would then prove more than sufficient to fuse
the anhydrous rock. On the other hand, where the fluid rock
spread out from these veins in far extended horizontal strata,
and penetrated through the upheaved tuff, the further develop-
ment of the igneous action on the surrounding rock must have
been arrested by the increased production of aqueous vapour;
the great specific heat of the water and the considerable latent
heat of its vapour, making a greater elevation of temperature
impossible in the tuff strata. The magnitude and extent of
these igneous actions, at considerable distances from the veins,
stand in a direct relation to the preponderance of the trap strata
over the tuff deposits.

If this view be correct, the fumerole actions, which are
owing to the vast development of aqueous vapour, must neces-
sarily be found to occur in the tuff and trap masses which

OF ICELAND. 369

served as a focus for their activity. Appearances fully corrobo-
rate these views. For the amygdaloid which is so rich in
hydrated silicates, and alternates with tuff and trap, presents
a faithful picture of the phenomena of decomposition which may
still be met with in full activity in the Icelandic suffiones and
thermal springs. In these vast amygdaloid rocks, which present
the mineralogist with one of the richest and most interesting
sources of mineral species, we may easily recognise the origin of
the metamorphosis of the rock, which occurs at the points of
contact between the trap and the tuff strata. Their principal
mass consists of a ferruginous or siliceous clay, which occurs as
a product of fumeroles still active, and after passing through
innumerable phases of decomposition merges into the perfectly
undecomposed strata of both these species of rock. Pyrites,
chalcedonous quartz, and opal, which are so characteristic of
fumerole structures, are scarcely ever absent, whilst everything
justifies the assumption, that this remarkable penetration of
aqueous and plutonic formations is to be ascribed to a great
manifestation of fumerole activity which has immediately fol-
lowed the catastrophe of eruption; the original rock being
converted into aluminous amygdaloid, by a separation of its con-
stituents, into soluble and insoluble silicates, in a manner analo-
gous to what may still be observed. By these means, the plastic
clay might easily be filled, through the agency of vapours and
gases, with those innumerable vesicular cavities in which the
products of crystallization of the soluble silicates which pene-
trate the clay are traced as supplementary constituent parts.
Whether I shall succeed with the materials at my command in
explaining the local conditions, on which the formation of the
various silicates depends, is a question that can only be decided
by an experimental investigation, at the place and focus of the
phenomena treated of in these observations.

The relations of superposition of the Doubly refracting
Iceland spar, which have been so carefully investigated by M.
Descloizeaux, and which appear so inexplicable on a first exami-
nation, can easily be explained by considering the fumerole actions
from the above point of view. In like manner all the difficulties
attending the explanation of the recent and unexpected discovery
of the occurrence of petrified infusoriae in the decomposing crust

2 B

370 PSEUDO-VOLCANIC PHENOMENA OF ICELAND.

of volcanic rocks, may easily be explained when we consider
that the mass of snch structures, converted by fumerole action
into clay at its surface, and then subsequently solidified by
the infiltration of soluble silicates, might easily contain within
itself the conditions necessary for the preservation of the micro-
scopic organisms, whose petrified remains are still found within
these strata.

THE END.

LONDON :

PRINTED BY T. R. HARRISON,

ST. MARTIN'S LANE.

^ 5^ I ;

jjfflffiflBi

University of British Columbia Library

DUE DATE

FORM 310

UNIVERSITY OF B-C. LIBRARY

3 9424 02308 0333

Illi

i,F H ill M

I lillllPllll

hi ! I ijljil'

ram

■BB1

, h -:i!J

K«M

I li 11

^^^^^H!i!ilii

BHH

WttM

ii!,;

i' lilil

ran j>'
■HP

HHl

! !

■

ill

i:,rl > ■ , ■

Internet Archive Audio

Featured

Top

Images

Featured

Top

Software

Featured

Top

Books

Featured

Top

Video

Featured

Top

Mobile Apps

Browser Extensions

Archive-It Subscription

Save Page Now

Full text of "Chemical reports and memoirs ..."

See other formats